The dialogical self of migrants: exploring the migration ...

THE DIALOGICAL SELF OF MIGRANTS:

EXPLORING THE MIGRATION EXPERIENCE THROUGH SELF-NARRATIVES

Mariel Sanchez-Rockliffe

Faculty of Health, Arts and Design

Swinburne University of Technology

Australia

A Report of an investigation submitted

as partial requirement for the degree of

Doctor of Psychology (Counselling Psychology)

December 2015

i

Declaration

I declare that that this thesis does not incorporate without acknowledgment any

material previously submitted for a degree in any university, or other educational

institution; and that to the best of my knowledge and belief it does not contain any

material previously published or written by another person except where due reference

is made in the text.

I further declare that the ethical principles and procedures specified by the

Psychology Department’s document on human research and experimentation have been

adhered to in the preparation of this report.

Mariel Sanchez-Rockliffe

ii

Abstract

Hermans’ theory of the dialogical self and the self-confrontation method (SCM)

are used to study the effect of migration on the self. A radical change in the social

environment, in this case migration, is shown to be reflected in the structure of the

dialogical self. Relationships between migrants’ I-positions relating to life in their

countries of origin and in Australia are explored.

The study is based on a sample of 38 migrants to Australia who perceived

themselves as having two identities, one for their country of origin, the other for their

new country. Participants were interviewed under the procedures of the SCM to elicit a

total of 1500 valuations (narratives) relating to their I-positions. These were scored

along the standard four SCM dimensions: self-enhancement (S), union-with-the-other

(O), positive affect (P) and negative affect (N). Participants also provided demographic

data, and self-reports of psychological well-being and distress.

The first study employed cluster analysis to group participants into four types.

“Progressives” steadily moved towards a valued goal; “Low-stables” experienced little

affective change, remaining at a lower level than “High-stables”, who likewise changed

little but were at a higher level of well-being. “Regressives” were the opposite of

“Progressive”. The types were then characterised in terms of self-confrontation indices,

demographics and psychological self-reports, and their differences identified. In the

second study participants were classified according to the nature of their valuations and

a transition matrix constructed linking the major themes of their valuations for country

of origin and for Australia. The transition matrix was statistically tested for the null

hypothesis that migration had no effect on I-position. The dialogical self was treated as

a hypothetical probability distribution of themes of valuations whereby the SCM is a

technique for generating a sample of valuations. The properties of the underlying

iii

distribution could then be inferred, as in standard sampling theory. The third study was

idiographic, analysing case studies representing each of the participant types identified

by the first study.

The main finding is the overall stability of the thematic content of valuations.

Where differences occur, they are intelligible. For example, many respondents seemed

to be responding positively to the peaceful, liberal social conditions they find in

Australia. The research corroborates previous research that identifies two modes of

biculturalism: compatible and oppositional. Low-stable and High-stable migrants

viewed their heritage and host cultures as complementary, whereas Progressive and

Regressive migrants viewed them as discrepant and conflicting. Many migrants had I-

positions with unambiguously different thematic content. This is evidence for Hermans’

dialogical view of the self and against the hypothesis of an underlying integrated self.

Most migrants showed self-enhancement (S) in their Australian than in their country of

origin I-positions, whereas union-with-the-other (O) did not differ significantly,

suggesting that migration accommodates self-enhancement (S) proportionately more

than union-with-the-other (O). Migrants varied in their psychological states in a

plausible and consistent way. In general, Progressive and High-stable migrants were

lowest on distress and highest on well-being; Regressive migrants were the opposite;

and the Low-stable migrants were intermediate. Lastly, good matches were found

between the case studies and the participant types they represented.

iv

Acknowledgments

It is pleasure and deep gratitude that I acknowledge the help, support and advice

extended to me in completing this thesis.

I am indebted to Swinburne University for the opportunity to undertake it, and for

their financial support and encouragement in attending international conferences

pertinent to my research.

To my supervisor, Glen Bates, I extend my deepest thanks for his guidance and

support. Without his patience and understanding this thesis would surely never have

been completed.

To my friend and mentor, James Symons, I extend my thanks for his untiring

efforts and advice in developing and testing hypotheses relating to the thesis. To my

husband, I extend my gratitude for his assistance in data management and editorial

matters. And to both I offer my sincere appreciation for their keen critical insights,

which were all the more valuable as they were drawn from disciplines outside my own.

Lastly, I extend my thanks to all those—friends, family and colleagues too

numerous to mention—who have stood by me.

v

Table of Contents

Declaration i

Abstract ii

Acknowledgments iv

Table of Contents v

List of Tables vii

List of Figures ix

CHAPTER 1: Introduction and Overview 1

CHAPTER 2: Self and Identity 3 2.1 James’ Theory of the Self 3 2.2 Symbolic Interactionism 6 2.3 Identity 11 2.4 The Dialogical Self 24 2.5 Summary 34

CHAPTER 3: Migration and Acculturation 36

3.1 Australian Migration 36 3.2 Acculturation 44 3.3 Acculturation Outcomes 58 3.4 Individualism versus Collectivism 67 3.5 Summary 73

CHAPTER 4: The Dialogical Self and Acculturation 75 4.1 Globalisation and the Illusion of Stability 75 4.2 A Dialogical Theory of Acculturation 82

4.3 Cultural Continuity 87 4.4 Research Aims and Questions 94

CHAPTER 5: Research Methodology 99 5.1 Participants 99 5.2 Self-report Measures of Psychological Distress and Well-being 104 5.3 Self-confrontation Method (SCM) 114 5.4 Procedure 118

5.5 Treatment of the Interview Data 121 5.6 Comparison with Previous Studies 127 5.7 Personal Reflections on the Experience of Migration 130

CHAPTER 6: Study 1 – Identification and Comparison of Migrant Groups 135 6.1 Preliminary Analysis 135

6.2 Distinguishing I-positions 143

6.3 Analyses of Participant Type 157

6.4 Comparison of Participant Types 170

vi

6.5 Summary 177

CHAPTER 7: Study 2 – The Dialogical Self as a Probability Distribution 180 7.1 The Dialogical Self as a Probability Distribution of Themes 180 7.2 Hypotheses 184 7.3 Analysis Based on Major Themes 185 7.4 Analysis Based on Self-confrontation Indices 195 7.5 Summary 199

CHAPTER 8: Study 3 – Case Studies 200 8.1 Euclidean distance 200 8.2 Case study: The ‘Progressive’ Migrant (Type 1) 204 8.3 Case study: The “Low-stable” Migrant (Type 2) 211

8.4 Case study: The “High-stable” Migrant (Type 3) 218 8.5 Case study: The “Regressive” Migrant (Type 4) 227 8.6 Summary 233

CHAPTER 9: General Discussion 236 9.1 Summary of the Empirical Work 236 9.2 Main Findings 239 9.3 Directions for Future Research 265 9.4 Conclusions 271

Glossary 273

References 275

vii

List of Tables

Table 3.1 Migrant stock: selected countries of large-scale immigration 39

Table 5.1 Characteristics of the sample, by gender 101

Table 5.2 Means and standard deviations for Years in Australia, Age and

Individualism, by gender 102

Table 5.3 Number of participants, by Hofstede’s individualism index 103

Table 5.4 Dimensions of Ryff’s Psychological Well-being Scale (PWB) 110

Table 5.5 Questions of the Self-confrontation method 116

Table 5.6 Composition of indices in the SCM 122

Table 5.7 Composition of major themes 125

Table 5.8 Empirical studies using the Self Confrontation Method 128

Table 5.9 Demographic difference between minority and majority cultures

for selected studies 128

Table 6.1 Means and standard deviations for the self-report measures 136

Table 6.2 Correlations among self-report measures 137

Table 6.3 Correlations among self-confrontation indices 139

Table 6.4 Correlations among self-reports and self-confrontation indices 141

Table 6.5 Means and standard deviations of self-confrontation indices 144

Table 6.6 Number of participants by cluster: country of origin 149

Table 6.7 Mean self-confrontation indices for country of origin I-position,

by cluster in country of origin 151

Table 6.8 Number of participants by cluster: Australia 154

Table 6. 9 Mean self-confrontation indices for Australian I-position, by

cluster in Australia 155

Table 6.10 Number of participants, by cluster in country of origin and in

Australia 157

Table 6.11 Mean self-confrontation indices for the I-position, by location

and participant type 159

Table 6.12 Tabulation of demographic characteristics, by participant type 171

Table 6.13 Mean and standard deviation of demographic characteristics, by

participant type 172

Table 6.14 Mean psychological distress, by participant type 173

Table 6.15 Mean psychological well-being, by participant type 174

Table 6.16 Mean self-confrontation indices for country of origin I-position,

by participant type 175

Table 6.17 Mean self-confrontation indices for the Australian I-position, by

viii

participant type 176

Table 7.1 Rules for assigning valuations to major themes 186

Table 7.2 Number of valuations, by participant type and major theme 187

Table 7.3 Thematic profiles showing “important” major themes, by

participant and location 190

Table 7.4 Average differences in proportions of major themes 194

Table 7.5 Thematic transitions 197

Table 7.6 Average differences in thematic scores (Australia minus country

of origin) 198

Table 8.1 Valuations with major themes, by location and period: Yoana 210

Table 8.2 Valuations with major themes, by location and period: Juan 217

Table 8.3 Valuations with major themes, by location and period: Luisa 225

Table 8.4 Valuations with major themes, by location and period: Carla 234

ix

List of Figures

Figure 6.1 Agglomeration schedule: cluster in country of origin 147

Figure 6.2 Dendrogram of 2-cluster solution: cluster in country of origin 148

Figure 6.3 Sum of squares from centroid: cluster in country of origin 150

Figure 6.4 Mean self-confrontation indices for country of origin I-position,

by cluster in country of origin 151

Figure 6.5 Agglomeration schedule: cluster in Australia 152

Figure 6.6 Dendrogram of 2-cluster solution: cluster in Australia 153

Figure 6.7 Sum of squares from centroid: cluster in Australia 154

Figure 6.8 Mean self-confrontation indices for Australian I-position, by

cluster in Australia 156

Figure 6.9 Mean self-enhancement for country of origin and Australian I-

position, by participant type 161

Figure 6.10 Mean union-with-the-other for country of origin and Australian

I-position, by participant type 162

Figure 6.11 Mean positive affect for country of origin and Australian I-

position, by participant type 163

Figure 6.12 Mean negative affect for country of origin and Australian I-

position, by participant type 164

Figure 6.13 Differences in self-confrontation indices for country of origin

and Australian I-positions, by participant type 165

Figure 6.14 Mean self-confrontation indices, by participant type 167

Figure 7.1 Hypothetical probability distribution of valuation themes:

oblique view and plan view (inset) 182

Figure 8.1 Euclidean distance between profiles for country of origin and

Australia based on mean self-confrontation indices of

valuations 202

Figure 8.2 Euclidean distance between profiles for country of origin and

Australia based on proportions of major themes of valuations 203

Figure 8.3 Proportion of valuations, by major theme and location:

Progressive migrants and Yoana 206

Figure 8.4 Mean indices of valuations, by location: Yoana 207

Figure 8.5 Mean indices of valuations, by location and period: Yoana 208

Figure 8.6 Proportion of valuations, by major theme and location: Type 2

participants and Juan 213

Figure 8.7 Mean indices of valuations, by location: Juan 214

Figure 8.8 Mean indices of valuations, by location and period: Juan 215

Figure 8.9 Proportion of valuations, by major theme and location: Type 3

participants and Luisa 220

x

Figure 8.10 Mean indices of valuations, by location: Luisa 221

Figure 8.11 Mean indices of valuations, by location and period: Luisa 222

Figure 8.12 Proportion of valuations, by major theme and location: Type 4

participants and Carla 229

Figure 8.13 Mean self-confrontation indices, by location: Carla 230

Figure 8.14 Proportion of valuations, by location and period: Carla 231

1

CHAPTER 1: Introduction and Overview

The overarching aim of the thesis is to strengthen the empirical evidence base of

the theory of the dialogical self. To this end, it explores the effect of migration on the

self within the theoretical context of Hubert Hermans’ theory of the dialogical self. It

does so by means of Hermans’ self-confrontation method (SCM), an investigative tool

widely used to access and explore the dialogical self. It shows how a radical change in

the social environment—in this case migration to a new country—is reflected in the

repertoire of I-positions of migrants. In so doing, it develops the nomothetic aspect of

the SCM by identifying significant associations between migrants’ I-positions relating

to life in their countries of origin, and their I-positions relating to life in Australia.

The thesis is structured as follows. Chapters 2, 3 and 4 review the relevant

literature. Chapter 2 (Self and Identity) explores the psychological concept of the self.

This is central to the thesis as it is what is accessed and measured by the SCM. The

thesis discusses current theories, including the theory of the dialogical self, and explores

the proposition that people possess multiple I-positions. Particular attention is paid to

previous studies of migrants.

Chapter 3 (Migration and acculturation) describes psychological theories of how

and why people migrate, and what happens to them at a psychological level when they

do. It explores the proposition that some migrants preserve their former cultural identity

while taking on a new one for the their country of settlement; it is “integrated” migrants

such as these who were the subject of the current research.

Chapter 4 (The Dialogical Self and Migration) brings together the previous two

chapters — one on theories of the self, the other on theories of migration and

acculturation. It shows how the theory of the dialogical self applies to migration and the

acculturative stress that it entails. From this synthesis, research questions are developed.

2

Chapter 5 (Method) explains how the data were gathered from guided interviews

of 38 migrants. The data collection process is described and the analytical procedures

set out. The data were analysed in two ways, termed Studies 1. 2 and 3, that differed in

approach.

Chapter 6, 7 and 8 present the results of three separate studies using the same data

set. Study 1 uses cluster analysis as a data reduction technique to group participants into

four types, and analyses their characteristics. Study 2 classifies participants into groups

on the basis of their “valuations” (to use Hermans’ terminology, which is defined later).

By treating the dialogical self as a hypothetical probability distribution of themes of

valuations, it permits formal hypothesis testing of transition matrices linking the

participants’ group in country of origin, with their group in Australia. Study 3 presents

four case-studies identified in the course of the preceding studies, one to exemplify each

of the four migrant types identified under Study 1.

Lastly Chapter 9 discusses the results of all three studies. In terms of external

validity, the thesis had the advantage of being based on a non-clinical respondent

sample; but its conclusions are limited to migrants of a particular kind (“integrated”),

and particular countries of origin and settlement. More work would be needed to extend

the validity of its conclusions. In terms of method, in Study 2 the thesis introduced a

novel way of characterising the dialogical self: as a frequency distribution of valuations.

A frequency distribution of this kind does not capture the full richness of the dialogical

self nor the I-positions that compose it, but it does offer powerful insights for exploring

the changes in the dialogical self that may be of value to future researchers, particularly

in the area of formal hypothesis testing.

3

CHAPTER 2: Self and Identity

This chapter identifies the antecedents of the theory of the dialogical self and

traces its evolution. It begins with William James’ theory of the self, since it was James

who arguably set the agenda for future social psychological inquiry about self and

identity (Suls, Tesser & Felson, 2000). Next, it discusses contemporary issues of self

and identity: the collective self, multiple selves versus a unified self, and identity—

personal, social, and bicultural. The concept of the dialogical self is then introduced.

Dialogical self theory, it is argued, is a powerful conceptual model for understanding

the process of acculturation in that it addresses weaknesses of previous acculturation

models, principally that of Berry (1990).

2.1 James’ Theory of the Self

James (1890) was not the first thinker to explore the nature of personal identity;

the Ancient Greek philosophers, notably Heraclitus, had pondered the notion. But James

was one of the first to articulate a distinction between the I (or self-as-subject) and the

Me (or the self-as-object)—a notion that came to be known as the duality of the self. For

James, the I was the knower, and the Me was the known, being the empirical aggregate

of everything objectively known about the self. James viewed the I as the subjective self

because it organised and interpreted one’s experiences.

According to Westen (1995), it is in James’ I that the sense of self resides; and it

is in James’ I that individuals experience themselves as the thinkers of their thoughts,

feelers of their feelings, and agents of their actions. Harter (1996) notes that, for James,

the “I-self” also demonstrated more specific types of awareness, such as a concern with

one’s personal continuity through time, and an awareness of the uniqueness of one’s life

experiences leading to the distinctiveness of oneself as a person. Although it was

4

essential for James to posit a knowing “I-self” as the active agent responsible for

constructing the “Me-self”, it is the Me that came to be labelled the self-concept and that

has received most attention in the field of self psychology (Harter, 1988; 1996).

James (1890) used the term “empirical self” to refer to the ways in which people

think about themselves:

The Empirical Self of each of us is all that he is tempted to call by the name of

me. But it is clear that between what a man calls me and what he simply calls mine

the line is difficult to draw. We feel and act about certain things that are ours very

much as we feel and act about ourselves. Our fame, our children, the work of our

hands, may be as dear to us as our bodies are, and arouse the same feelings and

the same acts of reprisal if attacked. And our bodies themselves, are they simply

ours, or are they us? (p. 291)

James’ (1890) definition of the “empirical self”—or “Me”—is broad:

In its widest possible sense, … a man’s Self is the sum total of all that he CAN call

his, not only his body and his psychic powers, but his clothes and his house, his

wife and children, his ancestors and friends, his reputation and works, his lands

and horses, and yacht and bank-account. All these things give him the same

emotions. If they wax and prosper, he feels triumphant; if they dwindle and die

away, he feels cast down—not necessarily in the same degree for each thing, but

in much the same way for all. (pp. 291–292, emphasis original)

This definition is broad enough to encompass most aspects of personhood

(Brown, 1998).

James subdivided his Me into three: the spiritual Me (the inner self or the

psychological self) at the top, the material Me (tangible objects, people, or places

designated as “my” or “mine”) at the bottom, and the social Me in the middle. It is

James’ notion of the social Me that appears to be the forerunner of most psychological

research and theories concerning personal identity. In James’ conception, the social Me

5

consists of the recognition a person receives from others, and hence there exists a

different social Me for each group whose opinion one is concerned about:

… a man has as many social selves as there are individuals who recognise him

and carry an image of him in their mind… But as the individuals who carry the

images fall naturally into classes, we may practically say that he has as many

different social selves as there are distinct groups of persons about whose opinion

he cares. He generally shows a different side of himself to each of these different

groups. (p. 294)

It is through the recognition of others that what he calls “the different sides of

oneself” begin to develop. Thus, by having different people or groups respond to one in

different ways, one begins to acquire various social roles—that is, distinctive patterns of

behaviour that are displayed in, and because of, different social situations.

For James, each of these social selves (for instance, “mother” versus “student”)

has its own set of expectations and behaviours; which are sometimes at variance and

occasionally in conflict due to competing demands from others.

Many a youth who is demure enough before his parents and teachers, swears and

swaggers like a pirate among his ‘tough’ young friends. We do not show

ourselves to our children as to our club-companions, to our customers as to the

laborers we employ, to our own masters and employers as to our intimate friends.

From this there results what practically is a division of the man into several

selves; and this may be a discordant splitting, as where one is afraid to let one set

of acquaintances know him as he is elsewhere; or it may be a perfectly

harmonious division of labor, as where one tender to his children is stern to the

soldiers or prisoners under his command. (p. 294)

For James’, how people think of themselves largely depends on their social roles

(Roberts & Donahue, 1994). Difficulties arise, however, where roles clash (Brown,

1998). People furthermore seek recognition; they affiliate, he argued, not just for the

company but to seek recognition and status, otherwise known as social climbing.

6

A man’s Social Self is the recognition which he gets from his mates. We are not

only gregarious animals, liking to be in sight of our fellows, but we have an innate

propensity to get ourselves noticed, and noticed favorably, by our kind. No more

fiendish punishment could be devised, were such a thing physically possible, than

that one should be turned loose in society and remain absolutely unnoticed by all

the members thereof. (p. 293)

These social selves have come to be referred to as social identities. Deaux, Reid,

Mizrahi, and Ethier (1995) distinguished five types: personal relationships (husband,

wife etc.), ethnic or religious affiliation, political affiliation, stigmatised group, and

vocation. Some, one is born with, others are attained. So for example, a person might in

this way see herself as “mother”, “Hispanic”, “Catholic”, “left-wing” and a “teacher”.

2.2 Symbolic Interactionism

Symbolic interactionism is a conceptual model of the self-concept first conceived

by Cooley (1902) and Mead (1934). It is concerned with the process of socialisation,

that is, how culture is acquired and perpetuated, and how people come to adopt the

values, standards, and norms of the society into which they are born. It posits that the

self-concept, being a product of social interaction with others, is in constant flux. For

both Cooley and Mead the self is fundamentally social in origin. Mead held that social

interaction creates an individual’s sense of self. People learn to take the role of others

and in this way learn who they are and how they should behave. Cooley, with his notion

of the “looking glass self”, held that people obtain a sense of who they are by how

others perceive them. The implication of their theories for the present thesis is that since

society itself is not constant (as is the case for migrants), the change of environment

creates a new self, which arises as a result of the new social commerce in the new

society; migrants experience a social rebirth in fact.

7

Cooley (1902), who gave a sociological perspective to the self, acknowledged that

he was strongly influenced by William James. He developed a conceptual framework

for dealing with the relationship between self and society, which he considered to be

aspects of the same whole. In explaining the origin of the self, he said, “self and society

are twin born … and the notion of a separate and independent ego is an illusion” (p. 8,

cited in Breakwell, 1983). The mind is a social construct, and society a mental

construct, and the relationship between them is best understood by the metaphor of a

“looking glass self”, by which he meant that what one internalises as one’s own is based

on information about oneself that one receives from others (Weinreich, 2003). In other

words, people see themselves as others see them because they learn about themselves

from others, who act as mirrors in which a person sees himself or herself.

James’ idea of the formative effect of the responses of others on the development

of a social self was pursued by social psychologist George Mead, who in a development

of Cooley’s theory elaborated on James’ social self and produced a more extensive

theory of its development (Mead, 1934). Mead differed markedly from James in his

conception of the social self: for James it was one constituent of the self; for Mead it

was the entire self, for the self is wholly a social construct, and is created particularly

through the use of language (Breakwell, 1983). According to Mead (1934, p 135):

the self is something which has a development; it is not initially there, at birth, but

arises in the process of social experience and activity, that is, develops in the

given individual as a result of his relations to that process as a whole and to other

individuals within that process.

Mead (1934) conceptualised the self “as that which can be an object to itself, is

essentially a social structure, and it arises in social experience” (p. 140). The person gets

to know himself only through others; that is, he cannot experience himself directly as

such, “but only indirectly, from the particular standpoints of other individual members

8

of the same social group, or from the generalised standpoint of the social group as a

whole to which he belongs” (p. 138). In this way Mead affirmed the primacy of the

social context—particularly the family—in which an individual is born and socialised

that defines a station in life in terms of kinship relationships, race and class. For Mead

the self-concept arises in social interaction. It is in the course of this socialisation that

people learn to take the “perspective of the other”, to place oneself in the other’s shoes.

According to Mead, the self is defined through linguistic exchanges (that is, symbolic

interactions) within communities. Initially people take on the roles of the significant

others in their immediate world: parents, playmates, and so on. Over time, they extend

their framework to take on the role of the generalised other—their culture. This abstract

and generalised other represents the broader society and culture into which they are

born. Socialisation is conceived as role learning; it forms a person’s identity as a

consequence of social interaction. The acquisition of roles results in the development of

self-conception. In time, the person is able to view the self, as it were, from outside the

self—as mirrored reflections of others. People learn how their culture interprets

gestures, and in this way they develop a self-concept: they see themselves through a

cultural lens. Through the learning of a culture (that is, a sort of symbolic “language”)

one can both predict the behaviour of others and anticipate the predictions that others

make of one’s own behaviour (Mead (1934); Breakwell, 1983). The opposite is also

true: migrants and others, who are immersed in a foreign culture, are unable to gauge

others’ reactions, and hence may become unsure of who they really are.

Mead (1934) conceptualised his “looking glass self” as a consensus about the

self’s typical characteristics as reflected by the “generalised other” that represents an

amalgamation of others who view the self’s enduring ways. The Meadian “self” is

firmly situated within a nexus of others, without which there would be no notion of

9

“self”. Identities are given by the roles people occupy as defined by society—one’s

identity as “mother”, “student”, “migrant”, and so on. In Mead’s symbolic interactionist

view of the self, identities are not self-contained but are the products of social

interactions (Weinreich, 2003). Thus James’ (1890) argument that the individual has “as

many social selves as there are individuals who recognise him” (p. 178) is further

elaborated in Mead’s view that multiple roles are a normal consequence of life in a

complex social world.

Mead’s focus was primarily on the Me, or self as object. With his “generalised

other”, Mead claimed to have solved the problem of how one can be an object to oneself

(Burns, 1979). He differentiated between the self as known (the Me) and the self as

knower (the I) by the context in which each operates. The Me operates in the context of

group membership, status, roles, and interactions with others; the I operates in the

context of asserting individuality:

The simplest way of handling the problem would be in terms of memory. I talk to

myself, and I remember what I said and perhaps the emotional content that went

with it. The “I” of this moment is present in the “Me” of the next moment. There

again I cannot turn around quick enough to catch myself. (Strauss, 1956, p.242).

So as Mead’s I moves along, it leaves Me in its wake and provides for the self a

way of thinking about the self—self-knowledge. Burns (1979) noted that Mead’s “I-

Me” dichotomy differs in a major way from James’ initial formulation: Mead’s I is the

impulsive tendency, the undisciplined, disorganised potential of the individual (almost

parallel to the Freudian id). Every behaviour commences as an I but evolves into a Me

as it comes under the influence of society’s constraints. In Mead’s analysis, the Me is

essentially a social construction, eminently compatible with James’ social self.

While James, Cooley, and Mead often refer to the self in the singular, its lived

presence in the world of everyday life needs to be plural (Holstein & Gubrium, 2000).

10

James called attention to the socially manifold self when he stated: “Properly speaking,

a man has as many social selves as there are individuals who recognise him and carry

an image of him in their mind” (emphasis in the original, p.189-190).

Mead (1934) saw the self as developing out of social interactions and, as in

James’ theory of multiple social selves, allows for the existence of several selves within

one person: “… we divide ourselves up in all sorts of different selves with reference to

our acquaintances. We discuss politics with one and religion with another. There are all

sorts of different selves answering to all sorts of different social situations” (p. 140).

As Holstein and Gubrium (2000) asserted, James’, Cooley’s, and Mead’s selves

were plural because their attention was focused on the varied “we’s”, “us’s”, and

“them’s” that people are to one another as they go about their daily lives. The social self

was always the “self-at-hand, the socially operative sense of who we are to ourselves

and to others” (p. 20). Both Cooley and Mead stress the role that significant others play

in shaping the self-concept through social interactions. Here the individual considers not

only feedback from others, but the process through which more generalised attitudes

towards the self are adopted. Cooley moreover hints at a developmental process of

internalisation whereby the reflected appraisals of others are incorporated in the form of

enduring attitudes about the self—a process with implications for the stability of the

self-concept. Given that the self emerges in the context of a complex, organised,

differentiated society, it too must be complex, organised, and differentiated, reflecting

Stryker’s (1980) view that “self reflects society”, and James’ (1890) notion that an

individual has as many selves as social roles . As Stryker (1980) noted, a person has “an

internalised positional designation” (p.60), for each role the person holds in society.

To summarise, this section has reviewed some important antecedents of many

current theories of the self that hold that individuals’ self-concepts derive meaning

11

through relationships with other people and groups. The next section examines

contemporary issues that have emerged from these historical roots and are relevant to

the present thesis’ emphasis on understanding the philosophical and theoretical

underpinnings of dialogical self theory.

2.3 Identity

Personal versus social identity: one self or many? The core tenet of the theory

of the dialogical self is that individuals have within them not one self but many

(Hermans & Kempen, 1993). It is therefore appropriate to investigate where this notion

arose, and how well-founded it is. This section outlines prominent psychological and

sociological theories that embody this idea.

It was James (1890) who first drew the distinction between the spiritual self

(being one’s inner or subjective being and dispositions) and the social self (being the

recognition one receives from others). Since then, psychologists and sociologists have

long debated whether the personal (internal) or social (external) aspects of identity are

more important in understanding human behaviour. Tappan (1998) pointed out that

much of the theoretical and empirical work on the self has been conducted from two

perspectives: psychological and sociological. The multiplicity of identity, or senses of

self, was proposed early in the twentieth century. As noted earlier, James (1890)

observed that there may be incompatibility between certain role-related selves leading

to the “conflict of the different Me’s”.

By the mid-twentieth century, however, the consensus of opinion had reverted to a

unitary conception of the self. In the explicitly psychological perspective (Damon &

Hart, 1988), theorists such as Jung (1957) and Maslow (1961) asserted that a person is

most truly his or herself and closest to his or her authentic identity when experiencing a

deep sense of personal uniqueness. Allport, Rogers, and Maslow, for example, spoke of

12

the self in the singular; they accepted that a person might exhibit subsidiary social

selves or roles, but saw them as less real, central, or authentic (Martindale, 1980). And

although Jung (1956) and Assagioli (1976) accepted that people were indeed possessed

of multiple subselves or subpersonalities, they argued that these are (or could be)

integrated in a deeper-level unitary self. In such a paradigm, identity was simply a

personal characteristic, and its development a result of internal cognitive processes.

Debate over the unity of the self is not new. As Martindale (1980) observed, it has

concerned scholars for centuries. Does a person have a single dominant and enduring

self, or a multiplicity of subselves that compete for control? In tracing the dominant

viewpoints about the self through time, he noted that in the early nineteenth century the

theoretical consensus, exemplified by theorists such as Kant and Bain, favoured a

unitary self or transcendental ego. But by the end of the century opinion had swung the

other way: psychiatrists and psychoanalysts explicitly denied the existence of a

superordinate self and instead postulated the existence of multiple selves. The

phenomenon of multiple personality was adduced as evidence; supposedly, in the

normal personality these different selves were elicited appropriately as circumstances

required, but in the abnormal personality one or more attained an unwonted autonomy

(Martindale, 1980). Other social psychologists and sociologists defined identity almost

exclusively in social terms: “Identity is socially bestowed, socially sustained, and

socially transformed” (Berger, 1963, p.98). They focused mainly on the social

component of identity because “the shaping of one’s identity is dependent on the

valuations (sanctions, reinforcements) placed on one’s public conduct by relevant

others” (Sarbin & Allen, 1968, p.550).

Sociological role theorists such as Goffman (1959) took the distinction between

components of identity to its logical extreme. They maintained that people adopt short-

13

term selves, donning them as roles and shedding them as easily. People are actors in a

drama where, for a time, the role dictates the appropriate actions. Goffman did not

explain who the actors are behind the roles they play, nor who they are when no role is

available, for according to Goffman they always have roles to perform. As Breakwell

(1983) noted, one could conclude from Goffman that the individual is nothing but an

agglomeration of roles, and that the self is always socially constructed. That individuals

present differently in different social settings casts doubt on the stability of a self that

transcends social roles. Gergen (1968) argued as much: “the popular notion of the self-

concept as unified, consistent, or perceptually whole psychological structure is possibly

ill-conceived” (p. 306). In his experiments he showed that individuals do indeed modify

their self-presentation, and also the way they see and experience themselves, with

different people in different situations; and he concluded: “The individual has many

potential selves. He carries with him the capacity to define himself as warm or cold,

dominant or submissive, sexy or plain. The social conditions around him help determine

which of these options are evoked.” (Gergen,1972, p.66). By maintaining that the self

consists solely of social roles, theorists such as Gergen (1982, also Shotter & Gergen,

1989), deny the existence of a separate self; they see the self and identity as exclusively

social and lacking personal agency or authenticity. Other theorists find this position

extreme, however, and adhere to James’ position that a person’s social identities share

an enduring sense of self (Brown, 1998).

Further evidence for the multiplicity of the self was provided by the Twenty

Statements Test (TST) developed by Kuhn and McPartland (1954). Linking identity to

the question “Who am I?”, they found that responses are typically couched in terms of

social positions or social roles that respondents perform (or aspire to), that is, they name

their “masks” or personas (defined as that aspect of one’s personality that one displays

14

to others). Since a person occupies multiple social positions and performs multiple

social roles, it follows that he or she would have multiple social identities. Indeed, Kuhn

and McPartland further point out that when asked to give twenty separate answers to the

question “Who am I?”, nearly all participants named masks or personas, and many

named up to twenty. Gecas (1973, 1982; Gecas & Schwalbe, 1983) found that research

using the TST has focused on socialisation into the identity (e.g., gender-roles),

evaluations of the identity (e.g., racial and ethnic), and role transitions (e.g., adjustment

to widowhood, identity loss in the family).

Social roles also feature in research by Rosenberg and Gara (1985), who argued

that personal identity consists of a person’s various social and “personal” roles—kinship

roles, occupational roles, religious affiliation, group membership, intellectual concerns,

and so on. Thus, a person could, for example, have an identity consisting of such roles

as mother, stockbroker, Catholic, university student, avid reader, and so on. Each role,

or “identity” as Rosenberg and Gara call them, is “an amalgam of features—personal

characteristics, feelings, values, intentions, and images—experienced by the individual”

(p. 90) that can be enacted in different situations. They conceptualised relations between

identities in various ways, including that of subset-superset and disjunction. The subset-

superset relation occurs when one identity includes all the features of another identity;

thus someone’s identity as a mother might include all the features of her identity as a

caregiver. The disjunction relation occurs when two or more identities have nothing in

common, as where the identity of a devout church-goer might share no features with the

identity of an aggressive stockbroker. Although their main concern is with the social

roles played by the individual, these roles are defined in terms of feelings, values, and

so on, experienced by the individual rather than in terms of social interaction. This

resembles James’ (1890) view that the social self is in the mind, the major difference

15

being that for Rosenberg and Gara the social self is in the mind of the person playing

the role, while James’ social self is in the minds of those who recognise the person.

In another insight, Markus and Kunda (1986) noted that “the self is not a unitary

structure or even a generalised average of images and cognitions. Instead, the self-

concept encompasses … a wide variety of self-conceptions—the good selves, the bad

selves, the hoped-for selves, the feared selves, the not-me selves, the ideal selves, the

possible selves, the ought selves” (p. 858). They reason that at any given moment

people function according to a working self-concept: the most currently available view

of ourselves.

In his more recent discussion of the postmodern era, Gergen (1991) observed that

in the modern world, people confront “a swirling sea of multiple social relationships”.

They adapt by constructing multiple selves, but sometimes at a cost to themselves, for

their different selves may not necessarily harmonise. Their enduring, core self is

compromised by playing the “social chameleon”. The postmodern identity, dubbed

“multiphrenia” by Gergen (1991), is “the splitting of the individual into a multiplicity of

self-investments” (p. 73–74). He claimed that “In place of the enduring and identifiable

self, we find fragmentation and incoherence… the postmodern sensibility questions the

concept of ‘true’ or ‘basic’ self, and the concomitant need for personal coherence or

consistency” (p. 172–178).

Lifton’s (1993) postmodern “protean self” (named after Proteus, the shape-

shifting Greek god), develops the same theme. The protean self emerges out of

“confusion, from the widespread feeling that we are losing our psychological moorings”

(p. 1); “the older version of personal identity, at least insofar as it suggests inner

stability and sameness, was derived from a version of tradition culture in which

16

relationships to symbols and institutions are still relatively intact—hardly the case in the

last years of the twentieth century” (p. 4).

In summary, this section outlines prominent theories that hold that within every

individual there is not a single self but many. This notion is a central to the concept of

the dialogical self, which holds that not only are individuals composed of multiple

selves, but those selves are in constant dialogue (Hermans & Kempen, 1993). The

attraction and strength of dialogical self theory is that it integrates these multiple strands

of thought into a single and unifying theoretical framework that is amenable to

empirical testing, conceptual refinement, and widespread application. It is furthermore

for this reason that dialogical self theory underpins the exploratory tools used in this

thesis to explore the acculturative aspects of migration.

Identity as memory. The previous section discussed the multiplicity of the self;

this section broaches another key strand—identity as continuity of memory. Both are

core aspects of the theory of the dialogical self, since the dialogical self is accessed

through narratives, which themselves are founded in memory. As stated before, the

singularity and unity of personal identity has long held a central position in Western

philosophical and psychological thought (Tafarodi, Lo, Yamaguchi, Lee, & Katsura,

2004). James (1890) challenged this orthodoxy by arguing for a plurality of selves:

“Each of us, when he awakens says, here’s the same old self again, just as he says,

here’s the same old bed, the same old room, the same old world” (p. 334).

While Tafarodi et al. (2004) agree that James’ “same old self” is the familiarity of

some “formless, elusive I”, they recognise that it is also much more. It lies in the

perceived continuity of the traits, intentions, beliefs, attitudes and emotions that people

use to define themselves in society (Tafarodi et al., 2004). So, although people conduct

17

themselves differently according to the social demands of the circumstances, they do

not lose their essential sense of self.

People are not static. Just as James’ material Me grows old, so his social Me

evolves: roles are added and moulded, and old ones discarded (James, 1890). One can

assume that James’ spiritual Me changes as well. Over a lifetime, hardly a single aspect

of the person remains the same. Nevertheless, most people have an unmistakable sense

of continuity. James (1890) was one of the first to consider the sense of personal

continuity and the sense of distinctness from others as constructs of identity. For James

(1890), the sense of personal identity was most closely associated with the I or the “self-

as-knower”. A stable self-identity derives from a sense of continuity of the self-as-

knower. According to James, “each of us spontaneously considers that by the “I” he

means something always the same” (1961, p.63). James believes that “the worst

alterations of the self” (1948, p.207) are associated with disruptions of the sense of

personal continuity. For James, this feeling of individuality derives from the subjective

nature of the self-as-knower: “Other men’s experiences, no matter how much I may

know about them, never bear this vivid, this peculiar brand” (1961, p.71). It is this

special feeling associated with one’s own experience which is the basis for one’s sense

of personal uniqueness. But according to James the identity of self-consciousness across

time is “only a loosely construed thing”: James gives the example of adults failing to

recognise photos of themselves as children. In this case, the disjunction is not only due

to a lack of memory for “what it was like to be” (Nagel, cited in Tafarodi et al., 2004)

but also to dissimilarity in physical appearance. People furthermore may change in other

ways and so revise aspects of their inner selves, and accordingly revise specific beliefs

about who and what they are (Bem, 1972).

18

Giles (1997) commented that the way in which James (1890) developed his

argument about personal identity as memory was by noting that everyone is familiar

with an aspect of existence—a central nucleus or core—that seems peculiarly one’s

own. The I is the stream of consciousness that creates one’s knowing of the Me, which

is the aggregate of things known about the self. James’ I is fundamentally a stream of

thoughts in which each thought can “remember” the one before, and it is this that lends

a sense of continuity across time. James’ solution to understanding personal identity is

that personal identity is to be found in treating the spiritual self in a concrete way—that

is, treating it as a seamless flow of uninterrupted perceptions. Each perception flows by

and they are joined because they are part of the same stream of consciousness. It is this

consciousness that provides a feeling of continuity of the self. Even if there are gaps in

this continuity (due to sleep or otherwise) there is a sense in which consciousness before

the gap “goes together with” consciousness after the gap. For James, consciousness

coheres over time, and mainly through the effects of memory (Apter, 1983).

Moreover, according to James (1890), it is the feeling associated with each

perception that provides the tie that binds them. Each perception carries a distinctive

feeling that people recognise as theirs and theirs alone:

Each thought, out of a multitude of other thoughts of which it may think, is able to

distinguish those which belong to its own [self] from those which do not. The

former have a warmth and intimacy about them of which the later are completely

devoid … (p. 330).

So, for James, personal identity is one’s uninterrupted memory for prior

perceptions and the affect that accompanied them. Giles (1997):

A uniform feeling of “warmth”…pervades [our various selves] and this is what

gives them a generic unity, and makes them the same kind…where the

19

resemblance and continuity are no longer felt, the sense of personal identity goes

too. (p. 335).

For James (1890), it is not only the continuity of consciousness but also the

continuity of warmth and intimacy of a person’s memories that underlies personal

identity. In this way, personal identity requires more than memory; it requires an ability

to recapture the feelings associated with the experience:

If a man wakes up some fine day unable to recall any of his past experiences, so

that he has to learn his biography afresh, or if he only recalls the facts of it in a

cold abstract way … he feels, and he says that he is a changed person. (p. 336)

According to James, then, one’s sense of personal identity is underlain by one’s

autobiographical memories and what it felt like to experience them. The notion of

identity as a felt sense of similarity and continuity became better known through the

work of Erikson (1950, 1956, 1959), who sought to establish a bridge between personal

and social phenomena:

The term identity expresses a mutual relation in that it connotes both a persistent

sameness within oneself (self-sameness) and a persistent sharing of some kind of

essential character with others. At one time it [identity] will appear to refer to a

conscious sense of individual identity; at another, to an unconscious striving for a

continuity of personal character; at a third, as a criterion of the silent doings of

ego-synthesis; and finally, as a maintenance of inner solidarity with a group’s

ideals and identity. (p. 102)

In this way he characterised identity by two essential experiences: an enduring

and unique personal existence over time, and the sharing of an essential unity at some

level with others. In his definition, Erikson follows the general definition—of

continuously being one and the same person—but adds a social component, namely the

recognition of this sameness and continuity by others and the wider sociocultural

context the person lives in (Erikson, 1950). His conception of identity invokes the

20

interplay of individual and context: a person can feel embedded in his or her context. In

contrast, Erikson (1964) has discussed the feelings of “uprootedness” of migrants. They

are missing that sense of embeddedness.

In conclusion, the theory of the dialogical self, which is elaborated later, inherits a

long philosophical and psychological tradition that holds that identity is founded in

memory—a crucial aspect of dialogical self theory and the interview technique that is

employed to access it.

Identity as narratives. The previous section argued that identity is founded in

memory. This section explores how memories are accessed and marshalled to create

identity, and how both memory and identity are in a sense social; for memory comes

alive in the act of narration, and identity is realised when the self is presented to another.

Human identities are emergent constructs based on social interactions over a

lifetime. Narratives on the other hand are specific to time and place, use a particular

language, and refer to a particular stock of historical conventions, beliefs and values

(Crites, 1971). Narrative theories have in the past tended to focus on textual analysis

(such as rhetorical devices and form). More recently, theories have focused on the

meanings, relationships and socio-cultural context (such as motives and plots) (Gergen

& Gergen, 1986; Hermans & Hermans-Jansen, 1995; McAdams, 1985, 2014;

Polkinghorne, 1988, Hoshmand, 2005). For example, Mancuso and Sarbin (1983) focus

explicitly on the integration and coherence of self. The individual is the narrator of the

story, constructing plots that give meaning to important events. They distinguish

between the “I” as narrator, agent, organiser, and the “me” as actor, player, performer;

and “emplotment” is the process by which meaning is assigned so as to develop a

unified story:

21

The underlying assumption in employing the narrative as a root metaphor is:

Human beings think, perceive, imagine, and dream according to a narrative

structure … When one probes into the ordered or patterned perceptual response, it

becomes immediately apparent that a plot is imposed on the disparate inputs …

Only when the raw data are placed in the context of a plot structure does the

possibility exist for establishing a coherent plot. (p. 234-235).

The self-as-a-story-or-theory approach, also known as the narratology approach

(Shotter & Gergen, 1989, McAdams, 2014), has two central tenets. One is the appeal to

an informal story that is a theory of self; it is the story that the self uses to explain to the

self and to others why one is as one is, and why one acts as one does. For example, a

person’s much rehearsed story of disadvantage and abuse that explains a proneness to

being victimised. The second is the notion that certain people express a dominant view

of themselves and the world in terms of their story, to which all else is subordinated; for

example a story in which one is a victim of circumstances that are continually recalled

as being the reason for all subsequent experiences—a “poor me” dominant story.

In the “identity is a life story” approach (McAdams, 1998), “the problem of

identity is the problem of arriving at a life story that makes sense—provides unity and

purpose—within a socio-historical matrix that embodies a much larger story” (p. 18.).

Thus, the individual recurrently constructs and reconstructs the story that is ultimately

the answer to the existential question, “Who am I?” Furthermore, “the story provides a

coherent narrative framework within which the disparate events and the various roles of

a person’s life can be embedded and given meaning” (p. 19). McAdams’s narrative

approach to identity focuses explicitly on the integration and coherence of self, which

are key aspects of the sense of identity in Erikson’s terms. McAdams (1996) argued that

people organise their life stories in the context of a personal narrative. The individual is

the narrator of the story, constructing plots that give meaning to important events.

22

Adopting a symbolic interactionist perspective, Gergen and Gergen (1988) hold

that self-identity grows out of the narratives people tell to others and to themselves.

Stories, they contend, “serve as a critical means by which we make ourselves

intelligible within the social world … We use the story form to identify ourselves to

others and ourselves’ (p. 35). People’s experiences of self and others, the Gergens say,

is not based on “a series of discrete, endlessly juxtaposed moments, “but is instead

derived from “goal-directed, coherent’ narratives in which they attempt to “understand

life events as systematically related”. And moreover, they also emphasise that

“narratives of the self are not fundamentally possessions of the individual; rather they

are products of social interchange: “Narratives are essentially social constructions,

undergoing continuous alteration as interaction progresses” (pp. 20-21).

Gergen and Gergen (1988) proposed some of the rules of narrative form and the

types of narratives that are commonly employed, maintaining that cultural conventions

determine the validity of the story rather than the “absolute match between word and

thing”. Narrative structure often revolves around temporal forms or recurring sequences.

These rules of narrative structure lend coherence and directionality to a life. Plots are

made chronologically linear with respect of evaluative shifts over time; and within

them, sub-plots and overarching themes may involve stability, progressive change, or

regressive change (Gergen & Gergen, 1983). A paradox of human relationships,

however, is that “functioning viably in a relationship often depends on one’s ability to

show that one has always been the same, and will continue to be so, and yet,

contrapuntally to show how one is continuing to improve” (Gergen & Gergen, 1983,

p.266). Focusing on the protagonist’s developmental movement, Gergen and Gergen

(1988) differentiated between “progressive” narratives, where there is a positive change

in self-evaluation; “regressive” narratives, where there is a negative change in self-

23

evaluation: and “stable” narratives (elaborated later), where there is no change in self-

evaluation. In their schema, a life story establishes the main plot and, within it, which

incidents are construed as advancing (“progressing”), retreating (“regressing”) or

digressing. They proposed that progressive and regressive narratives be viewed as

moving through a two-dimensional space. One dimension refers to its evaluation,

positive or negative, the other its direction: “progressive”, “regressive” or “stable”. In a

progressive narrative, events are linked “in such a way that one steadily progresses

toward a goal” (p. 27): and in a regressive narrative one is “continuously moving away

from a valued state” (p. 27). Thus Gergen and Gergen’s (1986) framework for the

analysis of progressive and regressive narrative structures accounts for direction, where

direction has to do with moving towards or away from the goal.

Gergen (1991) noted that in different situations people view themselves

differently: that is, they make up a different story about themselves. The multiplicity of

the self here is understood to follow from the proposition that individuals harbour the

“capacity for a multiplicity of narratives forms” (Gergen & Gergen, 1988, p.35). They

emphasise the changes in self-identity as the person engages in different interactions in

different situations. The point is that in the postmodern world, our main task is just to be

different people in different situations.

Because narratives are communicative acts with a particular listener in mind, the

same story requires major adaptation to appeal to different audiences, who value

different things and judge narratives by different criteria. It is commonplace to tell

conflicting stories in different contexts, as the different voices in Hermans’ dialogical

self cannot all be satisfied with the same narrative (Hermans, 1996a). Thus, the demand

for narrative consistency, highlighted by McAdams et al. (1997), clashes with the need

to please different audiences with different accounts of the same actions. As has been

24

shown, some contemporary theorists challenge the view that the mature individual

necessarily strives towards a consistent identity. Instead, they argue that personal

sameness and continuity are a social construct, and since modern societies are

constantly in flux, personal sameness is no longer a sign of maturity (Hermans,

Kempen, & Van Loon, 1992; Gergen, 1991; Lifton, 1993; Sampson, 1995).

To summarise, the literature contrasts traditional and post-modern conceptions of

personal identity: the former holds that good mental health is founded on a largely

coherent and harmonious life story; the latter holds that lives—even healthy lives—are

messy and disorganised in narrative terms.

2.4 The Dialogical Self

Since the 1980s there has been a resurgence of interest in the self (Harter, 1988).

New models have proliferated, though researchers have sought guidance from past

scholars, typically paying intellectual homage to James (1890) and symbolic

interactionists such as Cooley (1903) and Mead (1934). Hermans’ (1992) theory of the

dialogical self is one of the more influential models that have appeared since that time.

As previously discussed, it draws on James’ (1890) distinction between I (the “self-as-

knower”) and Me (the “self-as-known”) as the two main components of the self. This

section describes the building blocks of dialogical self theory, and explains how they fit

together.

I-positions and voices. According to Hermans (1996a), dialogical self theory

… conceptualized the self in terms of a dynamic multiplicity of relatively

autonomous I positions in an imaginal landscape. In this conception, the I has the

possibility to move, as in a space, from one position to the other in accordance

with changes in situation and time. The I fluctuates among different and even

opposed positions and has the capacity to imaginatively endow each position with

a voice so that analogical relations between positions can be established. The

25

voices function like interacting characters in a story, involved in a process of

question and answer, agreement and disagreement. Each of them has a story to tell

about his or her own experiences from his or her own stance. As different voices,

these characters exchange information about their respective Mes, resulting in a

complex, narratively structured self (p.33).

This quote captures the two main ways in which Hermans’ view differs from that

of James. Where James saw the I “as a unifying principle that is responsible for

organizing the different aspects of the Me as parts of a continuous stream of

consciousness” (Hermans, 2001, p. 246), Hermans sees it as possessing “the possibility

to move from one spatial position to another in accordance with changes in situation

and time” (p. 248). The distinction is one of stasis versus flux: James’ I is continuous

and corporeally bound—both essentially static notions; Hermans’ I possesses

discontinuity as well as continuity, and combines temporal characteristics with the

spatial—both essentially dynamic.

Dialogical self theory furthermore posits that multiple I-positions can be occupied

by the same person in space and time (Hermans, 2001), giving rise to identities that are

fluid and multiple. At the same time it accepts a self that remains continuous through

space and time (Richardson, Rogers, & McCarroll, 1998). In this way it belongs to the

“multiple selves” school of thought in the sense that it echoes ideas current at the turn of

the century, such as those of James, to which Hermans and colleagues (Hermans,

Kempen & Van Loon, 1992; Hermans & Kempen, 1993) pay explicit homage; but it

also transcends them.

Derivatives of I-positions. The concept of the I-position has generated a number

of derivative concepts, of which three—the “meta-position”, the “coalition” of I-

positions, and the “third position”—are central to a dialogical understanding of

acculturation.

26

I-positions form a “society of mind” (Hermans (2004_, p. 2), and like members of

any society they can form coalitions and hierarchies. Primary I-positions can therefore

be grouped in the mind of the observer into “coalitions” of I-positions. Likewise, “meta-

positions” stand in a hierarchically superordinate relation to coalitions of primary I-

positions in that they allow individuals to observe their own I-positions “from the

outside, as an act of self-reflection” (Hermans and Hermans-Konopka, 2010; p. 9).

According to Hermans (2001)

… a meta-position has several specific features: it creates a certain distance

toward the other positions; it provides an overarching view; it enables the

participants to interrelate the positions as part of their personal history; it provides

an opportunity for evaluating the several positions and their organization; and,

finally, the direction of change and the importance of one or more positions in

view of future possibilities become apparent. In summary, the meta-position

contributes, more than most other positions, to the integration and unity of the

repertoire. (p. 354)

In this manner, acculturating individuals are able to group their I-positions into

two coalitions (among others), one comprising I-positions that relate to their country of

origin, the other to Australia; and to survey each coalition independently but as a

collectivity by means of a meta-position unique to each coalition. It is this integrative

function of meta-positions and coalitions of I-positions that render the concepts crucial

in understanding acculturation.

Lastly, “third positions” (Hermans and Hermans-Konopka, 2010) are formed from

the dialogical creative interaction of I-positions. Far from being a simple amalgamation

of their formative primary I-positions, third positions may be completely or partly

novel, whereby ‘relationships of tension are transformed into conciliation and creation’

(Surgan & Abbey, 2011; p. 165). Surgan & Abbey and Valsiner (2002) propose various

27

mechanisms by which this takes place in the case of transcultural migrants, many of

which fall under the broad category of cultural “hybridisation” (see Chapter 4).

Emotional voices. “Dialogical voices can be reasoned or emotional” (Hermans,

2010; p. 41). It is natural to suppose that the “reasoned” content of dialogical voices

should matter, as reason facilitates rational dialogue. But not all dialogue is rational in

the narrow sense. By colouring the dialogical relationship, emotion constrains the range

of possible dialogical outcomes in terms of ethnic identification: majority culture,

minority, or hybrid. Dialogue that is coloured by antagonism is unlikely to lead to

resolution; conversely, shared emotional ground provides a basis for negotiating an

outcome that benefits both parties; and both are reflected in mental wellbeing. In the

cultural context, this may make all the difference between adaptive acculturation and

maladaptive.

What actually determines this emotional content is, however, another matter.

Hermans and Hermans-Konopka (2010) and Kinnvall (2004), point to power disparities

between minority and majority groups. When confronted by racism, whether outright or

in its more subtle manifestations, migrants may seek to counter its impact though

strategies that ascribe comforting emotional content to acts and objects—strategies such

as localisation and homesteading (see Chapter 4). Mahmoud (2009), Malhi (2009) and

Verkuyten (2002) see emotion (or “feeling”, as they put it) as a mediating factor in

determining ethnic identity. The relationships they describe are complex, but in essence

ethnic identity is partly a personal choice, but one that is governed by social context and

constrained by social norms. In dialogical terms all these models require that dialogical

voices be freighted with emotion.

Dialogue. Hermans (1996a) describes dialogue as implying “an interchange

between mutually influencing voices” (p. 31). The concept draws on the Russian

28

literary critic Bakhtin’s (1973) concept of the “polyphonic novel”. Hermans, Rijks and

Kempen (1993) conceive of the polyphonic novel as one “where different voices, often

of a markedly different character and representing a multiplicity of relatively

independent worlds, interact to create a self narrative” (p. 208). In the polyphonic novel

the author adopts a number of viewpoints corresponding to different imaginary

characters, who are not necessarily consistent in their attitudes, knowledge or beliefs

(Hermans, Kempen & Van Loon, 1992). According to Hermans and his colleagues, the

self is best regarded as a composite of characters that exhibits the variety one might find

in such a novel. Some—the I-positions—correspond to different authorial viewpoints;

others—the Me-positions—are the characters. Like the characters in such a novel, they

engage in dialogue, hence the term “dialogical self”. This “multivoiced” self contrasts

sharply with the conventional view of the self as organised around a single centre or

core. As author of its own life story, the I moves from one I-position to another,

adopting different perspectives. The self emerges as the totality of such I-positions,

together with their inter-relationships, that is, dialogues.

Self-narratives. Hermans, Rijks and Kempen (1993) hold that the metaphor of

the polyphonic novel extends the original narrative conception of the I as author and the

Me as actor (Sarbin, 1986). But unlike Sarbin, who assumes that a single narrator tells a

story about himself or herself as actor, Hermans, Rijks and Kempen consider that

narrators adopt numerous vantage points. According to Hermans and colleagues (1993;

Hermans, 1996a; 2001; 2002) the conception of the self as a polyphonic novel goes

even further in that it assumes that:

The one and the same individual lives in a multiplicity of worlds with each world

having its own author who may tell a story relatively independent of the authors

of the other worlds. It is thus assumed that the individual consists of multiple

29

authors entering into dialogical relationships with each other and creating a more

complex organization of the self (p. 213).

In their view each author relates a story that is entirely independent of that told by

another. Authors relate to one another, but they do so dialogically, with none having

ultimate power, though each may dominate at different times and in different places. It

is in the spatial dimension that Hermans’ dialogical model departs from those of other

advocates of a narrative approach, such as Sarbin (1986), Bruner (1986), Gergen and

Gergen (1988), and McAdams (1993). Following Bakhtin, time and space are seen as

equally important. This is suggested by the words “position” and “positioning”, which

Hermans employs in place of the traditional term “role” (Hermans, 2001); and also by

Bakhtin’s (1929/1973) usage of the term “juxtaposition”, which presupposes a plurality

of voices that are heterogeneous and even opposed, much less identical or unified.

Valuations. In the theory of the dialogical self, a “valuation” is “anything that a

person finds to be of importance when thinking about his or her life situation”

(Hermans, 1988, p.792). Valuations include loved ones, those one dislikes, disturbing

dreams, problems, cherished opportunities, memories of important events, plans or

goals, and so on. Each valuation is a “unit of meaning”, and each has either a positive,

negative or mixed (ambivalent) emotional quality. Through self-reflection, people

organise their valuations into self-narratives that situate them in time and space—a

valuation system. This means that the “person lives in the present and is, from a specific

point in space and time, oriented toward the past as well as the future” (Hermans &

Kempen, 1993, p.81). By reflecting on past and future experiences, the individual

composes a personal story consisting of significant events by. Moreover valuations may

differ through the life-cycle because the individual’s reference point changes (Hermans,

2006; 1991).

30

The latent–manifest distinction. Hermans (1993) introduced a latent-manifest

distinction into valuation theory. It assumes

that a small set of basic motives are latent in the affective component of the

valuation system. These motives, moreover, are… assumed to be similar across

individuals and to be continuously active within each individual. But, at the

manifest level, valuations vary phenomenologically not only across individuals

but also within a single individual across time and space. (p. 442)

Hermans (1989) notes that the reflective individual, as an observer of his or her

own life, “can look back with nostalgia on the past, be worried about the present, and

have hope for the future” (p. 14), but is always limited in his or her perspective by the

circumstances of the present. Since Hermans conceives personality as a process in time

and space, the perspective of the “I” is not fixed but changes from moment to moment:

The question is now how this concept of change can be related to the manifest and

latent levels of life meaning representation. It is expected that when the

perspective of the “I” changes, there may be a concomitant change in the manifest

experience of an event. Whether or not this change also implies basic

reorganization at the latent level remains to be seen. (p. 14 )

According to Hermans, valuations can be interpreted in terms of two primary

motives: the self-enhancement motive and the union-with-the-other motive. Following

Bakan’s (1966) concepts of agency and communion, the self-enhancement motive is

self-oriented—a striving for superiority, expansion, power and control. The union-with-

the-other motive is other-oriented—a longing for contact, union and intimacy. The two

motives give direction and organisation to the valuations in a person’s life story.

Sampson (1988) echoed the same idea when referring to the work of Bakan (1966). In

every individual there coexist two mutually opposing motives. “Agency”, manifests as

the urge for self-assertion, self-protection, and self-aggrandizement; “communion”,

manifests as the urge to seek union with something larger than the individual. These two

31

motives are paralleled in other research. McAdams (1985) distinguished power and

intimacy as basic motives. Many researchers in cross-cultural psychology (e.g.,

Triandis, Bontempo, Villareal, Asai, & Luca, 1988; Schwartz, 1990) make use of a

notion that is variously labelled “individualism–collectivism” and “idiocentrism–

allocentrism”, which in all its guises closely resembles Herman’s “self-enhancement”

and “union-with-the-other”.

Another way of describing valuations is by their affectivity, positive and negative.

Each valuation reflects positive and negative feelings, which indicate the extent to

which basic motives referred to above are fulfilled. Different I-positions may be

compared and contrasted, then, with respect to the extent to which their valuation

systems exhibit different degrees of self-enhancement motive, union-with-the-other

motive, positive affect, and negative affect. The following example clarifies the

relationship between the motivational and affective components of a valuation. The

valuation, “I gained my doctorate by hard work”, evinces pride and strength; the self-

enhancement motive is manifest in these affects as part of the valuation. Similarly, the

evaluation, “I enjoyed the time when I was together with my family,” evinces

tenderness and intimacy.

Identity. That personal identity is essentially dialogical is a natural extension of

the notion of a narrative self, and follows from the proposition that self-narratives are

collaborative, involving one or more listeners (Sarbin, 1993). In the contextualist view

of human conduct (Sarbin, 1976, 1986), human identities are evolving constructions;

they emerge out of continual social interactions in the course of life. This is a theme that

permeates the novels of Amin Maalouf, a Lebanese migrant to France: “Identity isn’t

given once and for all: it is built up and changes throughout a person’s lifetime.”

32

A person’s social identity may designate the collectivity with which a person

identifies (e.g., an ideology or religion) or with a location on the social map (Berger,

1966). Greenwald (1988) used the term collective self to refer to the individual’s

identification with a reference group, which leads to adoption of the group’s values and

goals. He pointed out that a range of groups can provide the basis of the collective self

(e.g., religious, racial, ethnic, national, political, corporate, and professional groups).

Although dialogical self theory posits that the dialogical self is “social” (Hermans,

2001a), this sociality is “not in the sense that a self-contained individual enters into

social interactions with other outside people, but in the sense that other people occupy

positions in the multivoiced self” (p. 250). Commenting on the relationship between self

and collectivity (e.g., Miller & Prentice, 1994; Turner, Hogg, Oakes, Reicher, &

Wetherell, 1987), Hermans (2001a) notes that the

idea of the Cartesian individualist and centralized self, which is separated from

group and culture, can be contrasted with his dialogical approach that opens the

realm of collective voices. (p. 262)

For Hermans, Baktin’s multivoicedness entails not only “the simultaneous

existence of different individual voices in the self, but the simultaneous existence of an

individual voice and voice of a group” (p. 262). In answer to his question regarding how

are these individual and collective voices related to one another, Hermans comments:

Empirical evidence suggests that the two voices function as relatively autonomous

parts of the self …The relative autonomy of the personal and the collective parts

of the self requires us to study their dialogical relations. As individuals people

may agree or disagree with the collectivities to which they belong. (p. 262)

Individuals engage in dialogue at a collective level as well as an individual one

(Hermans, 2002a, 2002b, 2003). The individual and collective voices within the

dialogical self parallel the personal and social positions in which people may find

33

themselves. Social positions are defined by the expectations and prescriptions of

society, personal positions by how individuals choose to organize their lives. Hermans

(2001a) illustrates this point with the following example:

when a person in a particular culture is defined as a woman, this social position

carries specific expectations regarding the person’s dress, movements, behaviour

and emotion regulation. However, from the perspective of her personal point of

view, she may feel feminine in some situations (e.g. in dress), but masculine in

other situations (e.g. in sexual behaviour). In this case, the meaning system of this

person is constructed in a field of tension between her social position and one or

more of her personal positions. (p. 263)

Thus, in Hermans, dialogical voices are not always harmonious but are engaged in

“negotiation, disagreement, power, negation, conflict, domination, privileging and

hierarchy” (Bhatia & Ram, 2001, p.301). For Hermans (2001a) these individual and

collective voices coexist and are in constant tension between “(symmetrical)

interchange and (asymmetrical) social domination” (p. 266). In this way the theory

accommodates the ways in which power asymmetries emerge between cultures and

social groups.

In this sense, culture operates as a collective voice that organises the meaning of

the dialogical relations (Hermans, 2001a). Because the self and culture are conceived in

the theory in terms of a “multiplicity of positions among which dialogical relationships

can develop” (p. 243), it then follows that cultural changes bring about changes in the

dialogical self by creating a movement of I positions as well as a change in how these

positions relate mutually. The culturally constituted dialogical self can be observed and

understood from personal narrative accounts (Hermans, 2004; McAdams, 1993;

McAdams & Pals, 2006). Since a personal narrative embodies memories, feelings, and

experiences, narrators generally tell their stories in highly selected ways that depend on

34

changing situations and social expectations (Hermans, 2000). This means that the

created world of the self-narrative, where the recounted actions and the dramatised

voices take place, is always context-dependent. Different narrative identities are

constructed according to linguistic, social and physical context. Speakers make narrative

choices in order to display a particular portrait of themselves, constructed for a

particular context and a particular audience. They may also tell different and sometimes

contrasting stories from different positions at different times (Hermans 2001a).

In short, dialogical self theory provides a conceptual toolkit with which to explore

the psychological aspects of culture and acculturation. Applying it, however, demands

elaborations and extensions of the theory, which are described in Chapter 4.

2.5 Summary

This chapter discusses the concept of the self, showing how and why it relates to

the thesis. A pervading theme is the notion of multiple selves. This idea is not new, but

has now become mainstream, beginning with James’ theories, and progressing through

Cooley, Mead, and Hermans. Their theories differ, sometimes substantially, in detail

and approach, but all have in common the notion that people possess multiple selves,

and that different selves are manifest in different circumstances.

Since Hermans developed his theory of the dialogical self, researchers have

refined and extended his theory into practical areas of application, important among

them being migration and acculturation. Their work is also described, but because it

relates closely to theories of migration and acculturation, it is presented after the next

chapter on migration and acculturation theories.

The idea that people have multiple selves that coexist in time and space is central

to this thesis. It is furthermore necessary, for the kind of Cartesian individualist and

centralized self, separated from group and culture, is incapable of explaining the

35

changes that often (though not always) occur when migrants migrate. For that one needs

a multivoiced approach, and furthermore one that accepts that the self is moulded by the

external environment, in this case culture. The empirical component of the thesis

therefore focuses on participants’ multiple selves, namely those corresponding to the

participant as a member of the culture of their country of origin, and those

corresponding to the participant as a member of the culture of their country of

settlement, in this case Australia.

36

CHAPTER 3: Migration and Acculturation

Ce qui fait que je suis moi-même et pas un autre, c’est que je suis ainsi à la lisière

de deux pays … C’est précisément cela qui définit mon identité. Serais-je plus

authentique si je m’amputais d’une partie de moi-même? [What makes me myself

rather than anyone else is the very fact that I am poised between two countries …

It is precisely this that defines my identity. Would I exist more authentically if I

cut off a part of myself?]

Amin Maalouf (1949– )

In my case, the effort for these years to live in the dress of Arabs, and to imitate

their mental foundation, quitted me of my English self, and let me look at the

West and its conventions with new eyes: they destroyed it all for me. At the same

time I could not sincerely take on the Arab skin: it was an affectation only ….

Sometimes these selves would converse in the void; and then madness was very

near, as I believe it would be near the man who could see things through the veils

at once of two customs, two educations, two environments.

T. E. Lawrence, Seven Pillars of Wisdom (1922)

This chapter describes the migration experience from a psychological and

sociological perspective, with particular focus on the Australian experience. It explores

the process of migration and acculturation in order to understand the various outcomes

of intercultural contact, and concludes with a synopsis of acculturation theories and

their implications for the thesis.

3.1 Australian Migration

The global context. It is commonplace to say we now live in a ‘globalised’

world. One measure of this is language. Language is not the same thing as culture, but

being relatively easy to observe, it is a serviceable indicator of it. For instance, the

European Commission (2012) reported that just over half of Europeans were ‘able to

37

hold a conversation in at least one additional language…’ (p. 5). By that criterion, most

people in Europe at least have to engage at some level with a culture not their own.

Global figures are harder to come by, but the British Council (2014) estimated that ‘one

out of four of the world's population speak English to some level of competence;

demand from the other three-quarters is increasing.’

If second language acquisition is a sign of potential acculturation, migration is

even more so. The OECD (2013) recently reported:

Some 232 million international migrants are living in the world today. Since 1990,

the number of international migrants in the global North increased by around 53

million (65%), while the migrant population in the global South grew by around

24 million (34%). Today, about six out of every ten international migrants reside

in the developed regions.

In short, migration is large and growing, particularly from the ‘South’ (which the

OECD broadly defines as developing economies), to the ‘North’ (developed

economies). Given the differences between their cultures, it is clear that acculturation

issues are greater than they have ever been, and are moreover set to increase.

The historical perspective. Australian immigration has been characterised by

several phases. Since European settlement began over 200 years ago until the end of

World War 2, Australia was essentially a “British” country—British in population in

that its people hailed overwhelmingly from the British Isles; and British in customs,

more so than the British in some ways, since Australians saw themselves as “peoples of

the three kingdoms [England, Scotland and Ireland]… equal sharers in this new

community” Hirst (2014; p. 146). Even the Irish, whose country at that time was part of

the United Kingdom, were happy enough to regard themselves as “British” in Australia,

when they would have resolutely rejected the notion back in Ireland.

38

All that changed in the aftermath of World War 2, by which time Australia’s

proportion of immigrants had fallen from a high of about 30% in 1893, to a low of near

10% in the late 1940s (Australian Bureau of Statistics 2013, December). After the near

invasion of their country by Japan, Australians rallied to the slogan: ‘Populate or

perish!’ True, institutionalised racism (present since Federation) remained a formal part

of the new immigration policy, but given the shortfall of British migrants, Australians

were persuaded to accept—reluctantly at first—other northern Europeans, then

successively more southerners, and ultimately non-Europeans. And accept them they

did (Hirst, 2010, 2014; Markus et al., 2009). By the 1970s the proportion of foreign-

born neared 25%.

The White Australia Policy was finally abolished in the mid-1970s, leading to a

surge in non-European immigration, mostly from Asia and the Middle-East. By the

mid-1990s nearly 50% of new entrants came from Asia, although Britain and New

Zealand were still major source countries (Castles, Iredale & Vasta, 1994).

Current situation. Australia is exceptional in its migration history, which is

reflected in its current migrant composition. Other countries may have a higher

proportion of foreign-born residents, but those countries are not traditionally countries

of settlement in the way that Australia is: some are burdened by refugees, such as

Jordan; some have large guest worker communities, such as Saudi Arabia and

Switzerland; and many are small island countries that have little impact globally, such

as Bermuda (OECD & UNDESA, 2013). Only Australia has so many foreign-born,

increasing so rapidly. Even Israel, with which Australia is often compared in matters of

migration, has these days a lower proportion of foreign-born (and that against a

backdrop of Israel’s Law of Return, which accords any Jew the right to migrate to

Israel).

39

Table 3.1 compares Australia with comparable countries from a migration

perspective. Australia’s migrant stock, though proportionately lower than Israel’s in

1990, now exceeds it; and its rate of growth exceeds that of all countries but New

Zealand. Canada, with which Australia is also commonly compared, is culturally more

diverse (Hawkins, 1989) but has proportionately fewer migrants. Because of its size the

United States receives many more immigrants than Australia in absolute terms, but

proportionately its migrant stock is half that of Australia’s and in recent years has

grown half as fast.

Table 3.1 Migrant stock: selected countries of large-scale immigration

Migrant stock

1990 2013

Growth rate

2010-2013

Australia 22.7% 27.7% 2.5% pa

Israel 36.3% 26.5% 0.5% pa

Canada 16.3% 20.7% 1.3% pa

New Zealand 15.4% 25.1% 3.4% pa

United States 9.1% 14.3% 1.2% pa

Source: OECD & UNDESA (2013).

Demographic analysis based on the 2006 national census showed that out of a

population reaching 21 million, nearly a quarter of the population was born overseas,

and a third had at least one overseas-born parent. The UK and New Zealand still remain

the most common birth countries. The dominant culture remains European, with British,

Italian and German still the main non-Australian ancestries, but Australians are

becoming less European and more Asian. Chinese are now the fourth largest overseas-

born group, with Italian third. Overall, Australia has 250 ancestries speaking 400

40

languages at home, but 79 per cent only speak English at home. Italian, Greek and

Cantonese are the most common non-English tongues.

Policy settings. When mass immigration began after the World War II, the policy

objective was officially one of cultural assimilation: migrants would become Australian

in the quickest possible time, and to that end were termed ‘New Australians’ by

Government and agencies (Hirst, 2010, 2014; Markus et al., 2009). By the mid-1960s,

this had changed: cultural “integration” was officially the goal, by which was meant that

the authorities “still sought social cohesion but more gently” (Hirst, 2014; p. 160).And

by the mid-seventies the term “multiculturalism” was officially adopted in recognition

of a social environment became much more tolerant and accepting of difference, despite

the racist ideas of the past (Castles & Vasta, 2000; Castles, Iredale, & Vasta, 1994;

Ward, Bochner and Furnham, 2001). Although few countries are completely

monocultural, some are more culturally diverse than others, and Australia is often cited

as an example of a culturally diverse society, as are the United States and Canada

(Berry, 1997; Bochner, 1986; Bochner & Hesketh, 1994).

There is a pattern to this policy progression: the foundation for what Hirst (2014)

terms Australia’s migration “success” rested on Australians’ fears being allayed. At

first, non-British, European migrants were accepted, grudgingly, and then only because

they were needed: this was the era of ‘Populate or perish’. Even so, ‘old’ Australians

demanded that they adopt Australian ways. Later, when Australians realised the

newcomers caused no trouble (and indeed, brought interesting cuisines with them), they

were allowed to keep their cultural identity (Hirst, 2014; Markus et al., 2009). Finally,

those customs were accepted as part of the Australian social fabric. It took 30 years, but

the transformation was dramatic, and still continues.

41

Push and pull factors. In a review of post-war literature on Australia’s

immigration program, Betts (1996) observed that immigration was largely explained in

terms of pull factors, that is, ones intrinsic to the host country that draw migrants to it.

Such theories have flourished in Australia and overseas since the mid-1970s, despite the

fact that since 1974 changes in immigration were already beginning to make these

theories incomplete. Instead, she suggested push factors are now more important,

driving migration in the absence of the former official policy of active recruitment, and

often in the face of attempts at exclusion. She saw a need to further define the terms

“push” and “pull” factors more clearly—that is, not from the point of view of the host

society but of the potential immigrant. For example, in the “new economics” theory,

potential migrants weigh up the net costs and benefits of attempting to emigrate

(Massey et al., 1993). This means that they consider the pull of the attractions in the

new country together with the push of negative factors in their home country and

evaluate these against the expense and risk of moving.

Betts (1996) also noted that the distinction between push and pull is confusing

when perceived from the migrant’s point of view. If a person is contemplating a move

he or she must see some attraction (pull) in the target country as well as some reason to

depart (push) in the home country. It follows that a potential migrant’s actions are

neither governed by push or pull: if benefits are to be perceived, both push and pull are

required. The decision to move involves weighing up the benefits of both against the

cost of breaking old ties, making a journey, and so on.

Motivations. Young and Madden (1992) examined how people arrive at the

decision to migrate to Australia. In a study of over 500 migrants from the United

Kingdom, Ireland, Lebanon, Malaysia/Brunei, and Vietnam, they found that women

generally migrated for “family reasons”, while men more often migrated as skilled

42

workers. Men were more likely to be ‘primary’ migrants in all categories and women

were ‘secondary’ migrants.

Maydell-Stevens, Masgoret & Ward (2007) found that motives were a key

influence on migrants’ adaptation to a new culture, to the extent even of determining

their acculturation strategy. Boneva and Frieze (2001) found that achievement

motivation, power motivation, and affiliation motivation all influence the desire to

migrate: intending migrants tend to be above average in achievement and power

motivation and lower in affiliation motivation. Migration is an expression of an

individualistic orientation: those high in achievement motivation may migrate in order

to seek opportunities, and those high in power motivation may be more willing to take

risks in reaching their goals. The authors interpret migration as a life-changing event

that offers opportunity but also carries risk: by moving into a new and usually larger

society, migration offers career and economic enhancement but impedes social ties with

home.

Migrants also differ from other cross-cultural travellers in motivation and

commitment to country of resettlement. Unlike refugees, who are “pushed” into an alien

environment, migrants are “pulled” towards a new country in pursuit of personal,

familial, social, financial and political goals (Winter-Ebmer, 1994). Economic factors

exert the strongest influence, though family reunification and political pressures play a

part. In a psychological analysis of migration intention, Winchie and Carmen (1988)

found that migrants are more likely to be dissatisfied with their employment in their

home country and come from a lower socio-economic stratum than non-migrants.

A typology of migrants. Taylor (1969) identified three types of migrant.

“Resultant” migrants are pressured by circumstances to move; they move essential

because they have to. “Dislocated” migrants move to avoid “dislocation” from their

43

primary group, often to join family or friends who have already migrated. But the most

are “aspirers”, who migrate to better themselves and their children. The “dislocated”

and “aspirers” are responding mainly to “pull” factors; the “resultant”, to “push” factors.

Modern economic theories of migration place the cause for moving within the

immigrants’ perceptions of their situations rather than within the host countries’

economic and political institutions, which may explain the selectivity of migration

(Massey et al., 1993). These factors point to economic, demographic and network

factors in order to explain why people leave their countries of origin (Massey et al.,

1993). They do not, however, explain why some people become migrants and others in

similar socio-economic conditions do not.

It may be that some people are psychologically predisposed to migration, that is,

they may possess characteristics that predispose them to migrate and help them endure

the challenges of migrant life. Jennings (1970) coined the term “mobicentric man” for

individuals who value action and are constantly on the move. Morrison and Wheeler

(1976) coined the term “pioneering personality” and suggested that the desire for

novelty might be as important a motivator as the desire for material advancement.

Other studies support the notion of a “migrant personality” and specify its

characteristics. Boneva and Frieze (2001) note that not everyone in poor countries wants

to migrate to richer ones, and would-be migrants are not all from the poorest class. They

accept that economic prospects, migration policies, and network support in the receiving

country create the preconditions for the desire to migrate, but argue that at the level of

the individual, migration is largely driven by personality. They found that primary

migrants (that is, those making the initial decision) to be more work-oriented, and

motivated more by achievement and power than by affiliation or family centrality, than

44

secondary migrants (that is, those who follow in order to be reunited with family)

(Boneva & Frieze, 2001; Frieze et al., 2004).

3.2 Acculturation

Terms and concepts. Acculturation is “the process by which individuals change,

by being influenced by contact with another culture and by being participants in the

general acculturation changes under way in their own culture” (Berry, 1990, p.204). It is

important to differentiate acculturation of the group from acculturation of the individual.

Sociologists and anthropologists were the first to study acculturation, and were mainly

interested in group-level changes; Redfield, Linton, and Herskovits (1936) were among

the first to provide a formal definition:

Acculturation comprehends those phenomena which result when groups of

individuals having different cultures come into continuous first-hand contact, with

subsequent changes in the original culture patterns of either or both groups. (p.

149)

Later, Graves (1967) introduced the concept of “psychological acculturation” to

describe changes as a result of personal contact with the other culture, and participation

in acculturation at the group level (Berry, 1990), so recognising migrants’ emotional

attachment to the host culture and country of origin (Betancourt & Lopez, 1993).

According to Berry, the phenomena are different at the two levels: at the group level,

changes occur mainly in the group’s social structure, economic base, and political

organisation; at the individual level, changes occur in identity, values, attitudes, habits

and motives (Berry, Kim, Power, Young, & Bujaki, 1989; Berry, Poortinga, Segall, &

Dasen, 2002).

Acculturative changes are normally adaptive in that they help individual to

transact two different cultures. This is illustrated in Figure 3.1.

45

Figure 3.1 The acculturation process (adapted from Berry, 2005)

Group-level factors (at top) precede acculturation in the society of origin, and pre-

exist in the society of settlement. Individual-level factors (at bottom) precede

acculturation in the migrant. In the middle are factors that arise in individuals during the

course of acculturation. Both societies act jointly to initiate and shape psychological

acculturation, which is described in terms of stressors, responses, and outcomes (for a

review, see Berry & Kim, 1988; Berry, Kim, Minde & Mok, 1987; Berry, 2005).

Individuals differ in how they experience acculturation. Some perceive acculturative

changes as stressors that expose deficits in coping skills; others perceive them as benign

opportunities for growth. Acculturative stress is manifest in affective, behavioural or

Stress

and skill

deficits

Responses:

affective,

behavioural,

cognitive

Outcomes:

psychological

sociocultural

adaptation

Cross-cultural

transition

experience

Societal-level variables

Individual-level variables

Society of origin

Social factors

Political factors

Economic factors

Cultural factors

Society of settlement

Social factors

Political factors

Economic factors

Cultural factors

Person

Personality

Language fluency

Training and experience

Cultural identity

Acculturation strategies

Values

Reasons for migration

Situation

Length of cultural contact

Amount of intra- and inter-group contact

Quality of intra- and inter-group contact

Cultural distance

Amount of life changes

Social support

46

cognitive responses. Variations in adaptation (psychological or sociocultural) range

from the well-adapted to maladapted.

Intercultural contact. The social psychology of cross-cultural relations and the

psychological effects of culture contact are important areas of research in cross-cultural

psychology, (e.g., Ward, Bochner & Furnham, 2001; Bochner, 1982; Brislin, 1981).

Bochner (1982) found intercultural contact to be of two kinds: that between members of

a culturally diverse society, and that which takes place when a person from one culture

travels to another with a particular objective, for example, to work, play, study, exploit,

convert, or assist. Over the last 30 years empirical research on intercultural contact has

flourished, particularly as concerns sojourners, refugees, migrants and expatriate

workers. Migrant adaptation has been found to be influenced by pre-migration

conditions, by the transitional experience in moving from one country to another, by the

characteristics of the migrants themselves, and by the conditions of the receiving

country, including government policies and economic factors. Migrants generally

relocate with the intention of long-term, if not permanent, resettlement (Ward, Bochner

& Furnham, 2001).

Berry (1990) observed that, technically, overseas-born, first generation settlers

should be described as migrants, and their descendants as members of ethnocultural

groups. Ward and colleagues (2001) pointed out that it is the initial contact between the

first generation migrants and the members of the receiving culture that most closely

resembles a between-culture contact. Over time, this interaction merges into a within-

culture contact, with successive generations as migrants groups evolve into established

ethnocultural communities within the larger society.

Migration and identity. As common as migration is, it is not the experience of

most people. Migration is a complex psychosocial process with lasting effects on

47

individual identity. For most migrants, their identity impinges on almost everything they

do. So how do migrants, who move between cultures, reconcile their social identities?

Oddly, James (1890, p.213), who “regarded [personal identity] as the most

puzzling puzzle with which psychology has to deal,” had little to say about migration;

but as Brown (1998) points out, James wrote at a time when psychology was mostly the

province of white males operating in their own cultural milieu. Not so Freud, possibly

because as a Jew in ethnically diverse Vienna he felt his ethnicity more personally. In a

1926 address to the B’nai B’rith (a Jewish community organisation) he referred to his

“irresistible” attraction to being a member of an ethnic group, so echoing his own

writings, which claimed the search for identity to be one of the mind’s basic motives

because it guaranteed safety and wish fulfilment (Graafsma, 1994).

Whenever I felt an inclination to national enthusiasm, I strove to suppress it as

being harmful and wrong, alarmed by the warning examples of the peoples among

whom we Jews live. But plenty of other things remained over to make the

attraction of Jewry and Jews irresistible—many obscure emotional forces … as

well as a clear consciousness of inner identity, the safe privacy of a common

mental construction. And beyond this there was a perception that it was to my

Jewish nature alone that I owed two characteristics that had become indispensable

to me in the difficult course of my life. Because I was a Jew, I was free from

many prejudices which restricted others in the use of their intellect; and, as a Jew,

I was prepared to join the Opposition and to do without agreement with the

“compact majority” (Freud, cited in Zavalloni, 1983, p.205)

According to Brown (1998), two aspects of collective identity have received

attention by researchers. One is how people evaluate identity: some minority members

resent or even disown their ethnic identity to avoid the stigma they feel (Lewin, 1948).

The other aspect explored how people maintain their ethnic identity as a minority in a

48

dominant culture. Amin Maalouf (2000), a Lebanese-born writer who has lived and

worked in France for decades, wrote:

How many times, since I left Lebanon in 1976 to live in France, have people

asked me, with the best intentions in the world, whether I felt ‘more French’ or

‘more Lebanese’? And I always give the same answer: ‘Both!’ I say that not in the

interests of fairness or balance, but because any other answer would be a lie. What

makes me myself rather than anyone else is the very fact that I am poised between

two countries, two or three languages and several cultural traditions. It is precisely

this that defines my identity. Would I exist more authentically if I cut off a part of

myself?

So am I half French and half Lebanese? Of course not. Identity can’t be

compartmentalized. You can’t divide it up into halves or thirds or any other

separate segments. I haven’t got several identities: I’ve got just one, made up of

many components in a mixture that is unique to every individual.

Sometimes, after I’ve been giving a detailed account of exactly why I lay claim to

all my affiliations, someone comes and pats me on the shoulder and says, “of

course, of course – but what do you really feel, deep down inside?” (pp. 3-4)

Maalouf has a life story that according to McAdams’s (2003) theory of identity

captures what he is in all his complexity. Unity is important for Maalouf, but he wants a

unity that is comprehensive, not one that cuts out portions of his story to arrive at a false

coherence. His audiences, however, demand a simplification that is fundamentally in

error. Biculturalism can foster pride in one’s origins, solidarity with compatriots, and a

rich appreciation of history; equally it can bring confusion, disappointment and cultural

conflict (Haritatos & Benet-Martınez, 2003). Migrants find themselves simultaneously

immersed in multiple cultures, and so typically have multiple group allegiances.

Liebkind (2008) brought another perspective when he observed that ethnicity is

both ascribed (in that one cannot choose one’s ethnic origin) and achieved (in that it is a

matter of choice). Identification with one’s ethnic roots therefore involves both a

49

cognitive self-categorisation and a sense of attachment. Liebkind (1996) notes that

ethnic/cultural identity has been empirically treated not only as “mere self-

identification” but also as “feelings of belongingness and commitment” towards one’s

group (p. 164).

On its own, ethnic self-identification says little about a migrant’s attitudes towards

the heritage culture or strength of identification with the self-applied category. For

example, Ethier and Deaux (1990) asked Hispanic American college students to name

their ethnic identities and list the characteristics they associated with each. Most (but

not all) claimed a Hispanic identity but associated with it differed meanings: some were

positive, others ambivalent. Likewise, in a study of students from ethnically diverse

high schools in the United States, Phinney and Devich-Navarro (1997) found no

statistical link between self-ascribed ethnicity and degree of involvement in their origin

ethnic group and/or American society. Likewise, in a study of first- and second-

generation Chinese Australian and Chinese American adolescents, Rosenthal and

Feldman (1992) found that ethnic identification (that is, labelling oneself as Chinese)

and knowledge and practice of Chinese culture declined over time, but subjective

evaluation of their Chinese identity and the importance attached to Chinese cultural

practices did not. They concluded that although the behavioural aspects of ethnic

identity diminished, its internal components did not.

In a review of relevant studies, Phinney (1990) found contradictory evidence to

link ethnic identification with degree of acculturation. Some studies showed the two

dimensions to be unrelated: Hutnik (1986) for instance found no significant correlation

between ethnic identity and cultural attitudes among young Indians in Britain. But other

studies gave a different picture: in separate studies of Israeli students in the United

States, Elias and Blanton (1987) found some correlations of behavioural and affective

50

aspects of ethnic identity. Phinney concludes from such studies that “ethnic identity is

not necessarily a linear construct” (p. 509). At a conceptual level, therefore, there exist

two main acculturation models that differ according to whether acculturation is seen as

unidimensional or bidimensional. Gordon (1964) proposed a unidimensional model. At

one end of the spectrum, migrants cling to their own culture; at the other, they adopt the

host culture’s values, attitudes, and behaviours. In a refinement of this simplistic view,

the so-called ‘balance’ model permitted a both cultures to coexist in the form of

biculturalism. However, it still viewed heritage and host cultures as interdependent

rather than orthogonal, and failed to accommodate those who identified strongly (or

weakly) with both cultures. Biculturalism was considered a temporary state in the

transition to full adoption of the host culture (Goldlust & Richmond, 1974;

LaFromboise, Coleman, & Gerton, 1993; Ryder, Alden, & Paulhus (2000).

To address these difficulties, acculturation models were devised that considered

heritage and host cultures as independent influences. These so-called “bidimensional”

models accommodate biculturalism and intensity of cultural involvement as orthogonal

dimensions (e.g., Berry, 1980; La Fromboise et al., 1993; Szapockznik, Kurtiness, &

Fernandez, 1980). They are based on two core assumptions: that individuals differ in the

degree to which their identity includes values, attitudes and behaviours from their

heritage culture; and that individuals can possess multiple cultural identities of varying

independence and strength (Ryder, Alden, & Paulhus, 2000). In models based on the

first assumption, immigrants can indeed relate to both heritage and host cultures

differently, but in different domains of life; they may for instance learn the host

language but still cling to the values they were brought up with. In models based on the

second assumption, the influences of both cultures are independent of each other. The

51

leading model of this kind is the one proposed by Berry (1980, 1984, 1997), described

next.

Berry’s acculturation model. Berry (1980, 1984, 1990, 1997, 2005) developed a

widely accepted framework for acculturation that embodies factors pertaining to both

the group and the individual, and to both the society of origin and that of settlement.

Berry (1980, 1984; Berry & Sam, 1997) was the first to propose that cultural identities

were separate dimensions rather than extremes of a single dimension; or to put it

another way, identifying more with one culture did not mean identifying less with the

other. There were instead four possible psychological responses to cultural change,

which Berry calls “acculturation strategies”: identification with the societies of both

origin and settlement indicates “integration” (real biculturalism); identification with the

culture of settlement alone indicates “assimilation”; identification with the culture of

origin alone indicates “separation”; and identification with neither culture indicates

“marginalisation” (e.g., Berry, 1980, 1985, 1990, 1997; 2005; Berry & Sam, 1997; Sam

& Berry, 2006). Most long term migrants adopt a strategy of “integration” or

“assimilation”, depending on whether or not they respectively retain or discard their

culture of origin. But not all migrants take this path: some either choose or are forced

into a strategy of “separation” or “marginalisation”, again depending on whether or not

they respectively retain or discard their culture of origin.

According to Berry’s (1980, 1984) psychological acculturation model, migrants

settled in the host society face two fundamental choices: whether or not to retain their

migrant culture; and whether or not to seek relations with the host society. There are

four possible psychological response combinations, which Berry calls “acculturation

strategies”: integration, assimilation, separation, and marginalisation. Integration means

maintaining some features of the migrant culture while adopting other features of the

52

host culture, and integrating the two. Assimilation means relinquishing the migrant

culture in its entirety and replacing it with that of the host culture. Separation is its

opposite: maintaining the migrant culture in its entirety and rejecting the host culture.

Finally, marginalisation means rejecting both cultures—their own and that of the

majority. These four strategies can be adopted at the individual or level (Berry, 1980,

1984). Acculturation may, however, vary across the domains of behaviour and social

life: “For example, one may seek economic assimilation (in work), linguistic integration

(by way of bilingualism), and marital separation (by endogamy)” (Berry, 1990, p.217).

Numerous empirical studies have shown the Berry model to have strong

descriptive and explanatory power (Berry, 1997; Redmond, 2000; Sam, 1995). Berry

and his associates (Berry, 1980, Berry, Kim, Power, Young, & Bujaki, 1989)

investigated the acculturation strategies of the following first-generation immigrants to

North America among others: Portuguese, Hungarians, and Koreans in Canada (Berry et

al., 1989); Lebanese in Canada (Sayegh & Lasry, 1993), and Indians in the United

States (Berry & Krishnan, 1992). They found integration to be the preferred

acculturation strategy, followed by assimilation and separation; marginalisation was

least preferred. Other studies confirm migrants’ preference for integration, and

furthermore see flow-on mental health benefits (Niles, 1999). For example, migrants to

Belgium and the Netherlands seek to combine their own culture with that of the

mainstream (e.g., Phalet & Hagendoorn, 1996; Van de Vijver, Helms-Lorenz, &

Feltzer, 1999). In a Canadian study, migrants who opted for integration were minimally

affected by acculturative stress, whereas the marginalized and separated tended to be

highly stressed (Berry, Kim, Minde & Mok, 1987); a tendency confirmed by Berry

(1997), who found that the assimilated experienced intermediate levels of acculturative

stress. However, none of this negates the distinction migrants draw between public and

53

private domains; for example, Turkish migrants in the Netherlands were found to favour

separation in the private domain but integration in the public (Ourasse & van de Vijver

(2005).

Beyond Berry. For all its elegance, Berry’s acculturation model has come under

increasing criticism. In particular, Berry’s “integration” says little about how people

maintain two cultures or why biculturalism is hard to sustain. He assumes that

acculturation to the new culture and to the culture of origin are independent (Berry,

1980; 1990); and although he offers a taxonomy of acculturation, he does not explain

how acculturation type is manifest in migrant identity, nor how migrants negotiate two

cultures. Ryder, Alden, and Paulhus (2000) remedied this by showing that migrants who

prefer integration (in Berry’s terminology) are bicultural in the sense of reporting more

than one cultural identity; and that cultural identifications are independent of measures

of personality, self-identity, and adjustment.

Some researchers consider Berry’s model to be too simplistic to account for the

complexities of acculturation (e.g., Boski, 1998; Weinreich, 1998). Bourhis, Moise,

Perreault, and Senecal (1997) replaced Berry’s social contact and participation

dimension with one pertaining to host culture adoption: “Is it considered to be of value

to adopt the culture of the host community?” (p. 378). Ward (1999) likewise identified

different underlying dimensions of cultural identity. Just as acculturation strategies are

classified on two dimensions (own cultural maintenance and involvement with host

society), migrants’ self-perceptions are also constructed along two dimensions:

identification with their country of origin’s heritage (their cultural or ethnic identity),

and identification with the host society (their created national identity) (Phinney, 1990;

1999; 2000). Moreover, the two dimensions of acculturation strategy are orthogonal, as

are the two dimensions of cultural self-perception. The strategies that emerge, mirror

54

Berry’s four acculturation strategies. Phinney (1990) described four possible outcomes

based on strength of identification with the majority and minority groups. Those who

identify with the dominant culture, while still identifying strongly with their cultural

background, are said to be acculturated, integrated, or bicultural. Those who abandon

their ethnic identity for that of the dominant culture, are said to be assimilated. Those

who decline to identify with the dominant culture are said to be separated. And those

who lose their ties to both cultural groups are said to be marginalised. Research suggests

that a bicultural identity is the most adaptive kind for migrants (Phinney, Horenczyk,

Liebkind, & Vedder, 2001), and bicultural competence, or the ability to interact in two

cultures without tension, is essential (LaFromboise et al, 1993). In short, it is possible to

possess two cultural identities at the same time.

Birman (1994) suggested two models of biculturalism: the alternation model and

the fusion model. In a review of biculturalism similar to that of LaFromboise et al.

(1993), she draws a distinction between the behavioural aspects of biculturalism

(LaFromboise et al.’s “bicultural competence”) and the psychological (“bicultural

identity”), recommending that both be considered. By “alternation” she means the type

of biculturalism where individuals feel acculturated to either one culture or the other

depending on context, and feel a sense of belonging to both. By contrast, according to

the fusion model, migrants exhibit a blended pattern of multicultural behaviour; rather

than having two distinct cultural repertoires that are invoked in different cultural

contexts, they belong to neither culture exclusively but instead to both (Birman, 1994;

Hermans & Kempen, 1998; LaFromboise et al., 1993). In Hermans and Kempen’s

(1998) view an acculturating individual fuses both cultures in a “hybrid” identity that

combines the best of each.

55

In a study of Mexican Americans and African Americans in California,

participants provided quantitative and qualitative data (Phinney, 1992). The study

generated a typology of bicultural identification. “Blended” biculturalism describes

those who identified equally with both cultures and dealt with them by combining them;

they are migrants who, when asked their ethnicity, might answer, say: “I am both

Mexican and American” (see for instance Phinney & Devich-Navarro, 1997).

“Alternating” biculturalism describes those who identified more with their ethnic

culture but were comfortable in both; they are migrants who, when asked to describe

their ethnicity, might answer, say: “I am a Mexican in America,” but not “I am Mexican

American” (Phinney & Devich-Navarro, 1997). Alternating biculturals identify with

both cultures (as do blended biculturals) but are also conscious of cultural discrepancies

that could be a source of internal conflict, so they keep their cultural identities separate

(Phinney & Devich-Navarro, 1997; Vivero & Jenkins, 1999).

There are thus two apparent modes of biculturalism: compatible and oppositional.

The one views their heritage and host cultures as complementary, the other as discrepant

or conflicting. Neither depends on how long an individual has lived in the host culture

(Phinney & Devich-Navarro, 1997) nor on attitudes toward biculturalism (Vivero &

Jenkins, 1999). However, as Vivero & Jenkins note, oppositional biculturalism can

generate internal conflict, which reduces well-being.

Benet-Martinez, Leu, Lee, and Morris (2002) elaborated the compatible-versus-

oppositional dichotomy by proposing that it arose from “bicultural identity integration”

(BII), an individual difference that moderates cultural “frame-switching”, or the act of

moving between cultures in response to situational cues. Bicultural individuals, whether

low or high on BII, exhibit Berry’s acculturative strategy of integration and hence show

56

a bicultural/bilingual pattern of cultural identification and language proficiency (Benet-

Martinez et al., 2002; Benet-Martinez et al., 2006).

In a later study, Benet-Martinez and Haritatos (2005) found that migrants

internalize potentially conflicting cultural identifications. Using their purpose-designed

Bicultural Identity Integration Scale, they scored a sample of first-generation Chinese-

American migrants on BII and correlated the results with measures of personality traits

and bicultural stressors. They inferred the existence of two modes of bicultural identity

integration: cultural conflict, which perceives that cultures clash, and was correlated

with neuroticism, discrimination, and strain in intercultural relations; and cultural

distance, which perceives that cultures are separate, and was correlated with low

openness and living in a culturally isolated environment. They then identified two

groups: the high BII, who feel their cultures can coexist, and who identify more with the

mainstream culture, even though competence in their ethnic culture of origin is often

similar for both groups; and the low BII, who feel conflicted about their two cultures

and see them as oppositional. These findings accord with the “alternation” model for

second culture acquisition (e.g., LaFromboise, Coleman, & Gerton, 1993), as

experimental research showed that often those high in BII also take part in cultural

frame-switching.

Benet-Martinez and Haritatos (2005) also found low-BII persons to be culturally

more complex than high-BII persons, which they attributed to the greater cultural

conflict that low-BII migrants experience. They suggested that low-BII people,

particularly migrants, feel “culturally homeless”, quoting from Vivero & Jenkins’s

(1999) study, in which one bicultural individual is further recorded as stating:

You start building a home in one place within one culture … but do not complete

it. Then you continue to build your home within another culture …. At the end,

57

you have different pieces of home in different places. You can never put them

together, because they may contradict or conflict with one another (p. 7).

Expanding on Benet-Martinez et. al’s (2002) concept of bicultural identity

integration, Boski (2008, p.148) distinguished “harmonious” and “conflictual”

biculturality. The harmoniously integrated experience a consistency between the two

identities: when either is activated, both are parts of the same cultural context and

correspond to each other. But the conflictually integrated experience dissonance

between cultures. This implies two modes of integration, irrespective of competence in

both cultural contexts. According to Boski (2009, p.148), to be bicultural means to

experience a “psychological parallelism” where a person switches between cultures and

functions effectively and coherently in both. But it can also mean that a person

experiences both antagonistically, to the detriment of psychological well-being. Both

depart from the premise that the person is equipped with competence in both cultures:

this is the paradox. What comes out of this is that integration in Berry’s occurs only

when the two identities are harmonious, as it is a contradictory to say that a person is

biculturally integrated when the two identities are in conflict. And yet, Cheng, Lee and

Benet-Martinez (2006) report that both harmoniously and conflictually bicultural

Chinese-Americans endorse Berry’s integration style.

Psychological theories of acculturation focus on changes in attitudes and mental

health (Berry, 1990). According to Phinney and Flores (2002), such changes must be

explained in terms of a two-dimensional model of acculturation. Like Berry and Sam

(1997), they propose that acculturation is not one-dimensional in that an increase in

involvement in one culture implies a decrease in involvement in the other; but two

dimensional in that involvement in one culture is independent of involvement in the

other. To test this proposition, they examined measures of acculturation as predictors of

58

traditional gender role attitudes in a sample of first-, second-, and third-generation

Hispanic migrants. Acculturation was measured in terms of ethnic retention (Spanish

language, Hispanic friendships), mainstream involvement (English language, non-

Hispanic friendships), and traditional gender role attitudes. English language and non-

Hispanic friendships correlated negatively with traditional gender role attitudes; Spanish

language and Hispanic friendships were uncorrelated. Thus the bicultural Hispanic can

be highly involved in, and share the values of, mainstream America, and yet retain

Hispanic social networks and Spanish language. Furthermore, “migrant generation”

(that is, first, second, third etc.) was unrelated to social networks, Hispanic and non-

Hispanic, or to English language usage. Thus, knowing a person’s migrant generation

reveals little about the effects of cultural contact on the individual. Such findings

powerfully support Berry and Sam’s (1997) two-dimensional model of acculturation in

that mainstream adaptation and ethnic retention are revealed to be independent of each

other and to have different outcomes.

Researchers (e.g., LaFromboise, Coleman & Gerton, 1998) agree that

“successful” migrants acquire social skills, cultural knowledge, and other aspects of the

society of settlement. Those who do not achieve bicultural competence may experience

acculturative stress in the form of hostility, anxiety, identity confusion, depression, and

feelings of marginality and alienation. The implication is clear: acculturation strategy

matters, because it affects well-being; likewise cultural identity, since an acculturation

strategy is defined by the migrant’s cultural identity (or identities).

3.3 Acculturation Outcomes

The literature views acculturation outcome from three conceptual angles:

psychopathology, behaviour, and acculturative stress. Psychopathology sees it as a

stressor, hence potentially problematic. Migrants, it implies, require assistance in

59

coping, But according to Bonanno (2004) and Ward (1996) this cannot be so, as most

migrants relocate successfully, with health, psychological and social outcomes like

those of the mainstream. When viewed in terms of observable behaviour, acculturation

is usually seen as unproblematic. It comprises three sub-processes: culture shedding,

culture learning, and culture conflict (Berry, 1992). In a process of “adjustment” (Ward

& Kennedy, 1993b), culture shedding and culture learning entail replacement of

inappropriate behaviours with ones that sit better with the host culture. They are

perceived as accomplished with little difficulty, but if conflict occurs, it is resolved by

the migrant’s yielding to the norms of the mainstream. Assimilation is the likely

outcome. Lastly, when viewed in terms of acculturative stress, acculturation is judged to

be problematic but surmountable (Berry, Kim, Mindle & Mok, 1987). Migrants

experience changes that challenge their cultural norms in ways that cannot be dealt with

by adjusting or assimilating.

The psychopathology of culture shock. Migration always causes a certain

“culture shock” (Garza-Guerrero, 1974). The anxiety that ensues challenges the stability

of the migrant’s “psychic organisation” and

a growing sensation of discontinuity of identity emerges. It is as if, out of his

usual habitat, the newcomer no longer has the necessary corroborative

environmental feedback for his ego identity …. The severity of the threats to the

individual identity runs parallel to the severity of concomitant mourning. Thus,

the more serious the break with the newcomer’s continuity of his identity, the

greater this yearning for those lost love objects (abandoned culture) which in the

past provided a comfortable sense of continuity. On the other hand, the greater the

longing for those love objects, the more afflictive the threats to his identity.

(Garza-Guerrero 1974, pp. 418–419)

For Volkan (2004), issues of personal identity are paramount for migrants.

Migrants go through a mourning process, which may be complicated by feelings of guilt

60

over loss of parts of the self (e.g., the migrant’s previous identity, his or her investment

in the language, culture, land, and people left behind). Guilt may trigger other

psychological processes, such as depression, identity confusion, shame, helplessness,

and culture shock. It is this culture shock that embraces adaptive processes that may

include reactions of anxiety, depression, fear, uncertainty, and anger.

Among the best known early writers on international relocation is Oberg (1960)

who discussed “culture shock”, the disorientation and confusion that results when a

person enters a second culture and familiar cues are absent. This results in feelings of

loss, confusion over roles and role expectations, and feelings of loss of control. Oberg’s

detailed four phases of emotional reactions: (a) the “honeymoon”, characterised by

euphoria, enchantment, fascination, and enthusiasm; (b) the crisis, characterised by

feelings of inadequacy, frustration, anxiety and anger; (c) the recovery, including crisis

resolution and culture learning; and finally (d) adjustment, characterised by enjoyment

of, and functional competence in, the new environment. Oberg had in mind people who

were more or less suddenly exposed to a completely unfamiliar setting and largely

overwhelmed by it. Some of the affective responses consistently mentioned in the

literature included confusion, anxiety, disorientation, suspicion, bewilderment,

perplexity and an intense desire to be elsewhere.

This “honeymoon period” of culture shock may be contrasted with hypotheses

arising from the stress-and-coping literature which predict that, in contrast to “entry

euphoria”, immigrants suffer the most severe adjustment problems at the initial stages

of transition when the number of life changes is the highest and coping resources are

lowest (Ward, Bochner & Furnham, 2001). Early research on culture shock compared it

to a disease, to be cured with appropriate treatment. This approach highlights the factors

that reduce the distress of culture contact, including personal resources such as self-

61

efficacy and emotional resilience, and interpersonal assets such as social support. The

affective outcome of a well-adapted migrant in the new culture is a psychological

adjustment.

Acculturation stress. The migrant must accommodate different beliefs, routines,

and social roles (Hovey, 2000). The term “acculturative stress” captures the stress and

loss this entails. For Lazarus & Folkman (1984), acculturative stress is an outcome of

the losses that occur when adopting a new system of beliefs, routines, and social

roles (Hovey, 2000). It draws on the broader paradigms of stress, adaptation, and

coping. Berry, Kim, Minde & Mok (1987) concur, seeing acculturative stress as a

reaction to life events rooted in the experience of acculturation, with depression, anxiety

and psychosomatic problems as some of its adverse consequences (Berry, 1997; Berry

& Sam, 1997).

Under this approach, the long-held prevailing view was that acculturation

inevitably brought about a decline in mental health. However, this view is no longer

supported (Berry & Kim, 1988). Many factors mitigate acculturative experiences such

that mental health is unaffected or only slightly so. Through adaptation the

consequences of acculturative stress may be positive as migrants develop skills and

behaviours that enhance mental and physical health, and social functioning.

In this regard, Okazaki (1997) explored the role of ethnicity in emotional distress

in a sample of Asian-born migrants to the United States. Drawing on Markus and

Kitayama’s (1991) model of self-construals, she found that Asian Americans are

socialised to hold highly interdependent self-construals; in consequence they are more

attuned to social cues and so more susceptible to social anxiety than others. After

controlling for depression and social anxiety, ethnicity and self-construal predicted for

social anxiety, and self-construal alone predicted fear of negative evaluation.

62

Building on work by Okazaki (1997), Norasakkunkit and Kalick (2002) measured

self-reported depression, social avoidance and distress, and fear of negative evaluation

in a sample of European Americans and Asians. They found that for the Asians, but not

the others, the measures were positively related to interdependent self-construal, and

negatively related to independent self-construal and self-enhancement. This was

explained in terms of the Asians’ greater sensitivity to relationships, given their greater

interdependent self-construal.

On the assumption that acculturative stress diminishes with time, researchers in

the 1980s and 1990s began to use acculturation level as a proxy for acculturative stress.

Their findings are contradictory. For example, studies of migrants’ attitudes in Canada

showed that an integration strategy produces the lowest acculturative stress (Berry,

1990; Berry & Kim, 1988; Berry et al. 1987). Conversely, in a study of acculturative

stress and depression, Hovey and King (1996), found that acculturation level and stress

to be unrelated. It may therefore be wrong to assume that the less acculturated (e.g.,

separated) experience more acculturative stress than the more acculturated.

Many factors contribute to successful acculturation at the individual level:

personal experience of each individual, often associated with age and gender; social

status such as education, prior contact with diverse cultures, familiarity with the

language and culture of the host community; and personality. Moreover, the greater the

cultural distance between the origin and host society, the more difficulties migrants

encounter (Ward & Kennedy, 1992; Ward & Searle, 1991; Hirst, 2014).

Acculturation adaptation. Except perhaps for social identity theory (Ward,

2001), acculturation studies usually conceptualise individual changes either as a

response to stress (Berry, Kim, Minde & Monk, 1987), or as a response to a need, in

this case to learn cultural skills so as to thrive and survive (Furnham & Bochner, 1986).

63

Acculturation results in the migrant’s learning skills that are adaptive and functional

(Berry, Poortinga, Segall & Dasen, 1992). Adaptation takes various forms, though all

are largely stable changes at an individual or group level in response to external

demands (Berry, 1997; Ward, 2001). It does not require that migrants become more like

their environment (a strategy of assimilation); it may imply resistance, efforts to change

the mainstream, or distancing (a strategy of separation). And the outcome varies

accordingly, from material, social and mental success, to the inability to function

effectively in the new society (Berry, 1997).

As illustrated in Figure 3.2, the literature distinguishes three theoretical

approaches to acculturation, one that invokes psychological models of stress-and-coping

as applied to cross-cultural adaptation; a culture learning approach; and a social

identification approach (Ward, Bochner & Furnham, 2001).

The stress-and-coping approach views cross-cultural transitions as a series of

stress-provoking life changes that require coping responses from the migrant. The

culture learning approach focuses on the skills and behaviours required in order to

succeed in a new culture; adaptation is a result of learning new and appropriate skills

(Bochner, 1972, 1986). Lastly, the social identification approach is associated with

social identification theories, and is concerned with values, attitudes, beliefs,

expectations, self-identity, cultural identity (Taft, 1977).

Another distinction found in the literature is that between psychological and

sociocultural adaptation (Searle & Ward, 1990). Psychological adaptation covers

depression, anxiety and psychosomatic problems, which are the most common mental

health consequences of acculturating individuals (Berry, 1997; Berry & Sam, 1997).

Sociocultural adaptation covers the relationship between individuals and their cultural

context, and concerns the success or otherwise of participation in the host society

64

(Ward, Bochner & Furnham, 2001; Searle & Ward, 1990; Ward et al., 2001; Ward &

Kennedy, 1994). In this framework, the stress of acculturation results in the acquisition

of adaptive and functional coping skills (Berry, Poortinga, Segall & Dasen, 1992)..

Figure 3.2 The “ABC model” of acculturation, adapted from Ward, Bochner &

Furnham (2001)

Though separate, psychological and sociocultural adaptation are linked, for

success in the host culture are likely to boost feelings of well-being, leading to a

virtuous circle. Two studies provide empirical support for this: Berry (2003) found

correlations in the range of .4 to .5 between measures of psychological and sociocultural

Affects Behav-

iours

Percep-

tions

Developing

changing,

maintaining

identity

Acquiring

specific

skills

Coping

with

cultural

change

Cognition Social

identification

theories

Affect

Stress and

coping

theories

Behaviour

Culture

learning

theories

Processes

Outcomes

65

adaptation; Ward and Kennedy (1999) also found them to be positively correlated.

Moreover, the strength of the association was positively related to with cultural

proximity to, and integration in, the host society (Ward, 1999; Ward & Kennedy, 1996;

Ward & Rana-Deuba, 1999).

Several studies on cross-cultural adjustment have focused on differentiating

psychological and sociocultural forms of adjustment. Berry and colleagues (2002)

argued that, while empirically related to some extent, psychological and sociocultural

adaptation are distinct in time-scale and causation. Psychological problems often

increase soon after contact, declining over time, whereas sociocultural adaptation

typically proceeds linearly. Likewise, psychological adaptation is mainly predicted by

personality and social support, whereas sociocultural adaptation is predicted by other

factors: cultural knowledge, degree of contact, and intergroup attitudes (Berry et al.,

2002).

In order to elaborate and test the distinction between psychological and

sociocultural adjustment Searle and Ward (1990) investigated a sample of Malaysian

and Singaporean students in New Zealand. Depression (a psychological measure) and

social difficulty (a sociocultural one) were tested for their relationship to expected

difficulty, cultural distance, quantity and quality of extraversion, life changes, and other

variables. They found that social difficulty, satisfaction with host contacts, extraversion,

and life changes accounted for 34% of the variance in psychological adjustment and

cultural distance, expected difficulty, and depression accounted for 36% of the variance

in sociocultural adjustment. No significant differences were found for either

psychological or sociocultural adjustment on length of residence. They concluded that

satisfying interactions with members of the host culture predicted psychological

adjustment. More specifically, they found that while the quantity of host national

66

contact predicted socio-cultural adjustment, it was the quality of the contact that

predicted psychological adjustment. For relationship satisfaction implied adequate

social support, and relationship dissatisfaction implied increased psychological

distress—findings that are congruent with the clinical literature on social support and

mental health (Ward & Kennedy, 1992). In contrast, Ward and Kennedy (1993) found

that it was relations with co-nationals which form the important social support network,

and that it was satisfaction with these relations that underpins psychological adjustment.

Acculturation strategies and adaptation. Any acculturation strategy that

migrants prefer can be experienced as stressful if they have to strive hard to realise their

aims regarding successful adaptation to a new culture (Berry, 1992; Berry & Sam,

1996). In her review of research on acculturation, Ward (1996; Ward & Kennedy, 1993)

found that both psychological and sociocultural adaptation are helped by adopting an

acculturation strategy of integration and by minimising cultural distance. Research

relating adaptation to acculturation strategies allows for some generalisations (Berry,

1997; Ward, 1996). In most cases, the integrated appear to be best adapted, the

marginalised are worst adapted, and the assimilated and the marginalised lie in between.

Other research into acculturation strategies and psychological adaptation as an

outcome variable has been contradictory. Sing (cited in Ourasse & van de Vijver, 2005)

found that identification with the host culture was positively related to stress; Padilla

(1986) found the opposite. Ghaffarian (1987) found that identification with the host

culture was negatively related to depression; Kaplan and Marks (1990 found the

opposite. Buriel, Calzada, and Vasquez (1982), moreover, found that a balanced

identification with both cultures was positively related to psychological adjustment;

Ward and Kennedy (1994) found differently: identification with the host culture was

67

positively associated with sociocultural adjustment, but identification with the migrant

culture was negatively related to psychological problems.

For many researchers, integration is the preferred acculturation strategy, mainly

because it is seen to promote mental health (Niles, 1999). They find, for example, that

most migrants to Belgium and the Netherlands want to combine their native mainstream

cultures (e.g., Phalet & Hagendoorn, 1996; Van de Vijver, Helms-Lorenz, & Feltzer,

1999). Likewise, the psychological adjustment of migrants to Canada could be ranked

from highest to lowest as follows: integrated, assimilated, separated and marginalised

(Berry, Kim, Minde & Mok, 1987; Berry & Annis, 1974)

3.4 Individualism versus Collectivism

This section examines the individualism–collectivism dimension that researchers

such as Hofstede (1980, 1984) have used to characterise cultures. Individualism–

collectivism is not the only dimension that has been used in this way (Hofstede initially

defined four) but it holds special relevance for this thesis because it parallels the

dimensions by which the dialogical self is characterised, and hence is discussed in depth

here and referred to in later chapters.

Using the previously mentioned self-confrontation method (Hermans & Hermans-

Jansen, 1995), the dialogical self can be characterised in terms of two motives, self-

enhancement and union-with-the-other, whereas cultures can be characterised as

“individualistic” or “collectivist”. Although the parallel is not straightforward, it

suggests that there is a link between self and identity on the one hand, and migration

and acculturation on the other.

Anthropologists and cultural psychologists have long argued that personhood is

constructed differently in different cultures (for instance Shweder & Bourne, 1984).

Geertz (1975) noted the distinctive features of the Western conception of personhood:

68

… the Western conception of the person as a bounded, unique, more or less

integrated motivational and cognitive universe, a dynamic centre of awareness,

emotion, judgement and action organised into a distinctive whole and set

contrastively both against other such wholes and against its social and natural

background is, however incorrigible it may seem to us, a rather peculiar idea

within the context of the world’s cultures. (p. 48)

In the West, the self is generally conceived of as autonomous and separate from

other people or social contexts (Geertz, 1975). Non-Western countries tend to stress

group solidarity and connectedness, viewing the self as embedded in social roles, duties,

and responsibilities. Shweder and Bourne (1984) suggested that the concept of the

person varies across cultures. They distinguish two main types of personal organisation:

“egocentric-contractual” (individualistic, subjected to and limited by sociocultural

rules), and “sociocentric-organic” (outward-directed, closely integrated with the

surrounding environment). Various other distinctions and perspectives exist:

achievement orientation and affiliation orientation, self-directed and other-directed or

field dependent.

For Sampson (1988), individualism is more a philosophical and practical principle

than a universally espoused standard of social life. Referencing Heelas and Lock (1981),

he observes that all cultures distinguish self and the nonself, but do so very differently.

American society promotes a self-contained individualism that fosters individual

freedom and achievement, whereas “elements of Japanese, Chinese, and Islamic

indigenous psychologies offer ensembled type [sic] of individualism” (Sampson, 1988,

p.17). Other researchers have largely concurred. For example, societies can be

characterised by their location on the individualism-collectivism dimension (Triandis,

Bontempo, Villareal, Asai, & Lucca, 1988). According to Hofstede (1980, 1984), in

individualistic societies, individuals are expected only to look after themselves and

69

immediate family; in collectivistic societies individuals are integrated into cohesive in-

groups that protect them in return for group loyalty. In such a culture the individual’s

“selves” largely derive from, and are sustained by, the groups to which they belong.

Research by Hofstede (1980), Triandis (1990), Triandis, Bontempo, et al. (1986)

and Triandis, Bontempo, Villareal, Asai, & Lucca (1988) found collectivism to be

highest in traditional societies, and in East Asia and most of Africa, Asia, Latin

America, and the Pacific Islands. Hofstede (1980) found that the most individualistic

countries were the United States, Australia, and the UK,. Moderate levels of

individualism are found in the rest of Europe. In a refinement of Hofstede’s (1980)

conceptualisation of individualism and collectivism as bipolar opposites, Triandis et al.

(1986) found that at the cultural level, each construct of individualism and collectivism

consists of two factors. Individualism he defines as separation from in-groups, plus self-

reliance combined with hedonism; collectivism he defines as family integrity, plus

interdependence combined with sociability. They suggested that individualism and

collectivism are not opposite poles of a single bipolar construct, but are independent

constructs.

Triandis, Leung, Villareal, and Clark (1985) distinguished between cultural

individualism and collectivism, and psychological individualism and collectivism,

which they term allocentrism and idiocentrism. Allocentric individuals subordinate their

personal goals to the interests of their collectivity and may derive satisfaction in doing

so. Idiocentric individuals, on the other hand set their personal goals above the interest

of any collectivity. While allocentrics in individualist cultures join communes and other

collectives, idiocentrics in collectivist cultures try to escape the dominance of their in-

groups (Triandis et al, 1985).

70

Along similar lines, many researchers contrast Western conceptions of the self,

which emphasise the separateness of the individual, with non-Western conceptions that

emphasise relatedness. For example, Markus and Kitayama (1991) found that the

Western cultures seek to maintain the independence of the individual as a separate, self-

contained entity; individuals are socialised to be unique, to express themselves, and to

realise and actualise their inner self and promote personal goals. The emphasis is on

personal attributes, demonstrated in public and confirmed in private through social

comparison (Shweder & Bourne, 1984), and to “be true to one’s own internal structures

of preferences, rights, convictions, and goals” (Markus & Kitayama, 1994, p.569).

Non-Western cultures, by comparison, do not value overt separateness but tend to

emphasise the “fundamental connectedness of human beings” (Matsumoto, 2000).

Under the interdependent self-construal (Markus & Kitayama, 1991), individuals focus

on their interdependent status with other people: the self is unbounded, flexible, and

contingent on context. Furthermore, the most salient aspects of the self are defined in

relationships—that is, those features of the self related to and inseparable from social

contexts. The overall aim is to maintain social harmony.

Pervin (1999) observed that although the distinction between the independent and

interdependent self is helpful, it is far from the whole story. Selves are not

dichotomously independent or interdependent; they may be both (Singelis, 1994). That

is, collectivism and individualism are two dimensions, not opposite ends of a single

dimension (Rhee, Uleman & Lee (1996). Moreover, there is considerable variation in

self-construal within a single culture. Matsumoto (2000) points out that the concept of

independent versus interdependent selves is not unlike other dualities of the self and

human nature, including Freud’s (1930/1961) “union with others” versus egoistic

happiness, Balint’s (1959) ocnophilic and philobatic tendencies, Bakan’s (1966)

71

communion and agency, Boen’s (1966) togetherness and individuality, Bowlby’s (1969)

attachment and separation, Franz and White’s (1985) individuation and attachment,

Stewart and Malley’s (1987) interpersonal relatedness and self-definition, and Slavin

and Kriegman’s (1992) mutualistic and individualistic urges (all cited in Guisinger &

Blatt, 1994).

In elaboration of this notion, Guisinger and Blatt (1994) acknowledged that

although mainstream psychology has historically placed more importance on self-

development for autonomy than for interpersonal relatedness, the two processes are not

mutually exclusive. Subsequently, Niedenthal and Beike (1997) proposed the existence

of both interrelated and isolated self-concepts. Like Guisinger and Blatt (1994), they

argue that these concepts exist not as dichotomies but as interrelated dualities, and that

an individual can possess both self-concepts in the same domain simultaneously.

According to Matsumoto (2000) the implication of the existence of these dualities

means that cultures may emphasise both types of self-construal rather than only one.

Finally, Hermans and Kempen (1998) argued that

Globalization involves social processes that are complex and laden with tension.

These processes fall squarely outside the scope of cultural dichotomies, which by

their nature are oversimplifying and insensitive to the apparent tensions that are so

typical of the relationships between cultural groups. (p. 1112).

They concluded that theories involving cultural dichotomies are flawed since they

take cultures to be homogeneous, distinct, and geographically delineated. According to

Hermans and Kempen, globalisation shows that cultural identities are neither fixed nor

static, but fluid, dynamic, negotiable, and in flux; they cite the example of individuals

who grow up in bicultural environments and experience cultural dialogues within

themselves that represent a dynamic and interactive multiplicity of different and

contrasting voices and roles.

72

Hermans is not without his critics. Roland (2001), a psychoanalyst who has

worked in India and Japan and with patients from diverse cultures and ethnic

backgrounds in New York, has expressed his reservations in these words:

[Hermans’] dismissal of the work of Markus and Kitayama (1991), Marsella

(1983), Shweder and Bourne (1984), Triandis (1980) and others as too

dichotomous, as being too governed by a Cartesian dualist of the ‘West against

the rest’, is really a continuation of his agenda to flatten out radically different

cultural/psychological worlds (p. 317).

He notes that his own research in inter-Asian analyses indicated that “radically

different cultural psychological worlds do exist” (Roland, 1988).

Tweed, Conway and Ryder (1999) have also responded to Hermans and

Kempen’s (1998) criticism that mainstream cross-cultural psychology sees cultural

differences in terms of strict, stable dichotomies defining homogeneous groups. They

note that Triandis’s (1989) research “has already hinted at the complexity of self and

identity with his assertion that each individual has multiple selves from which to

sample” (p. 838). In terms of cultural conceptions of the self, Triandis (1989) argued

that there are three kinds of self-conceptions: private, public, and collective selves.

Private selves are the aspect of self-conceptions that are shown and known to oneself;

public selves are those aspects that are presented to others; and collective selves

represent the self as members of various groups. Markus and Kitayama’s

conceptualisation of interdependent self-construal “conflates Triandis’s public and

collective selves into one type of conception of the self that emphasises its social and

contextual embeddedness” (Kashima, 2001, p.213). Kashima further notes that the

argument has been that Triandis’s public and collective selves, and Markus and

Kitayama’s interdependent self-construal, are more prevalent and emphasised in

collectivist cultures such as East Asia and Latin America, while public selves and

73

independent self-construal are more prevalent and emphasised in individualist cultures

such as Western European-based cultures including North America and Australia.

Triandis’s (1989) differentiation of the self along three dimensions has been

further developed by other researchers. For example, Brewer and Gardner (1996)

proposed a distinction between three sources of self-representations: the personal, the

interpersonal, and the collective. They propose that

At the individual level, the personal self is the differentiated, individuated self-

concept most characteristic of studies of the self in Western psychology. At the

interpersonal level, the relational self is the self concept derived from connections

and role relationships with significant others. This corresponds most closely to the

interdependent self as defined by Markus and Kitayama (1991). Finally, at the

group level, is the collective self, which corresponds to the concept of social

identity as represented in social identity theory and self-categorisation theory.

(Turner, Hogg, Oakes, Reicher, & Wetherell, 1987, p.84)

Along these same lines, Kashima (2001) postulated a concept of self in relation to

“at least three major classes of other entities: goals, individuals, and groups” (p. 214):

A self in relation to a goal is an agentic self; a self in relation to another individual

is a relational self; and a self in relation to a group is a collective self. These three

aspects constitute three components of the meaning of a self. (p. 214)

Recently something of a consensus has developed. Most researchers now agree

that all cultures accommodate multiple self-construals of the self to varying degrees,

and that different selves manifest in different contexts (Church, 2000).

3.5 Summary

This chapter first described Australia’s migration program: its history, the policies

that shaped it, and the social forces that sustained it—and still do. The chapter then

explored the psychological aspects of the acculturation that commonly accompanies

74

migration. James and the early psychologists paid scant attention to acculturation,

probably because they lived in an era of limited international migration by the

bourgeoisie, who were then the only ones to attract the serious attention of the

psychologists. Later, anthropologists and sociologists took an interest, but did so from

the perspective of their own disciplines. Arguably, Berry was the first to develop a

psychological model of acculturation, which subsequently attracted great attention from

psychologists and others; and for that reason it is explained at length. Also explained are

the many models that succeeded Berry but which, like Berry, are essentially static in

nature and not dynamic. The next chapter demonstrates how the theory of the dialogical

self—as a dynamic, process-oriented model—has the potential to address many

criticisms of Berry’s model and those of his successors.

75

CHAPTER 4: The Dialogical Self and Acculturation

The two preceding chapters described theories of the self and acculturation

independently. This chapter brings them together by showing how one particular theory

of the self, the dialogical self, offers a framework to analyse and explain the

psychological dimensions of acculturation in a way that Berry’s acculturation model

(and other models constructed along similar lines) do not.

The chapter opens with a discussion of the phenomenon of globalisation, for it is

this that has caused researchers from Hermans (1998) onwards to question previous

models of acculturation. Next follows a section showing how dialogical self theory

accounts for aspects of acculturation in an era of globalisation—in particular cultural

continuity—that were troubling for previous models. The chapter concludes with the

research questions addressed in this thesis, having first placed them in the context of the

theories of migration and acculturation.

4.1 Globalisation and the Illusion of Stability

Globalization is “a process by which cultures influence one another and become

more alike through trade, immigration, and the exchange of information and ideas”

(Arnett, 2002; p. 774). Although globalisation began centuries ago with the rise of the

great trade-driven European maritime empires (Friedman, 2005), it has come to the fore

in the public consciousness only in recent decades (Kinnvall, 2004; Hermans &

Dimaggio, 2007). But for all its boons, globalisation has a “shadow side”, to which

Hermans and Dimaggio attach the descriptor of “uncertainty”. It is against the backdrop

of globalisation, and the psychological influence that it exerts, that traditional models of

acculturation have been found wanting, and which dialogical self theory addresses.

76

The challenge of globalisation. By revealing weaknesses in previous

acculturation models, globalisation has thrown up challenges for cultural psychologists.

According to Kadianaki (2009; p. 477), whereas migration is “complex and dynamic”,

past theories focus “on states rather than processes”; and what is needed instead is a

theory that captures “life histories of individuals as they develop in space and time”.

Dialogical self theory addresses these issues: as a theory it is inherently process-

oriented (Andreouli (2013), and as a metatheoretical substrate for a range of

psychological “complex and dynamic” phenomena such as acculturation, it lends itself

to ready elaboration (Hermans & Dimaggio, 2007). It is, thus, a “language” in which

dynamic models and theories can be couched.

The remainder of this section makes the case that previous models, being static

and dichotomous in nature, do not adequately account for cultural identity on a world

increasingly populated by cultural hybrids.

Blurred boundaries: the hybridised identity. The impact of globalisation on

culture is nowhere better captured than in the concept of the “hybridised identity”. The

term seems to be of recent origin, along with globalisation itself. Over a century ago,

James (1890) recognized that the self is shaped by cultural and community, but wrote

little on the subject. A century later, Bruner (1986, 1990) provided a cultural

perspective, as did Hermans, Kempen, and van Loon (1992), who argued that ”the

concept of the dialogical self is based on the notion that people have told each other

stories and listened to stories across cultures and across times” (p. 108; emphasis

added).

In this vein, Hermans (2004) observed that cultural identities and boundaries are

far from fixed. With the metaphor of the dialogical self as a “society of mind”, he

highlighted its similarity to society itself, replete with shifting coalitions, tensions and

77

factional conflict (Hermans, 2002a). This multivoiced aspect of the dialogical self

therefore renders it particularly suitable for analysing cultural identity (Hermans &

Dimaggio, 2007) because the pace of social change has led to new forms of cultural

identity and “hybridized” identities (Hermans & Kempen, 1998):

...[such] as Mexican school girls dressed in Greek togas dancing in the style of

Isadora Duncan, a London boy of Asian origin playing for a local Bengali cricket

team and at the same time supporting the Arsenal football club, Thai boxing by

Moroccan girls in Amsterdam, and native Americans celebrating Mardi Gras in

the United States. (p. 1113).

Hybridised identities are not a developmental end-point but rather they shape a

person’s dialogical negotiations in the context of racism, gender, imperialism, and

power (Bhatia & Ram, 2001; Bhatia, 2002; Da Cunha, 1995). Hermans (2001a)

explained how dialogical self theory accommodates cultural misunderstanding:

When people are raised in one culture and then migrate to another, they arrive in a

situation in which two or more heterogeneous internal positions (e.g., I as

Egyptian and I as Dutch) interact with a multiplicity of very heterogeneous

external positions (e.g., the family of one’s culture of origin and individuals and

groups representing the host culture). Such positions (e.g., Egyptian versus Dutch)

may be felt as conflicting or they may coexist in relatively independent ways or

even fuse so that hybrid combinations emerge in the form of multiple identities (p.

258).

It is the multi-faceted nature of cultural identities (Hermans & Kempen, 1998)

that prompts the question of “how people, involved in a process of acculturation,

organize and reorganize their self-system in such a way that they are able to share with

other people cultural elements that may be highly divergent, partly unknown and laden

with power differences” (Hermans, 2001a, p.258). For Hermans, acculturation is an

78

“intercultural phenomenon par excellence because it takes place on the contact zone

between someone’s home culture and host culture” (p. 269).

Hermans (2001a) noted that Bhatia and Ram (2001) called into question Berry’s

(1980) acculturation model and strategies:

One of the main assumptions is that, although there are considerable variations in

the life circumstances of the cultural groups involved, the psychological processes

that operate during acculturation are essentially the same for all groups. (pp. 269-

270)

Berry, as has been stated, claimed that his “integration” strategy was optimal for

the acculturating migrant since it provided the best social support, allowing involvement

in the host culture while maintaining connections to home (Berry, 1980). But Hermans

(2001a), again echoing Bhatia and Ram (2001), observed that Berry’s model mistakenly

implied that the psychological processes were the same for migrants from Western

Europe as for those from former colonies. Berry’s model also considered integration “to

be a linear trajectory that leads to an end-goal” (Hermans, 2001a; p.271), but did not

show how that goal could be achieved given the power and status differences between

the majority and minority cultures.

In his review of Hermans (2001a), Roland (2001) agreed with Hermans in

rejecting cross-cultural psychology’s “assumption that [acculturation] is the same for

everyone and that everyone comes to it with the same psychological make up” (p. 318).

He further agreed that it is the process of acculturation that is more important than its

outcome or endpoint. Roland only questions whether an endpoint is ever achieved, as

“there is always the issue of how much of the indigenous self is kept or disavowed in

relationship to a newer self in the eventual makeup of a bicultural self” (p. 318),

especially when a person migrates from a different “civilisation”.

79

Nevertheless, Adams and Markus (2001) acknowledged that the dialogical self

model, with its emphasis on the multiplicity of identity and the agency of the self as a

constructor of identity, makes an important contribution as a model of identity

negotiation. They saw the model as illustrating how the self functions as a “zone of

proximal development”, where both personal-identity and cultural-identity categories

“make each other up” (p. 293). The dialogical self supersedes the one-dimensionality of

other paradigms, such as the acculturation studies paradigm (Berry & Sam, 1997),

which view cultures as mutually distinct and internally homogeneous, and presupposes

a linear conception of cultural change. In the dialogical self paradigm, identities can be

hyphenated and hybrid constructions made of heterogeneous and contrasting elements.

Lastly, Falmagne (2004) welcomed the dialogical self model for its analytical

sophistication in respect of the self, although she criticised it for its lack of a societal-

level frame of analysis. The model sees the self as “continuous with culture” (Hermans,

2001a), but according to Falmagne it has little to say about the politics of culture

(Squire, 2000), and the societal system of which culture is a constituent.

False dichotomies and reification. Thus far, traditional acculturation models

have been criticised on empirical grounds. Criticism has been levelled on

methodological grounds as well. Traditional acculturation models, it is said, force

acculturation into a false dichotomy. This leads to the logical fallacy of reification,

whereby a category that is implied by theory is treated as if it existed in reality when in

fact (given that the theory is incorrect) it does not (Schwartz, 2010).

Foremost among these critics is Rudmin (2003), who observed that a common and

serious failing of theorists was to ignore the obvious fact that in any acculturative

setting there is not one culture but two; so each of Berry’s four acculturative strategies

80

applies not only to the minority culture but to the majority culture as well, making 16

possible combinations, not the mere four that Berry and his fellow theorists discuss.

To suggest that minorities are psychologically reactive to intercultural contact and

that dominant groups are not almost implies that minority people are a different

species of psychological being, one distinct from the majority…. To the contrary,

as a result of the speed and ease of world travel, global communications, and

international marketing, all humans, everywhere, are subject to acculturation

processes, whether they know it or not and whether they like it or not (p. 6).

He went on to criticise researchers’ lack of logical rigour, which explains why

their studies have variously found measures of incompatible acculturative attitudes to be

positively correlated, and why bicultural integration and marginalisation are confounded

constructs, and why there is no robust evidence that biculturalism is most adaptive. This

latter presumption he found so irksome that he was moved to complain that “studies

showing minorities to have superior health are called ‘counter-intuitive’ and a

‘paradox’…. In fact many studies using large samples show immigrants to have

superior mental health…” (Rudmin (2010b, p. 302). He also drew attention to other

errors and claims that acculturation research has produced little and has been guilty of

confused concepts, circular reasoning, wrong research questions, and “mistaken

citations of mistaken interpretations of studies using mistaken measurements” (Rudmin

(2010a; 2010b, p. 299).

Gillespie, Howarth and Cornish (2012) have also criticised traditional

acculturation models on methodological grounds. Social categories, they claim, are

“perspectival” in that categorization stems from social position, history and particular

interests, none of which is objective; social categories are “historical” in that they and

their members are in constant flux; social categories are blurred by the movement of

people between categories; social categories are “re-constitutive of the phenomena they

81

seek to describe” (p. 1) because people react to categorisation; and social categories

oversimplify by blurring intra-group differences and accentuating inter-group

differences.

Even dialogical self theory itself has not escaped methodological criticism, though

unlike the criticism directed at older models, it is not deemed to be fatal. Adams and

Markus (2001) voiced misgivings as to “whether the dialogical self meets the objective

of a non-reifying account of culture and self” (p. 284). They observed that Hermans’

“treatment of culture as positions” as in itself entailing the “reification of dynamic

flowing [cultural] patterns into static, fixed entities” (p. 284). Further, they see

Hermans’ notion of positioning as based on the premise that “selves and ‘cultures’ can

be represented or ‘fixed’ as locations in space” (p. 284); and that the Hermans’

metaphor of dialogue “restricts cultural influence to knowable, objectified ‘me’ aspects

of the Jamesian self and appears to neglect or understate the possibility of cultural

influence on knowing, subjective ‘I’ aspects”, with the implication of “an essential, core

self—whether unitary and continuous or multiple and discontinuous—that is insensitive

to cultural variation” (p. 284). Notwithstanding, they conclude that

As a model of identity negotiation, Hermans’ dialogical self makes important

contributions: it emphasizes the multiplicity of identity, highlights the agency of

the self as a constructor of identity, and suggests the importance of psychology—

and the study of self, in particular— for the study of culture (p. 283).

In summary, researchers have criticised traditional acculturation models on two

main grounds, empirical and methodological. At the empirical level, researchers have

argued that traditional models do not adequately fit the facts since they cannot

accommodate the blurred boundaries of cultural hybrids; and at the methodological

level, researchers have argued that traditional models and the methods by which they

are applied are logically flawed.

82

4.2 A Dialogical Theory of Acculturation

The previous section made the case that traditional acculturation models do not

adequately explain acculturation. This section makes the converse case by showing how

dialogical self theory explains the epiphenomena of acculturation, such as localisation

(as a counterforce to globalisation), asymmetric power relations, and cultural continuity.

It is argued that all such epiphenomena have as their driving force a single impetus: the

desire by migrants and others to mitigate the debilitating and disturbing impact of

uncertainty.

Uncertainty and its discontents. Uncertainty has long been recognised by all

branches of the social sciences as undesirable and undesired. For instance, Hofstede

(1980) recognised uncertainty avoidance as one of his cultural dimensions, and in the

field of cross-cultural psychology, Hermans (1999) linked the discontents of uncertainty

to globalisation:

Globalization is easily understood as contrary to living one's “authentic life” in

peace, partly because authenticity and pureness, however conceptualized, is better

suited to a homogeneous, stable, localized, and predictable society than an

increasingly heterogeneous, changing, translocal, and unpredictable global world.

(p. 1118).

He went on to argue that uncertainty and the strategies mitigating it must be

understood in a dialogical framework; and that “uncertainty is not primarily in a

culture's core but in its contact zones” (pp. 1119), where migrants meet their hosts. In

his later studies, Hermans (2001a; Hermans & Dimaggio, 2007) saw the progression

towards hybrid identity in the following terms: globalization leads to identity

hybridisation, which leads to uncertainty, which leads to anxiety. The logic of this

progression is found in dialogical self theory. In dialogical terms, the hybrid cultural

83

identity is characterised by an unprecedented density and heterogeneity of I-positions,

and the self is subjected to larger “position leaps” than ever before.

To recap, uncertainty is an unavoidable outcome of globalisation, particularly for

migrants, who themselves are a product of globalisation. But uncertainty alone does not

create anxiety; for that to happen there must also exist the possibility of undesired

outcomes among life’s many uncertainties. This is where power relationships enter the

argument, for it is the asymmetry of power as between migrant and host country that

diminishes the control that migrants have over their lives, and hence the likelihood of

undesired outcomes. Therein lies anxiety.

Power relationships. Disparities of power colour the acculturation process

(Hermans (2001). Migrants find themselves in a strange and unfamiliar environment

where even the simplest exchanges can conceal pitfalls as words are misunderstood or

unintended offence given. Hermans (2001) depicted the problem in dialogical terms by

characterising cultures

...as collective voices that function as social positions in the self. Such voices are

expressions of embodied and historically situated selves that are constantly

involved in dialogical relationships with other voices. At the same time these

voices are constantly subjected to differences in power. (p. 272, italics added)

Identity formation takes place by means of an internal dialogue between personal

and collective voices, in which external power relations play a role (Hermans &

Hermans-Konopka, 2010). Psychological and cultural processes are moreover

intimately connected, such that cultural diversity is reflected in the self.

Subsequent researchers have expanded on this theme. Kinnvall (2004) observed

that migration “is often characterized by a sense of powerlessness and dependence as

insecurity is increasing among many migrants. This is frequently mixed with an acute

anxiety about their new circumstances and strong feelings of homelessness” (p. 747).

84

Bhatia (2002) and Chaudhary & Sriram (2001) drew attention to the role of race and the

history of colonialism that colours much North–South migration. Raggatt (2007)

identified power relationships as central to his taxonomy of positioning in the dialogical

self. Lastly, much of the work of O'Sullivan-Lago (2011a; 2011b; O’Sullivan-Lago &

de Abreu, 2008; 2010; O’Sullivan-Lago, de Abreu & Burgess, 2008), which concerns

immigrants to Ireland, is conducted against a backdrop of power asymmetry,

exacerbated by the inclusion of asylum-seekers in her sample (see also later section on

cultural continuity).

Power relationships are of importance because of their function—or rather, lack of

one—in traditional acculturation models. Models such as that of Berry (1980) are

limited largely because they do not recognise and formally incorporate the reality of

power relationships. According to Bhatia and Ram (2001), such models implicitly

assume that psychological processes are the same for all migrants, weak and strong.

Migrants, it is supposed, are free to choose a preferred acculturation strategy,

integration being touted as the one with the best outcome in terms of mental well-being.

This, argue Bhatia and Ram, simply misrepresents the reality: asymmetries of power

and status can often rule out integration, and sometimes even assimilation, leading some

migrants to choose separation and leaving others to be marginalised.

In dialogical terms, migration is often less a “blissful marriage of … cultures” and

more a series of “contested, negotiated and painful, rupturing experiences associated

with living between cultures.” (Hermans, 2001; p. 271). When people migrate, they

naturally take with them an existing heterogeneous set of internal I-positions, developed

in their home country in dialogue with its culturally specific external I-positions; which

then confront an equally heterogeneous, but thoroughly unfamiliar, multiplicity of

external positions I-positions from the adoptive culture. In Hermans’ (2003) words:

85

Voices from the culture of origin do not simply disappear when people are

involved in an acculturation process. Instead, the older or deeper voices are often

established parts of the self, and they are challenged, evoked, repressed, or simply

ignored (p. 95).

Strategies for uncertainty mitigation. According to Hermans and Dimaggio

(2007), in an effort to calm the anxiety arising from the turbulence and unpredictability

of globalisation, people may resort to “localising” and “homesteading” (best seen as a

type of localisation) as a counterforce to the globalising forces they see all around. The

authors couch their explanation in terms of the dialogical self, which, they argue, is the

appropriate theoretical framework for the task. They grant that dialogical self theory is

not the only theory that claims to explain localisation: the others are social identity

theory, in which membership creates self-categorization to favour the in-group; optimal

distinctiveness theory, in which people seek a balance between desire to fit in or stand

out; and intersectionality theory, in which race, class, and gender are seen as

intersecting and interlocking. But in their view, none of these alternative theories

adequately explains the epiphenomenon of localisation.

Hermans and Hermans-Konopka (2010) identify “Five strategies to cope with the

heightened level of uncertainty” (pp. 44–47). The first is to reduce the number of a

person’s I-positions and their heterogeneity—to simplify life and so restore tranquillity.

The second strategy is to let a single, powerful I-position take the lead; leaders may rely

on this in their followers either for good (when they termed “inspirational”), or ill (when

they are termed “populists”). The third strategy is to delineate more firmly the boundary

between in-group and out-group (“We are not like them”), whereby multiple internal I-

positions effectively become external and so can be subordinated and subdued. The

fourth strategy is to augment the number of I-positions by adding others that provide

solace—a strategy not without risk as the same turbulent I-positions remain in place.

86

Lastly, the fifth strategy is what Hermans terms “going into uncertainty”, that is, to

confront and wrestle with uncertainty dialogically, as one would in constructive debate.

Significantly, all five strategies draw on dialogical self theory, but in different

ways. Strategies one through four are dialogical only in the sense that they intentionally

manipulate the number and type of I-positions. The fifth strategy is exceptional in that

the intention lies in the dialogue, not the I-positions that may or may not result. Any

change in I-positions is therefore unintentional in the sense of being unpredicted; what

is intentional is the dialogue that causes them. The I-position changes are necessarily

unpredictable, emerging perhaps serendipitously, for if not they would have been

managed already by one or other of the other strategies. Therein lies the real power of

dialogical thinking.

Kinnvall (2004; 2010), like Hermans, argues that globalisation diminishes

people’s sense of security. Majority and minority communities in Europe have reacted

by searching for one stable and secure identity, which she terms “the securitization of

subjectivity” (Kinnvall, 2004; p. 763). One strategy for achieving this aim—in essence,

localisation—is what Kinnvall calls “homesteading”, a term coined by Kronsell (2002)

for the different physical circumstances of an academic establishment, but co-opted by

Kinnvall to describe “making and shaping a political space for oneself in order to go

beyond and surpass the life of contradictions and anxieties that characterizes

homelessness” (Kinnvall, 2010, p. 601). Another, topical, example (not Kinnvall’s) is

provided by the independence movements of such places as Catalonia and Scotland,

which can be interpreted as localising resistance to the globalising impact of the Spanish

and British states. And though it seems paradoxical that both regions should so strongly

support membership of the European Union, the paradox is resolved when if Brussels is

perceived as a more distant, and therefore acceptable, alternative to the stifling embrace

87

of the mother country, providing enough political certainty to mitigate the uncertainty

and insecurity of smallness.

4.3 Cultural Continuity

Thus far, the discussion has centred on the mechanism by which globalisation

generates anxiety, and on the means of mitigating that anxiety. This section

demonstrates how this process commonly manifests itself in practice: the search for

cultural continuity, that is, the efforts that migrants (and more broadly anyone

experiencing acculturative stress) make to preserve their mother culture, sometimes

with success, sometimes not.

Fear of fragmentation. Culture metaphorically cradles the self in the sense that it

creates the “others” to which the self must relate dialogically, and so generates many of

the external I-positions that constitute the dialogical self (Salgado and Hermans, 2005).

And in so doing, it shapes and moulds the self in its image—or a reflection of it. But

attractive as this metaphor is, it does not of itself explain what happens to the self when

culture changes. For whatever else takes place, migrants (except in exceptional and

possibly pathological cases) do normally preserve their sense of personal continuity. It

seems, then, that individual and cultural continuity are intimately entwined in the sense

that the one depends very much on the other. It is for example impossible to imagine

cultural continuity without individual continuity (or indeed any kind of continuity for

that matter). Conversely, it may not be impossible, but it is certainly problematic, to

imagine individual continuity without cultural continuity.

This theme was ably explored by Chandler and Lalonde (1998; 2004; 2008) in

their research into First Nations of Canada. Like many indigenous communities, the

First Nations are migrants of a kind—not geographical but temporal: they stayed put

while the world changed around them. But the acculturation problems they faced were

88

nonetheless real, and the worse for being involuntary and infected with the power

dynamics of racism. In an extensive survey of First Nations communities, Chandler and

Lalonde (2008) found that those “that preserve heritage culture, and control destinies,

are dramatically more successful in insulating their youth against the risks of suicide”

(p. 6). The reason is that the adolescents of communities, already at an age of precarious

selfhood, see no future for themselves, literally, unless sustained by the visible and

emotional signifiers of cultural continuity. Chandler and Lalonde (2004) state:

If, owing to some train of personal or collective mishaps, single individuals or

whole communities lose track of themselves in time, and so suffer some

disconnect with their past or future, then old responsibilities and new promissory

notes fly out the same window, and life, we argue, becomes cheap. (p. 3)

It remains to explain this phenomenon in dialogical terms.

Salgado and Hermans (2005) argued that the self exists only by virtue of its

context. Rejecting both Thomist monism and Cartesian dualism, they see the self as an

emergent property of dialogism: “the I emerges by reference with an Other” (p.10). In

this way they reconcile the notion that the self must change as its dialogical others

change. The implication is that if one’s culture ceases to exist, then the external I-

positions that it sustains likewise are extinguished, which goes to the very core of

selfhood.

An alternative, but still consistent, explanation is afforded by the phenomenon of

fragmentation. Fragmentation is one form of dissociation, the others being omission,

subduing and distortion (Hermans & Hermans-Jansen, 1995). Unlike the postmodern

self, the dialogical self does not normally give rise to fragmentation; instead, it bridges

the gap between the modern ideas of discontinuity and continuity (Hermans (2001a).

Voices from different I-positions may disagree, but they are normally in dialogue, that

is, the oppositional I-positions acknowledge the other’s existence and may compromise.

89

Where the strains are too great, however, fragmentation may occur, such as that

experienced by Chaudhary (Chaudhary & Sriram, 2001), but it falls far short of the

pathological:

Whatever I do or say… either exotic or weird… I resist, sometimes consciously,

sometimes unknowingly… but the foreign culture is affecting my psyche…

nothing I do is irrelevant, it is either a reaction or an act of conformity…

sometimes I go out of my way not to resist… I want to be open… beyond

culture… (p. 383)

In summary, cultural and personal identity are inseparable but it is nevertheless

possible for the self to persist throughout the acculturation process and still retain its

sense of continuity. It does so by means of dialogue with multiple cultures,. Strategies

for doing so are described next.

Strategies for engaging alterity. In a program of research spanning two decades,

Bhatia and colleagues explored the acculturation of mainly South Asian migrants to the

United States. What they found resembled less a linear transition from monoculture to

one of Berry’s four acculturated states, than a dialogically mediated struggle.

Bhatia & Ram (2001; 2007) found that race, gender, imperialism and power

relations coloured and shaped cultural choices. Migrant women in particular, including

second-generation daughters, were caught in a dynamic of push-and-pull dialogical

negotiation, alternately tugged this way and that by their separate cultural identities.

According to Bhatia & Ram (2001):

These negotiations are multi-layered and complex because the voices of parents,

peers, language, siblings, homeland and American society are all represented in

the dialogical self. (p. 305)

In an investigation of how diasporic women negotiate multiple and conflicting

cultural identities, Bhatia and Ram (2004; 2007) found that they adopted three

90

dialogical strategies: polyphonization, expropriation, and ventriloquation. By

‘polyphonization’ is meant a strategy whereby selves oppose each other’s voices in a

complex dialogue that despite its contradictions does not fragment; by ‘expropriation’ is

meant a monologue whereby social institutions that demands total obedience to social

norms and institutions attempt to suppress dissenting voices, usually without success;

and by ‘ventriloquation’ is meant the phenomenon of one voice infusing the other voice

to express its own voice or I position. These are all dialogical strategies that attempt to

mitigate (whether in a healthy way or not) acculturative stress; and consequently they

demand a dialogical theory of the self in order to be understood.

Bhatia and Ram (2009) describe the experience of South Asian migrants to the

United States during the terrorist attack on the Twin Towers. Having taken for granted

that they were “integrated”, they found to their shock that they were not; overnight they

came to be treated as the “Other” for no other reason than they looked Arab, at least in

some Americans’ eyes. Bhatia (2002; 2003; 2014) echoes this theme by describing

Edward Said’s struggles with forming a coherent identity, and describing the cultural

constraints places on migrants in an occasionally racist host culture: “Will we ever

belong?” The implication is that “an immigrant’s journey through acculturation is not

straightforward, direct, self-evident or ever complete” (p. 147).

Although integration and bicultural competency may be worthy goals to

achieve… for most people living in contemporary diasporas, their negotiation

with multiple cultural sites is fluid, dynamic, interminable and often unstable. (p.

148)

Homesteading is another dialogical strategy to assuage the existential insecurity

engendered by encountering alterity. According to Hermans and Hermans-Konopka

(2010), “A primary way of responding to these experiences is to seek reaffirmation by

drawing closer to any localized group that is seen as capable of reducing uncertainty and

91

insecurity” (pp. 41–42). Bhatia (2007) explains how South Asians migrants to the

United States have done so: by re-creating India in the home, through media

consumption (mainly movies and music), membership of social clubs etc.

Homesteading, like localisation, is an antidote to the homogenising impact of

globalisation in that it sustains external I-positions reflecting the home culture.

Yet another strategy is to develop “adaptive” I-positions, by which is meant ones

that are conducive to functioning in the host culture. Aveling and Gillespie (2008)

identify three distinct I-positions among adolescents of Turkish origin in the UK, and

show that they are adaptive to the fractured and asymmetrical social contexts in which

they are embedded. As an example, the researchers show that one such I-position can

undermine these constraining racial chauvinisms, allowing them to build cultural

bridges, as, for example, in the diversity of their friendship groups. Thus while

reflecting persistent and unresolved dialogical struggles resulting from the

asymmetries of power which structure their socio-cultural context, this movement

between positions is nonetheless adaptive. (p. 21)

Lastly, Gillespie, Kadianaki, and O’Sullivan-Lago (2012) and O’Sullivan-Lago

(2008; 2010a) draw on the theory of semiotics for ameliorative strategies. Semiotic

regulation refers to the way in which meanings and actions regulate the thoughts and

actions of self and other:

Encountering alterity cannot be reduced to contact in the sense of co-presence in

the same geographical space. For a genuine encounter with alterity to occur, the

perspective of the other has to permeate the self and to some extent, transform the

self.… In the same way that living tissue expunges alien objects, semantic barriers

operate within the semantic realm to expunge discomforting ideas; just as the

body regulates what enters it, the self regulates what enters the semiotic realm. (p.

706)

92

In summary, what these strategies have in common is the attempt to mould the

migrants’ mental and physical environment to accommodate the demands placed upon

them by the imposition of otherness, and this way to create a familiar and hence

restorative mental and physical space.

Strategies for identity maintenance. O’Sullivan-Lago and her collaborators

(O’Sullivan-Lago, 2011a; O’Sullivan-Lago & de Abreu, 2008; 2010a; 2010b) have for

many years researched the acculturative experiences of immigrants to Ireland. Ireland

has traditionally been a land of emigrants, so to start receiving immigrants was therefore

a novel, and to many Irish people, disturbing experience. O’Sullivan-Lago and

collaborators saw this therefore as fertile ground for employing dialogical self theory to

explore how both immigrants and Irish adapted to the situation.

O'Sullivan-Lago de Abreu and Burgess (2008) describe a strategy whereby both

immigrants and Irish adapted an I-position that permitted both groups to accept the

dignity of the other. As a basis for their research they drew on Chandler and Lalonde’s

(1998) research into youth suicide in Canadian First Nations communities, which found

enormously elevated rates in certain communities, and linked these to a lack of cultural

continuity. According to O'Sullivan-Lago et al. (2008), the relevance to Ireland is two-

staged: first, “an individual’s decision to end their own life might be made far easier if

they were suffering from broken personal continuity because they would lose their

personal meaningful connection to who they will become in the future” (p. 350); and

second, “we can infer from Chandler’s theory that… cultural continuity has a profound

impact at the individual level” (p. 355). Or to trace the causal link from its genesis,

cultural continuity affects individual continuity, which in turn affects suicidal

behaviour. Completing the syllogism, immigration has caused cultural discontinuity,

which has caused, not suicide, but mental ill-health—or would do so but for the

93

ameliorative I-position adopted by both sides: “I as a human being”, the function of

which is to maintain cultural, and hence individual, continuity in the face of

demonstrated stresses on cultural continuity:

The strategy of evaluating all as equal due to the possession of the ‘I as a human

being’ identity allowed each group to attempt to restore their broken continuity.

All are, and always will be, human beings despite divisions such as culture and

ethnicity. (p. 361)

In an elaboration of this theme, O’Sullivan-Lago (2011a) explored “how

solidarity might be formed between the groups now resident in the multicultural space

and examine[d] what identity spaces are available that will bridge the gap between self

and other” (p. 3.18). Besides the shared-humanity perspective identified by O'Sullivan-

Lago et al. (2008), several other representations of intercultural solidarity were found,

concerning religion, collective history, and EU membership. In an analysis of Roma

(otherwise termed Romanies or Gypsies), O’Sullivan-Lago (2011b) explored the same

theme from a negative perspective, that is, one where the “I as a human being” position

is not established. By constructing the Romany other as less than human… the

relationship is no longer dialogical but monological: the possibility of a relationship, of

dialogue, is removed” (p. 68). Though not achieved in this case, the shared-humanity I-

position is seen as supremely important for dialogue and thereby the maintenance of

culture.

O'Sullivan-Lago and de Abreu (2010b) approached cultural continuity from a

different angle, by asking not about coping strategies as such (as O'Sullivan-Lago de

Abreu and Burgess, 2008) but about their proposed identity strategies given the cultural

futures they foresee for themselves. Accepting the proposition that the dialogical self is

a “self-organizing system”, they asked “what happens in the Dialogical Self when

individuals experience cultural discontinuity” (p. 284). The answer depended on the

94

group in question. The native-born Irish represented Irish cultural ways as “normal”,

which carried the suggestion that migrants should assimilate. The voluntary immigrants

stressed their integration into Irish culture but at the same time emphasized the

maintenance of their own cultural identity positions—an “integration” strategy in

Berry’s terms. Lastly, the asylum seekers espoused a more or less assimilationist

position because they needed to stay in Ireland. What is interesting about these

strategies is not that they fit Berry’s taxonomy (which anyway covers the range of what

is possible) but that they are negotiated, malleable (the voluntary migrants might want

to repatriate, so keep their cultural options open), and above all constrained. All seek to

restore continuity, but find different ways to do so given their differing circumstances.

4.4 Research Aims and Questions

Aims. The preceding review of the research to date strongly suggests that

dialogical self theory has the potential to explain the nature of acculturation better than

previous theories. Supporting evidence is important, given that dialogical self theory is

a relatively new model of personality. Quantitative research is particularly needed, to

build on existing qualitative research. The present thesis seeks to investigate the

migration experience by utilising a dialogical-self perspective to add to this endeavour.

There are two broad aims. One is to test the theory of the dialogical self in a

natural context, namely the phenomenon of migration. The logic of the investigation is

as follows: the migration experience constitutes a transition in the lives of migrants,

which places them under acculturative stress; this cultural rupture, which broadly takes

the form of uncertainty, is then manifest in the repertoire of I-positions of migrants

when probed using the self-confrontation method. Verifying this analysis would buttress

dialogical self theory as both an explanatory and a diagnostic tool in a range of

applications.

95

The second aim of the thesis follows from the first. To recap, the first aim is to

use the theory of the dialogical self to establish whether or not migration is reflected in

the migrant’s repertoire of I-positions; the second is to determine the magnitude of any

such change.

Research questions. Five research questions (some being multi-part) are

investigated.

Migrant identity. The present thesis has presented theories of acculturation that

have implications for I-positions. In essence, a number of theories claim ethnic

identification as a causal factor in acculturation outcome. Research questions 1a and 1b

explore this proposition.

Research question 1a: Biculturalism. How is biculturalism reflected in I-

positions?

Research question 1b: Ethnic identification. How is ethnic identification reflected

in I-positions?

Stability and change in the dialogical self. The present thesis has shown that

people have multiple selves that coexist in time and space, and that this proposition is

widely, though not universally, accepted by contemporary theorists in the field. It is,

furthermore, a central and defining plank of the theory of the dialogical self. For these

reasons it is central to this thesis. The thesis moreover tests the existence of multiple

selves, for migration qualifies as exactly the type of life change that might propagate a

corresponding I-position. At the same time, however, it is not expected that all migrants

will exhibit changes in I-position, for even theories of multiple identity acknowledge

that change must take place against a backdrop of stability, without which it is

impossible to talk about the individual as an enduring construct. Research questions 2a

96

and 2b pertain to migrants who claim to have two cultural identities, and were explicitly

selected for the study sample.

Research question 2a: Stability of the dialogical self. Are country-of-origin and

Australian I-positions similar, and if so, how?

Research question 2b: Changes in the dialogical self. Do country-of-origin and

Australian I-positions differ, and if so, how?

Composition of I-positions. The present thesis has presented theories of identity

some of which suggest possible parallels with aspects of the dialogical self. Research

questions 3a, 3b and 3c explore aspects of this and similar propositions.

Research question 3a: Evidence of multivalence of the dialogical self. Can

individuals simultaneously possess I-positions that differ significantly in their major

themes.

Research question 3b: Evidence of adjustment to migration. If so, do individuals’

I-positions correspond to those for country of origin and Australia? In what way?

Research question 3c: Motives for migration. What motivates successful migrants:

the desire for self-enhancement or union-with-the-other?

Psychological state of migrants. The surveyed literature touched on migrants’

psychological response to acculturation in terms of culture shock and acculturation

stress. Although marked distress was not expected (the study sample had lived in

Australia long enough to have passed through any period of culture shock, or cultural

“honeymoon”), it was thought that psychological responses to migration might reveal

the degree of adaptation to life in Australia. Research question 4 is grounded in theories

that invoke a cultural adaptation as a response to the stresses imposed by trying to

survive and prosper in an alien culture, particularly the “ABC model” of acculturation.

97

It reveals migrants’ attempts to adapt to circumstances by exploring the relationship

between their I-positions and their psychological distress and psychological well-being.

Research question 4: Psychological distress and well-being. How do migrants,

when classified by their I-positions pertaining to country of origin and Australia, vary in

psychological distress and well-being?

Case studies. Research question 5 seeks to corroborate the answers to research

questions 1 and 2 by taking an idiographic approach.

Research question 5: Idiographic confirmation. Are differences between

individuals classified by their I-positions reflected in the life stories of specific case

studies?

Plan of the empirical work. The present research took the form of three parallel

studies. Study 1 was exploratory in nature. It investigated whether or not there might

exist naturally forming groups among migrants. Its specific aim was to investigate if

migration brings into being separate I-positions relating variously to the migrant’s

country of origin and to Australia; and if so, to determine if the nature of those I-

positions were related to the demographic characteristics, psychological self-reports and

self-confrontation indices of the migrants possessing them. Study 1 therefore addressed

the first stated aim of the research, namely to establish whether or not migration were

reflected in the migrant’s repertoire of I-positions.

The aim of Study 2 was to elaborate and thereby corroborate Study 1. Unlike

Study 1, which used cluster analysis to group individuals, Study 2 classified them by

comparing their valuations in their country of origin and Australia. Despite their

differences in analytical approach, both studies investigated similar patterns of stability

and change in I-positions. However, it was in the context of Study 2 that hypotheses

concerning stability of the self were developed and formally tested. Study 2 therefore

98

addressed the second stated aim of the research, namely to determine the magnitude of

any change observed under Study 1.

Study 3 adopted an idiographic approach to complement the nomothetic approach

of Study 1. Four case studies will be discussed, one of each migrant type: Progressive,

Low-stable High-stable, and Regressive. The aim is to exemplify and corroborate the

nomothetic principles identified in Study 1, and thereby to furnish a more complete

understanding of the individual.

All three studies employ data gathered from the same sample of 38 adult migrants

to Australia. To maximise comparability they were selected according to five criteria:

they perceived themselves as having two separate cultural identities, one for their

country of origin and the other for Australia; they were all born overseas and migrated

to Australia after the age of 12; they were voluntary migrants (as opposed to refugees);

they had lived in Australia for at least five years; and they had a good knowledge of

English. Chapter 5 documents the rationale for these criteria and details the

characteristics of the sample.

99

CHAPTER 5: Research Methodology

This chapter describes the research method adopted in the thesis. First, the

characteristics of the sample are described. Next, details are given for the data gathered

under the study; these were of two main kinds: self-report measures and self-

confrontation indices. After that, the self-confrontation method (SCM) is described, this

being the method by which the self-confrontation indices were derived. The chapter

concludes with a description of the procedure for gathering data, and the procedure for

computing the self-confrontation indices from the participants’ responses.

5.1 Participants

Sample selection. Data for the present thesis came from a sample of 38 migrants

(14 men and 24 women), who were recruited using a variety of methods. Initially it was

thought that the sample could be recruited on campus, and to this end advertisements

were placed on notice boards posted around the university campus (see Appendix A). In

the event only two student participants were obtained in this way. A “snowball”

sampling method (Valentine, 1997) was adopted instead, which had the advantage that

the respondents were drawn from the broader community. This was deemed appropriate

as the major research goal was to study a range of migrants’ valuations to see whether

patterns existed across them. First, a few participants were recruited from interested

applicants who contacted the researcher after she was interviewed about the study by a

local community radio station. These then suggested others, who in some cases

suggested still others.

To maximise comparability of participants, they were selected according to five

criteria:

100

Criterion 1: On questioning, the participants reported that they perceived

themselves as having two separate cultural identities—that is, they thought of

themselves as having achieved bi-cultural identification, one identity belonging to their

country of origin and the other identity belonging to Australia.

Criterion 2: The participants were all born overseas and had experienced the same

life changing event: migration to Australia after the age of 12. As such, they are

categorised as first-generation migrants (Rumbaut, 2004) since those who arrive in their

adolescent years have similar experiences and adaptive outcome as adult migrants.

Criterion 3: The participants were voluntary migrants (as opposed to refugees).

Reasons for migrating have been studied using the concepts of “pull/push motivations”

(Berry, 1990), where push motives characterise the involuntary or forced migration of

refugees. This is generally negative and mostly reactive, whereas pull motives

characterise voluntary migration of migrants, which is generally positive and mostly

proactive, seeking a better future (Richmond, 1993).

Criterion 4: The participants had lived in Australia for at least five years so as to

allow for sufficient time to fully experience migration and also to allow for enough time

for the adaptation process to take place (Taft, 1986).

Criterion 5: The participants had a good knowledge of English to avoid translation

difficulties. Also, following La Fromboise et al.’s (1994) focus on the concept of

bicultural competence, English language proficiency underlies the ability to function

effectively in mainstream Australia. A good language competence is known to facilitate

the adaptation process and be an important indicator of acculturation level (Berry,

1990).

Sample characteristics. Table 5.1 sets out the salient characteristics of the

sample.

101

Table 5.1 Characteristics of the sample, by gender

Gender Total

Male Female N %

Marital status Single 4 6 10 26.3%

Married 7 15 22 57.9%

Divorced 3 3 6 15.8%

Education Before Year 10 2 2 5.3%

Year 10 1 1 2.6%

Year 12 1 3 4 10.5%

Undergraduate 5 4 9 23.7%

Postgraduate 8 14 22 57.9%

Language at home English only 7 8 15 39.5%

Bilingual 6 6 15.8%

Mother tongue only 7 10 17 44.7%

Ethnic identification Belongs to ethnic group 1 7 8 21.1%

No ethnic belonging 13 17 30 78.9%

Fourteen participants were male and 24 female. Most were married and the rest

either single or in a few cases divorced. Most were highly educated—possibly as a

consequence of using a snowball sampling method. Just under 40% spoke only English

at home, slightly more spoke only their mother tongue at home, and a few were

bilingual. About one in five identified with their ethnic group, and the rest claimed no

particular ethnic affiliation. At first sight this is unexpected, as a criterion for selecting

participants was that they perceived themselves as having two cultural identities, one for

102

country of origin, the other for Australia. One would not, therefore, expect them to

express “no ethnic belonging”. A plausible explanation is that they perceived their

ethnic identity as private rather than public. Thus they have two private cultural

identities while at the same time not identifying publicly with their ethnic group.

Table 5.2 presents means of continuous variables by gender. No significant

differences were found for years in Australia, t(36) = –.46, p = .65 (two-tailed), or for

age, t(36) = –.33, p = .74.

Table 5.2 Means and standard deviations for Years in Australia, Age and Individualism, by

gender

Gender

Male Female All

Years in Australia Mean 18.7 20.4 19.8

Standard deviation 10.7 10.6 10.5

Age Mean 41.1 42.2 41.8

Standard deviation 10.4 9.6 9.8

Individualism Mean 38.4 43.1 41.4

Standard deviation 12.5 24.5 20.8

Hofstede’s (1980) concept of individualism was discussed in Chapter 2. Table 5.3

ranks the sample by individualism score of their country of origin. About three-quarters

of the sample was drawn from countries such China and Colombia that are low on

individualism (arbitrarily defined as 50 on Hofstede’s scale) and consequently high on

collectivism. These are cultures that stress collective identity, in-group loyalties,

emotional interdependence, sharing, duties and obligations, a need for stable and

predetermined friendships, and subordination of individual interests.

103

Table 5.3 Number of participants, by Hofstede’s individualism index

Country Individualism index n

United States 91 1

Australia (listed for comparison only) 90 0

United Kingdom 89 2

Italy 76 2

South Africa 65 1

Poland 60 1

Spain 51 2

India 48 4

Argentina 46 1

Japan 46 1

Iran 41 1

Egypt 38 1

Lebanon 38 1

Uruguay 36 2

Greece 35 4

Philippines 32 2

Mexico 30 3

Ethiopia 27 1

Kenya (Seychelles) 27 1

China 20 3

El Salvador 19 3

Colombia 13 1

Total

38

Conversely Australia, with an individualism score of 90, ranks second out of 53

countries on individualism (Hofstede, 1980), implying an emphasis on individual

autonomy, emotional independence, assertiveness, self-expression and individual

achievement. For most of the sample, therefore, the cultural contrast between country of

104

origin and Australia is substantial. It follows that most have had to make a big cultural

adjustment if it is assumed that they are representative of their birth cultures (which is

not necessarily the case). In the present study, however, men and women did not differ

significantly in the individualism score of their countries of origin, t(36) = –.67, p = .51

(see Table 5.2 for means).

Sample representativeness. In requiring that participants possess “two separate

cultural identities”, the sample made no claim to be statistically representative of

migrants in general, and this is evident in the sample characteristics. Most participants

were well educated (over 80% university educated) and all had a good command of

English (40% spoke it in the home, including three participants from English-speaking

countries). However, the intention was not that the sample be representative but that it

include migrants from a range of backgrounds. Tellingly, they were diverse in their

responses, some finding the migrant experience positive, some not. This is appropriate

to test the study’s research questions.

5.2 Self-report Measures of Psychological Distress and Well-being

Each participant completed a questionnaire that contained self-report measures of

psychological distress and well-being (Appendix A). Descriptive statistics (including

reliability coefficients) relating to the psychological scales are presented in Appendix C.

Psychological Adjustment Measures. Three measures were used. All were self-

reports and are discussed below.

Social Avoidance and Distress Scale. The Social Avoidance and Distress Scale

(SAD; Watson & Friend, 1969) was employed as a cognitive and behavioural

measurement of social adjustment. The SAD contains 28 items, 14 of which assess

social avoidance and 14 of which measure social anxiety. It uses a True/False response

105

format, evenly divided between true and false items. Scale scores on the true/false scale

range from 0 (lowest avoidance and distress) to 28 (highest avoidance and distress).

Social avoidance and distress as measured by the SAD refers to an individual’s

tendency to avoid social situations and to feel anxious while in such situations. The

authors distinguished between avoidance and the failure to approach, arguing that the

opposite of avoidance is not social approach, but simply lack of avoidance. Further,

they deliberately included only items that assessed subjective distress and behavioural

avoidance and excluded items that dealt with physiological indices of anxiety and

impaired behavioural performance (Leary, 1991). Patterson & Strauss (1972) used

factor analysis to verify the subscale structure for avoidance and distress; however, the

subscales have not been validated in this way, and most research using this scale has

used the full-scale version (Leary, 1991). Watson and Friend (1969) reported three

experimental and one correlational study of the SAD. The data indicated that people

defined as high anxious (those with scores of 11 and above) on the SAD were less

interested in participating in group activities and talked less when in an interpersonal

situation than the low anxious (those with scores below 3). Watson and Friend (1969)

reported a mean for university students of 9.1 (SD = 8.0), adequate internal consistency

(Cronbach’s alpha = .77), test-retest reliability (r = .68 to .79 over a 1-month period),

and construct validity (Watson & Friend, 1969). More recently, Hofmann, Dibartolo,

Holaway, and Heimberg (2004), in a study conducted to check the psychometrics of the

SAD on a sample of 199 college students, reported a Cronbach’s alpha of .93.

Cronbach’s alpha for the present thesis was .93.

Fear of Negative Evaluation Scale. The Fear of Negative Evaluation Scale (FNE;

Watson & Friend, 1969) was employed as a cognitive measurement of sociocultural

adaptation. The FNE is a 30-item measure of the degree to which people expect to be

106

negatively evaluated by others, to look foolish, and to make a bad impression. Watson

and Friend (1969) defined fear of negative evaluation as “apprehension about others’

evaluations, distress over their negative evaluations, avoidance of evaluative situations,

and the expectation that others would evaluate oneself negatively”. In elaboration, they

further stated that “fear of loss of social approval would be identical to FNE” (p. 449),

and that the opposite of high FNE is the lack of apprehension about other’s evaluations,

but not necessarily a desire or need to be evaluated positively.

The FNE’s items are scored true or false, split roughly equally between positively

and negatively scored items. True responses are scored as 1 point and scores range from

0 to 30, with higher scores indicating greater fear of negative evaluation. The FNE

norms (and also the SAD norms) were derived from a pilot study by Watson and Friend

based on an original sample of 205 college undergraduates that revealed a mean of 15.5

(SD = 8.6), with total scores of “18 or higher” and “9 or lower” on the FNE

corresponding to cut-off scores for the top and bottom 25%, respectively. They reported

that participants high on FNE became nervous in evaluative conditions, and were more

likely to expect disapproval from a leader and in addition more likely to seek approval

from others in order to avoid anticipated disapproval than low FNE participants. The

authors reported high internal consistency (Cronbach’s alphas = .94 to .96), strong test-

retest reliability (r = .78 to .94 over a 1-month period), and construct validity (Watson

& Friend, 1969).

Arkowitz, Lichtenstein, McGovern and Hines (1975), using a sample of 205 high

and low frequency daters in a college population, reported a mean FNE score of 7.85 for

high frequency daters, and 18.74 for low frequency daters. British FNE norms were

established by Stopa and Clark (1991) in two studies using a large sample of students.

107

Research on individual differences on the FNE has found that, compared with low

scorers, participants high in FNE avoid situations that could lead to potentially

threatening comparisons with others (Friend & Gilbert, 1973), are more affected by

negative evaluations about themselves (Smith & Saranson, 1975), worry more about

how they come across to other people (Leary, 1983), are more socially anxious and

more easily embarrassed (Heimberg, Hope, Rapee, & Bruch, 1988; Miller, 1995) are

more sensitive to criticism (Atlas, 1994), are more dependent (Overholser & Freheit,

1994), and are inclined to interpret others’ facial expressions as negative (Winton,

Clark, & Edelmann, 1995) than low FNE participants. The FNE has been validated

through significant positive correlations with selected personality measures (e.g., social

approval, locus of control; Corcoran & Fischer, 2000) and a negative correlation with

self-acceptance (Durm & Glaze, 2001). Cronbach’s alpha for the present thesis was .96.

Beck Depression Inventory, second edition (BDI–II; Beck, Steer, & Brown,

1996). The BDI-II was employed as a measure of psychological adjustment. The BDI-II

is the 21-item Beck Depression Inventory (BDI; Beck & Steer, 1993), which assessed

the severity of depression in normal and psychiatric populations of both adolescents and

adults, updated in light of criteria from the Diagnostic and Statistical Manual of Mental

Disorders (DSM-IV; American Psychiatric Association, 1994) for major depressive

disorder. The BDI–II comprises 21 items relating to symptoms and attitudes considered

to reflect depth or severity of depressive symptomatology in individuals aged 13 years

and older (Beck, Steer, & Brown, 1996). The BDI–II measures the cognitive, affective,

somatic, and vegetative aspects of depressive experience. Participants indicate the

extent to which they have experienced the symptoms over the last two weeks.

Individual items are rated on an ordered (0 to 3) scale, with higher scores indicating

higher levels of depression. The maximum total score for all 21 items is 63. According

108

to the BDI-II manual, scores from 0 to 13 indicate minimal depression, 14 to 19 indicate

mild depression, 20 to 28 indicate moderate depression, and 29 to 63 indicate severe

depression. More typically, clinically depressed or maladaptively nonclinical

populations score in the 14 to 28 range (Beck et al., 1996).

The BDI-II has demonstrated high internal consistency (alpha coefficients were

.93 among college students, and .92 among adult psychiatric outpatients (Beck et al.,

1996). The authors also demonstrated adequate validity (e.g. content, construct,

factorial), and established diagnostic discrimination. In their psychometric evaluation of

the BDI-II, Dozois, Dobson and Ahnberg (1998) established corresponding BDI-II cut-

off scores for classifying groups of research participants (0 to 12, nondepressed; 13 to

19, dysphoric; 20 to 63, dysphoric-depressed) based on cut-offs developed by Kendall,

Hollon, Beck, Hammen, and Ingram (1987) for the BDI. Dozois et al. (1998) also

reported a high level of internal consistency (Cronbach’s alpha of .91). More recent

research by Steer, Rissmiller, and Beck (2000) found that the total BDI-II score was not

significantly correlated with gender, age, or ethnicity of the participants. Cronbach’s

alpha for the present thesis was .93.

Psychological Well-being Scale. The Psychological Well-being Scale (PWB;

Ryff, 1989a; Ryff & Essex, 1992; Ryff, Lee, Essex & Schmutte, 1994) was employed as

a measurement of psychological adjustment. Derived from the literature on lifespan

development, mental health, and personal growth (Ryff, 1989b), the PWB consist of 84

items rated by a response scale ranging from 1 (strongly disagree) to 6 (strongly agree).

The version of the PWB used here contains six subscales: environmental mastery,

positive relations with others, autonomy, personal growth, purpose in life, and self-

acceptance.

109

Each PWB subscale was operationalised with a 14-item scale of positively and

negatively phrased items, which were mixed and, where appropriate, reversed to

produce a single inventory. Each subscale has a range of 14 to 84, high scores indicating

higher self-ratings of the dimension assessed. These six dimensions are found in Table

5.4 and describe psychological well-being in terms of the presence of positive attributes

rather than the absence of mental illness.

110

Table 5.4 Dimensions of Ryff’s Psychological Well-being Scale (PWB)

Dimension Score Definition

Environmental

mastery

High Has a sense of mastery and competence in managing the

environment; controls a complex array of external activities; makes

effective use of surrounding opportunities; is able to choose or

create contexts suitable to personal needs and values.

Low Has difficulty managing everyday affairs; feels unable to change or

improve surrounding context; is unaware of surrounding

opportunities; lacks sense of control over external world.

Positive

relations

with others

High Has warm, satisfying, trusting relationship with others; is

concerned about the welfare of others; is capable of strong

empathy, affection, and intimacy; understands the give and take of

human relationships.

Low Has few close, trusting relationships with others; finds it difficult to

be warm, open, and concerned about others; is isolated and

frustrated in interpersonal relationships; is not willing to make

compromises to sustain important ties with others.

Autonomy High Is self-determining and independent; is able to resist social

pressures to think and act on certain ways; regulates behaviour

from within; evaluates self by personal standards.

Low Is concerned about the expectations and evaluations of others;

relies on judgments of others to make important decisions; conforms

to social pressures to think and act in certain ways.

Personal

growth

High Has a feeling of continued development; sees self as growing and

expanding; is open to new experiences; has sense of realising his or

her potential; sees improvement in self and behaviour over time; is

changing in ways that reflect more self-knowledge and effectiveness.

Low Has a sense of personal stagnation; lacks sense of improvement or

expansion over time; feels bored and uninterested with life; feels

unable to develop new attitudes or behaviours.

Purpose

in life

High Has goals in life and a sense of directedness; feels there is meaning

to present and past life; holds beliefs that give life purpose; has

aims and objectives for living.

Low Lacks a sense of meaning in life; has few goals or aims; lacks sense

of direction; does not see purpose of past life; has no outlook or

beliefs that give life meaning.

Self-

acceptance

High Possesses a positive attitude towards the self; acknowledges ad

accepts multiple aspects of self, including good and bad qualities,

feels positive about past life.

Low Feels dissatisfied with self; is disappointed with what has occurred

in past life, is troubled about certain personal qualities; wishes to

be different from what one is.

111

Ryff (1989a) demonstrated the convergent validity of the scale with positive and

significant correlations with five prior measures of positive functioning: Affect Balance

Scale (Bradburn, 1969), Life Satisfaction Index (Neugarten, Havighurst & Tobin 1961),

Self-Esteem Scale (Rosenberg, 1965), Locus of Control – Internal (Levenson, 1974),

and Revised Philadelphia Geriatric Center Morale Scale (Lawton, 1975). Coefficients

ranged from .25 to .73.

Discriminant validity was shown via negative and significant correlations with

three established prior measures of negative functioning: Locus of Control-Powerful

Others, and Chance Control (Levenson, 1974), and Zung Depression Scale (Zung,

1965). Coefficients ranged from –.30 to –.60.

Intercorrelations amongst the six dimensions were all positive and significant and

ranged from .32 to .76 (Ryff, 1989a). The three highest intercorrelations (between self-

acceptance and environmental mastery, r = –.76, between self-acceptance and purpose

in life, r = .72, and between purpose in life and personal growth, r = .72) suggest that

these dimensions might be linked at some underlying causal level. Ryff (1989a),

however, found that they showed differential age profiles, and hence represent different

facets of positive psychological well-being.

Ryff (1989a) reported a high level internal consistency for the original 20-item

version scale, with Cronbach’s alpha ranging from .86 for autonomy to .93 for the self-

acceptance. She also reported excellent test-retest reliability over a six-week period,

with reliability coefficients ranging from .81 for environmental mastery and personal

growth, to .88 for autonomy (Ryff, 1989a).

The PWB scales have shown replicable patterns of significant age and gender

differences (Ryff, 1989a; 1991; 1995). Ryff (1989a) asked young, middle-aged, and

old-aged adults to rate themselves on each of the dimensions of well-being. She found

112

that environmental mastery and autonomy, increased with age, particularly from young

adulthood to midlife; personal growth and purpose in life decreased, especially from

midlife to old age; and positive relations with others and self-acceptance did not change

significantly (Ryff, 1989, 1995; Ryff & Keyes, 1995; Ryff et al., 2003). In a metastudy

of gender differences, Ryff (1995) found that women consistently scored higher than

men on positive relations with others and personal growth; but she found no significant

differences between men and women on the other subscales (Keyes, Shmotkin, & Ryff,

2002; Ryff, 1989; Ryff & Keyes, 1995).

Other studies using the Ryff scales explored the effect of education, and socio-

economic status on psychological well-being (Clarke, Marshall, Ryff, & Wheaton,

2001; Marmot, Ryff, Bumpass, Shipley, & Marks, 1997; Ryff, 1989; Ryff & Keyes,

1995; Ryff, Keyes, & Hughes, 2003). In respect of education, psychological well-being

in general, and purpose in life and personal growth in particular, were associated with

higher levels of education (Keyes & Ryff, 1998, chap. 7; Marmot et al., 1997; Ryff et

al., 2003).

In a study investigating how a life experience such as community relocation

relates to mental health, Ryff and Essex (1992) used the 14-item, shortened version of

the original PWB (Ryff, 1989a). Results were encouraging for the robustness of the

PWB; internal consistency for the reduced scales ranged from .86 to .93, correlations

with the 20-item original scales ranged from .97 to .98, and test-retest reliability over

ranged from .91 to .88 (Ryff & Essex, 1992).

In a 1997 study, Schmutte and Ryff examined the common affective

underpinnings between measures of personality and well-being, with 2 samples of

midlife adults completing measures of personality (NEO Five-Factor Inventory; Costa

& McCrae, 1992) and the shortened (14-item scales) versions of the original Ryff’s

113

(1989a) PWB. They found consistent relationships between the Big Five personality

factors and the six PWB subscales. Internal consistency (alpha) coefficients for the 14-

item scales ranged from .82 to .90.

More recently, Diehl, Hastings and Stanton (2001) used the shortened (14-item

scales) versions of Ryff’s (1989a) PWB in their study to investigate the extent to which

persons’ self-representations are different for different social roles and contexts, across

the adult life span, and their associations with psychological well-being. They obtained

coefficients of internal consistency of .91, .89, .87, .88, .84, and .87 for self-acceptance,

environmental mastery, purpose in life, positive relations with others, personal growth,

and autonomy, respectively.

Ryff (1989a, b) argued that previous perspectives on operationalizing well-being

“have little theoretical grounding” (Ryff, 1989a, p.1069) and neglect important aspects

of positive functioning. To address this shortcoming Ryff drew on the theories of

positive mental health (Jahoda, 1968), personal growth (e.g., Maslow 1968; Rogers,

1961; Allport, 1961), life span developmental (e.g., Erikson, 1959). In her view, well-

being is not simple the attainment of happiness but is characterised by “the striving for

perfection that represents the realization of one’s true potential” (Ryff, 1995, p.100). In

this view, happiness is “the by-product of a life that is well-lived” (Ryff & Singer, 1998,

p.5), and is reflected in six ideals: autonomy, personal growth, self-acceptance, purpose

in life, environmental mastery, and positive relations with others. And on that basis,

Ryff (1989a) developed a multidimensional instrument to measure them: Ryff’s

Psychological Well-Being (PWB).

The Ryff scales have also been used cross-culturally. For instance, Ryff (1995)

used it to investigate the psychological well-being in middle-aged Americans and South

Koreans. Interestingly, findings were broadly consistent with known cultural

114

differences. For instance, age and gender profiles were similar, but the Americans had

higher scores on all of the Ryff scales; Americans were more likely to attribute positive

qualities to themselves than were South Koreans; gender differences were the same in

both cultures, with women rating themselves significantly higher than men on positive

relations with others and personal growth; South Koreans placed greater emphasis on

the well-being of other people (e.g., children) in defining their own well-being than did

the Americans.

These and other findings lend credence to the robustness and usefulness of the

Ryff scales, including in a cross-cultural context. Cronbach’s alphas in the present

thesis were .91 for environmental mastery, .92 for positive relations with others, .88 for

autonomy, .88 for personal growth, .87 for purpose in life, and .91 for self-acceptance.

5.3 Self-confrontation Method (SCM)

Overview and rationale. Valuations were accessed and measured by means of

the self-confrontation method (SCM), which is an investigative tool used to access and

explore the dialogical self (Hermans, 1987a, 1987b, 1988, 1989, 1992; Hermans &

Hermans-Jansen, 1995). It did so by accessing I-positions. According to Hermans, “the

I … continuously organizes and interprets experience …”; people adopt different I-

positions depending on their life circumstances. If those circumstances are cultural, then

cultural identity is a special case of an I-position.

In the SCM, each valuation is assigned four indices, one each for the self-

enhancement motive, the union-with-the-other motive, positive affect, and negative

affect—respectively termed the S, O, P and N indices (or simply S, O, P and N for

short). Collectively, they represent the “theme” of the valuation. Although valuations

may in principle exhibit any combination of indices, many (but not all) can be assigned

115

to one or other of six “major themes”, each denoted by a characteristic feeling (Hermans

& Hermans-Jansen, 1995, p.73). (The SCM is elaborated in Chapter 4.)

Two additional concepts, “general feeling” (Hermans, 1988) and “ideal feeling”

(Hermans & Oles, 1996), are now explained. “General feeling” measures how a

particular valuation impinges on the general experience of the person; “ideal feeling”

measures how a valuation impinges on the ideal experience of the person. The more

generalising power a valuation has, the more influential is its affective component in

colouring the way the person generally feels, and the more idealising power a valuation

has, the more it says about the way the person would like things to be.

The SCM is administered in three stages: (a) valuation elicitation, (b) affective

rating, and (c) evaluation and integration (Lyddon, Yowell & Hermans, 2006). As an

investigative tool, the SCM is particularly suited to a study of this kind. First, it was

expressly designed to access I-positions. Second, it allows access to cultural identities

both before and after migration, so obviating a costly and difficult longitudinal

experimental design. Third, it avoids the bias of self-report questionnaires because its

purpose is not apparent to the participant (Berry, 1990, p.482). The present thesis used

an adapted version of the SCM. The procedure had three steps: (a) identification of the

cultural positions of the self that are current in the person’s life; (b) formulation of a

valuation system pertaining to each cultural position; and (c) assessment of the affective

implications of the valuations of the two positions.

Valuation elicitation. In the SCM, valuations are elicited by using a set of

standard, open-ended questions (see Table 5.5).

116

Table 5.5 Questions of the Self-confrontation method

Set 1: The Past

These questions are intended to guide you in reviewing one or more aspects of your life

which may have been of great importance to you.

Has there been anything of major significance in your past life which still

continues to exert a strong influence on you? Example: “After high school I never

managed to build up a group of friends; I felt left out and stuck between two

worlds.”

Was there in the past any person or persons, experience or circumstance which

greatly influenced your life and still appreciably affects your present existence?

Example: “My parents have always treated us as equals. They always tried to keep

me stimulated. I could always talk very well with my parents”.

Set 2: The Present

This set again consists of two questions which will lead you, after a certain amount of

reflection, to formulate a response:

Is there anything in your present existence which is of major importance to you or

exerts a significant influence on you? Example: “Bodily complaints: Every time I

find that I have a problem, my body signals it in one way or another”.

Is there in your present existence any person or persons or circumstance which

exerts a significant influence on you? Example: “Wendy is a good friend”.

Set 3: The Future

The following questions will again guide you to a response:

Do you foresee anything that will be of great importance for, or exert a major

influence on your future life? Example: “I want to pursue a combination of

studying and volunteer work”.

Do you feel that a certain person, persons or circumstance will exert a significant

influence on your future life? Example: “John and I want children in the future”.

Is there any future goal or object which you expect to play an important role in

your life? Example: “I would like to meet somebody who would teach me how to

have a normal life”.

Two final questions:

How do you generally feel these days?

How would you like to feel?

These questions are not answered by a formulation. Instead, you answer them by adding

two extra rows in the matrix using the same list of affect terms.

117

The questions are read aloud by the interviewer while sitting next to the

participant, and relate to people, experiences, circumstances and goals from the past,

present and future, that are significant in the judgment of the participant. The questions

invite participants to reflect on their life in such a way that they feel free to mention

those concerns that are most relevant from the perspective of the present. Participants

are free to interpret the questions how they want. They are also encouraged to phrase the

valuation in their own terms. During the interview, each elicited valuation is written by

the participant on a small card. The typical form of expression is a sentence or

sentences, as long as it as felt to be one total experience, where the participant brings

together those events that he or she feels belong together as elements of a personal unit

of meaning. A quick response is not required, and there is no one-to-one relation

between question and answer.

The participant is encouraged to mention all valuations that come to mind and, as

is typical of the SCM, each question leads to more than one valuation. The aim of each

interview is to arrive at an exhaustive assessment (Hermans, 1993) of the experiences

relevant to the participant from each I-position. At the end of each interview participants

are asked whether it covers all the experiences they consider (from their vantage point

of their present I-position perspective) to be important. If something is missing, they are

free to add this. At the end of the procedure the range of valuations elicited from each

participant is in most cases between 20 to 40 valuations for each I-position. Also at the

end of the valuation elicitation phase, participants are asked the following questions:

“How do you generally feel these days?” and “How would you like to feel?” Responses

are used to gauge the extent of generalisation (G) and idealisation (I) of a valuation

within the system. These questions do not ask for a specific valuation but instead assess

the participant’s “general feeling” and “ideal feeling”, respectively.

118

Affective rating. Next, the participant is presented with a standard list of affect

terms (see Table 4.6), is asked to return to each valuation and, after reflecting on the

experience it related to, to rate on a 0-5 point scale (0 = not at all, 1 = a little bit, 2 = to

some extent, 3 = rather much, and 5 = very much) to what extent he or she experiences

each of the 24 affects in relation to each valuation. Participants use a matrix they are

provided with in which the rows consist of valuations and each of the columns

represents an affect term in order to rate each valuation with the same list of affect terms

(see Appendix A). Next, participants also use the same 0-5 point scale to answer each of

the last two final questions (regarding their general and ideal feelings) directly by going

over all the 24 affect terms that are used to rate the valuations. The last two rows of the

matrix are used by the participant to answer the last two questions referring to their

general and ideal feelings.

5.4 Procedure

Pre-testing. To pre-test the measures, the instruments were trialled on the first

two respondents. Trialling took the form of analysing the resulting data to determine

that they were credible and hence that the instruments were working as intended. In the

event, no problems were encountered at the pre-test stage and the instruments were not

modified as a result.

Self-report measures. In a preliminary telephone interview, potential participants

who met the selection criteria were sent the research questionnaire with a return

envelope. Eight persons did not meet the criteria; an explanation was given to them and

they were thanked for their interest. Selected participants were told what was required

of them, and informed of the time commitment expected, the general purpose of the

study and the use to which the data would be put. Participants signed a consent form

attached to each questionnaire and returned it to the researcher in the envelope provided.

119

Four people who originally agreed to participate later declined, stating that they

did not want to relive the experience of migration. All ethical procedures were

observed: participation in the study was voluntary and confidential. To ensure

confidentiality in the statistical analysis an identification number was allocated to each

participant, and participants’ names have been changed in the case studies. As

recounting migration experiences might cause distress, participants were informed that

they could stop at any time. One participant took advantage of this; the interview was

discontinued and the participant removed from the study sample. After participants had

completed both the questionnaire and the SCM, they were fully debriefed.

The questionnaire took approximately 45 minutes to complete and was returned to

the researcher. The questionnaire included: a Demographics page, the Social Anxiety

and Distress Scale (SADS), the Fear of Negative Evaluation Scale (FNES), the

Psychological Well-being Scale (PWS), and the Beck’s Depression Inventory-ii (BDI-

II). After completing the questionnaire containing the self-report measures, participants

returned them to the researcher in a pre-paid envelope. Participants were then

telephoned to schedule the first interview under the SCM protocol.

Self-confrontation method interview. The procedure described below closely

follows that set out in the SCM Manual (see Hermans & Hermans-Jansen (1995),

Chapter 9).

Participants took part in two or more in-depth face-to-face interviews using the

SCM. All participants were interviewed by the researcher, who was trained in the use of

the SCM at an international two-week summer course at the University of Nijmegen in

the Netherlands. Interviews averaged an hour and a half in length depending on the

participant; to avoid fatigue none exceeded two and a half hours. Thirty-two participants

were interviewed at home and six at the psychology clinic at Swinburne University’s

120

Hawthorn campus. Participants were told they would be questioned about what they

saw as important experiences in their past and present lives in their country of origin

and in Australia. The response of the thirty-eight participants was enthusiastic. In fact,

the main problem in the interviews was to keep them focused and for many the

interview was therapeutic either because they found new insights because of the

questions they were asked or because the interview setting afforded them an opportunity

to relieve important stories of their migration experience.

Identification of cultural positions of the self. At the first interview, participants

were invited to select two cultural identities of their personality, after confirming that

the sense of two identities corresponded to what was required for the research. They

were told: “Many migrants feel that they have developed two identities as a result of

moving to a new country. One identity retains aspects of how they were in their country

of origin; the other identity relates to their new country”. Participants were allowed time

to reflect and describe their perception of themselves as having two distinct identities in

their own terms.

Formulation of valuation systems from each I-position. Also at the first

interview, participants were invited to focus on their cultural identity of their country of

origin; to think about their past, present, and future, and relate important experiences

from this perspective (I-position). This produced a valuation system for each participant

from the viewpoint of their society of origin (Hermans, 2001). At the second interview,

which usually took place a week after the first one, participants were invited to focus on

their Australian cultural identity and to think of their past, present and future, and to

relate important experiences from this perspective (I-position). This resulted in a second

valuation system of each participant (Hermans, 2001).

121

By the end of the interview procedure the number of valuations elicited from each

participant averaged about 20 for each I-position.

Affective ratings. After the interviews had taken place, participants were asked to

return to each valuation and, after reflecting on the experience it related to, to rate on a

0-5 point scale to what extent he or she experienced each of the 24 affects in relation to

each valuation. Participants used a matrix provided by the researcher (in which the rows

consist of valuations and each of the columns represents an affect term—see Appendix

A) to rate each valuation with the same list of affect terms. Next, participants used the

same 0–5 point scale to answer each of the last two final questions (regarding their

general and ideal feelings) directly by going over all the 24 affect terms that were used

to rate the valuations. The last two rows of the matrix were used by the participant to

answer the last two questions referring to their general and ideal feelings. The matrices

were then collected by the researcher.

5.5 Treatment of the Interview Data

Once the affective ratings for the different valuations were obtained, affect scores

were aggregated into four indices, each representing a different motive or affect

(Table 5.6): the S index (range 0 to 20) obtained by summing the scores for the four

affects indicating the strength of the self-enhancement motive; the O index (range 0 to

20) obtained by summing the scores for the four affects indicating the strength of the

union-with-the-other motive; the P index (range 0 to 40) obtained by summing the

scores for the eight affects indicating the strength of positive affect; and the N index

(range 0 to 40) obtained by summing the scores for the eight affects indicating the

strength of negative affect. Table 5.6 shows the composition of the four indices

(Hermans &Hermans-Jansen, 1995). The indices represent the affective character of the

122

valuations and give information about the affective organisation of each participant’s

valuation system in general.

Besides the four indices discussed, there is a further, derived, index that is

sometimes useful: well-being (W), defined as P – N (Hermans &Hermans-Jansen, 1995,

27). In this study, well-being is used in Study 2 to reduce dimensionality of the dataset.

(This use of “well-being” is not to be confused with a participant’s score on the PWB

scale, also termed “well-being” in the literature.)

Table 5.6 Composition of indices in the SCM

Self-

enhancement (S)

Union-with-

the-other (O)

Positive

affect (P)

Negative

affect (N)

Self-esteem Caring Joy Powerlessness

Strength Love Satisfaction Anxiety

Self-confidence Tenderness Enjoyment Shame

Pride Intimacy Trust Self-alienation

Safety Guilt

Energy Loneliness

Inner calm Inferiority

Freedom Anger

“General” and “Ideal” feeling. Two additional indices—a general feeling index

and an ideal feeling index were derived. The extent of generalisation of any particular

valuation within an I-position system of valuations was found by computing the

product-moment correlation between the affective profile that belongs to a specific

valuation and the affective profile representing how the person generally feels. The

more positive the correlation (between the two affective profiles), the more this

valuation is supposed to generalise within a particular I-position system of valuations.

123

The scores of the general feelings index were also used to assess the general well-being.

This is represented by the P–N difference belonging to the general feelings.

To measure the extent of idealisation of any particular valuation, the same

procedure was followed. The correlation between the affective profile belonging to a

specific valuation and the affective profile belonging to the ideal feeling indicates the

ideal quality of this valuation. The height of the correlation indicates the extent of the

idealisation of the valuation within a particular I-position system of valuations. The

more positive the correlation between the two, the higher the idealisation.

Reliability of indices. The present thesis used Cronbach’s alphas to establish the

reliability of the S, O, P, and N indices, consistent with Hermans (1987b) who also used

this method with a group of 43 students (20 men and 23 women) and a group of 40

clients (20 men and 20 women). Hermans reported that in the student sample, the

reliabilities (coefficient alpha) of the S, O, P, and N indices were .83, .86, .90, and .88,

respectively. The correlation between S and O was .27, and between P and N was –.80.

In the client group (who had identity problems and problems in social relationships), the

reliabilities of the S, O, P, and N indices were .83, .89, .95, and .91, respectively. The

correlation between S and O was .64, and between P and N was –.67. When the groups

were compared, the client group showed lower scores for S (p < .001), lower scores for

O (p < .001), and higher scores for N (p < .001) than did the student group. No

difference was found between men and women (Hermans, 1987b). More recently, the

differences in patterning of the indices between the valuation systems of the college

student sample and the clinical sample was cited as evidence for the discriminant

validity of the SCM (Lyddon, Yowell, & Hermans, (2006).

The Cronbach’s alphas in the present thesis were excellent: .96, .96, .98, and .96,

for the S, O, P, and N indices respectively.

124

Coding of major themes. Another way of comparing valuations in the SCM is to

arrange them according to six “major themes” (Hermans and Hermans-Jansen, 1995).

The numerical criteria for assigning valuations to major themes vary between

researchers: “… the choice of such minimum scores is quite arbitrary” (Hermans &

Hermans-Jansen, 1995, p.254). However, they must comply with the general directions

set out in Table 5.7. The numerical rules that were applied in the current research are set

out in the Results section.

Strength and unity. Valuations that refer to an experience of “strength and unity”

(labelled +HH by Hermans) are typically characterised by both high self-enhancement

and high union-with-the-other motives, and because they are normally associated with

positive feelings, they are also characterised by both high positive affect and a low

negative affect. These are valuations in which the fulfilment of both basic motives

coincide, that is, the fulfilment of one motive goes with the fulfilment of the other, thus

creating high levels of well-being, as in “I made a puppet—something from myself for

my girlfriend." (S = 16, O =14, P = 17, N = 0)” (Hermans & Hermans-Jansen, 1995,

p.95). A valuation with this profile of scores refers to and is abbreviated as valuation of

the +HH type.

125

Table 5.7 Composition of major themes

Composition

Description S O P N

+HH Strength and unity High High High Low

+O Unity and love Low High High Low

–O Unfulfilled longing Low High Low High

–LL Powerlessness and isolation Low Low Low High

–S Aggression and anger High Low Low High

+S Autonomy and success High Low High Low

Unity and love. Valuations that refer to experience of “unity and love” (labelled

+O by Hermans) are typically characterised by low self-enhancement and high union-

with-the-other motives, and because the wish was fulfilled, they are also characterised

by both high positive affect and a low negative affect. These valuations depict an

increased contact and union with something else (a person, group, animal, object, or the

surrounding world) which is satisfying as in “I enjoy singing with a group of people,

being together without any pressure. (S = 4, O = 13, P = 19, N = 0)” (Hermans &

Hermans-Jansen, 1995, p.83). A valuation with this profile of scores refers to and is

abbreviated as valuation of the +O type.

Unfulfilled longing. Valuations that refer to “unfulfilled longing” (labelled –O by

Hermans) are typically characterised by low self-enhancement and high union-with-the-

other motives, and because the longing is unfulfilled, they are also characterised by both

low positive affect and high negative affect. These valuations depict “the adoption or

maintenance of a loving orientation toward another person or object that is, or is

126

imagined to be, unreachable”, as in “I regret that it is over with Alice. I want to hold on

to this feeling, and I will not allow the clichés to prevail; I find her too valuable. (S = 6,

O = 16, P = 2, N = 13)” (Hermans & Hermans-Jansen, 1995, p.83). A valuation with

this profile of scores refers to and is abbreviated as valuation of the –O type.

Powerlessness and isolation. Valuations that refer to an experience of

“powerlessness and isolation” are typically characterised by low self-enhancement and

low union-with-the other motives, and because they are normally associated with

negative feelings, they are also characterised by both low positive affect and high

negative affect. These are valuations where both motives are unfulfilled at the same

time, and where the individual typically feels that there is no way out, as in “I was raped

about 8 weeks ago by a man; because of this I feel humiliated and sad. (S = 0, O = 0, P

= 0, N = 18)” (Hermans & Hermans-Jansen, 1995, p.91). A valuation with this profile of

scores refers to and is abbreviated as a valuation of the –LL type.

Aggression and anger. Valuations that refer to an experience of “aggression and

anger” (labelled –S by Hermans) are typically characterised by both high self-

enhancement and low union-with-the-other motives, and because they are normally

associated with negative feelings, they are also characterised by low positive affect and

high negative affect. These valuations depict “some kind of opposition in which

individuals feel opposed to somebody or something that is threatening their self-

esteem”, as in “I am opposed to those people who want to decide for me what I must do.

(S = 15, O = 0, P = 3, N = 16)” (Hermans & Hermans-Jansen, 1995, p.70). A valuation

with this profile of scores refers to and is abbreviated as valuation of the –S type.

Autonomy and success. Valuations that refer to an experience of “autonomy and

success” (labelled +S by Hermans) are typically characterised by high self-enhancement

and high union-with-the-other motives, and because they are generally experienced as

127

positive, they are also characterised by high positive affect and a low negative affect.

These valuations depict a high degree of autonomy and productive coping with the

environment, as in “I hope to achieve a better position through my evening courses (S =

17, O = 4, P = 12, N = 4)” (Hermans & Hermans-Jansen, 1995, p.74). A valuation with

this profile of scores refers to and is abbreviated as valuation of the +S type.

5.6 Comparison with Previous Studies

Table 5.8 lists the dimensions of previous studies. The current research covers 38

participants and some 1600 valuations. Though not particularly large in terms of

respondents, it is possibly the largest of its kind in terms of valuations. This may be

because most studies of the dialogical self to date have been case studies with an

idiographic approach, and valuable as these are, they do not, nor are they intended to,

generate big datasets, without which detailed statistical analysis is impossible.

Besides having a larger sample than previous studies, the present thesis also

differs in sample composition. Table 5.9 compares the respondent sample for the

present thesis with those of other studies. Each sample is drawn from a minority group

(mostly immigrants but in one case an indigenous culturally distinct population) within

a majority group, and each is ranked on six demographic dimensions according to the

similarity of the minority and majority groups. The table shows that sample used in the

present thesis is demographically more similar to its majority culture (in this case

mainstream Australia) than are the samples used by the other studies to which it is

compared. Table 5.8 supports this conclusion in that it shows that most empirical

analysis to date using the SCM has been based on student and clinical samples, which

may be atypical of the wider population.

128

Table 5.8 Empirical studies using the Self Confrontation Method

Author (year) Name of study N Nature of

participants

Avge no. of

valuations

Hermans,

Hermans-Jansen, &

Van Gilst (1985)

The basic motives of human

existence: Their expression in

personal valuation

43

40

Students

Clinical

n.a.

Alford, Lyddon, &

Schreiber, (2006)

Adult attachment and working

models of emotion

143 Students 5 (min.)

Hermans & Oles,

(1996)

Value crisis: Affective

organization of personal

meanings

48 Students 8

Hermans & Oles

(1999)

Midlife crisis in men: affective

organization of personal

meanings

104 Men in

midlife crisis

9

Hermans & Oles

(1994)

The personal meaning of

values in a rapidly changing

society

53 Students 10

Rioux & Barresi

(1997)

Experiencing science and

religion alone and in conflict

40 Students 4

Hermans (1989) The meaning of life as an

organized process

25 Non-severe

clinical

n.a.

Table 5.9

Demographic difference between minority and majority cultures for selected studies

Author of study

Dimension of demographic difference

Volun-

tariness

Socio-

economic Race Religion

Lan-

guage Overall

score

Chandler and Lalonde

(1998; 2004; 2008) 4

Van Meijl (2012) 3

O’Sullivan-Lago and de

Abreu (2008) 2.5

Bhatia and Ram (2007) 2

This study 1

Legend: = All members of minority culture differ on the stated dimension; = Some

members of minority culture differ on the stated dimension.

Note: Overall score was computed by scoring 1 for each dimension of difference where all

respondents differ and 0.5 where only some respondents differ, then totalling.

129

In surveying the state of research to date, several authors have found that

empirical advances lagged behind theoretical ones. Concerning the SCM, Lyddon,

Yowell, and Hermans (2006) identified a need for more empirical work:

The Self Confrontation Method is gaining support as an effective approach to

assessment in both case study research and in empirical studies involving larger

samples …, though the latter are not as common. … Although the Self

Confrontation Method has garnered promising qualitative and quantitative

research support, research using the method with larger samples is warranted. (p.

40)

Hermans and Dimaggio (2007) concurred:

[T]o be recognized as a respected science, it is necessary to develop the dialogical

field in a theory-guided, empirical direction, taking advantage of both quantitative

and qualitative methods and of both experimental and experiential approaches ….

[N]ew and challenging theories should be created that may profit from equally

developed assessment methods and research procedures that are essential to revise

and improve existing theoretical notions. (p. 55)

As did Hermans (2008):

… the literature in the dialogical self field … shows more theoretical advances

than methodological and empirical elaborations. Therefore, future empirical work

on the basis of this theory should be encouraged and stimulated so that theoretical

progress goes hand in hand with empirical study and practical application. (p.186)

Stemplewska-Żakowicz, Walecka, Gabińska, Zalewski and Suszek (2005) also

advocated more empirical work, but with a particular emphasis. Having conducted

experiments on positioning that they describe as “exemplars of a successful dialogue

between post-positivist and social-constructivist methodology”, they conclude “that the

positivist tradition in methodology may be useful for studies within dialogical self

theory.”

130

These conclusions all point to a relative insufficiency of empirical research, which

the present study, with its quantitative focus, goes some way towards filling.

5.7 Personal Reflections on the Experience of Migration

Conventional wisdom has it that the scientist, including the social scientist, is an

impartial observer, neither influencing what is observed nor filtering it through the lens

of subjectivity. But if this view was ever widely held, it is less so now than ever,

particularly in the field of narrative research which, being inherently dialogical, must

have an author in dialogue with an audience. According to Bruner (1990): "The-story-

of-a-life as told to a particular person is in some deep sense a joint product of the teller

and the told" (p. 124). Stiles (1993) stated: “… good practice recommends disclosure by

the investigator of his or her expectations for the study, preconceptions, values, and

orientation…” (p. 602). Most recently, Lawrence-Lightfoot (2005) stated: “All

researchers—whether working within the quantitative or qualitative methodological

paradigm—are selective in defining and shaping the data they collect and the

interpretations that flow from their findings…The shaping hand of the investigator is

counterbalanced by the skepticism and scrutiny that is the signature of good research”

(p. 11).

If, then, the researcher inevitably influences the outcome of the research, how is

this to be accommodated? Wortham (2000) and Potter and Hepburn (2005) have

developed and advocate analytical tools and procedures for delving more deeply into the

“interactional” (as opposed to “representational”) aspect of dialog. But these tools,

powerful as they are, are too time-consuming to use on the approximately 1600

narrative “valuations” elicited in the present research. However, many of these

problems, namely those that Potter and Hepburn term “contingent”, can be addressed by

131

the interviewer’s appropriate positioning in the narrative dyad. This was elaborated by

Hermans (2000):

the interactional nature of the autobiographical interview requires the interviewer

to develop a position repertory which helps him to tune in to the flux of

experiences on the side of the narrator. A broad repertory which enables the

interviewer to move flexibly from one to another position, in accordance with the

needs of the narrator, is a precondition for studying the person’s autobiography in

all its complexity and multifacetness.

Fortunately my experience as a migrant has equipped me with a broad repertoire

(to use Hermans’ terminology). In two migrant episodes at very different stages in my

life I have been, or have attempted to be, variously “separated”, “assimilated”,

“integrated” and “marginalised”, (to use Berry’s terminology) and sometimes more than

one at the same time, if in different domains of life—separated at home, and assimilated

(or attempting to be so) at work. This has enabled me to read, and empathise with, my

interviewees. It is for this reason that I write this section, and moreover do so in the first

person, to remind the reader that the present research was in some degree shaped by my

experiences as a migrant myself. What follows is a brief autobiographical sketch.

I was born and raised in the Dominican Republic. My first migration (though at

the time I had no idea how long-lasting it would be) was supposedly to further my

education. But the real reason was that my parents wanted me out of the house. That

suited me; I wanted to go. So they arranged for me stay with an aunt while I studied

English in New York.

My aunt’s household was Dominican from top to bottom—artefacts, music, food,

everything. She was a political exile. Her husband, a prominent political and intellectual

figure, had been assassinated by the former dictator. But by the time he was

assassinated in his turn, and her return became possible, her children, my cousins, had

132

become Americanised and completely bilingual, while she still spoke very little English.

Torn as she was between children and homeland, she stayed. Lawson (2000) describes

this existential migrant condition as “a state of in between-ness”, and I was part of it. I

inhabited a Dominican “ghetto” where everybody knew who I was and what I was. In

the words of Deaux (2000), I was “remoored”.

Coming from a privileged family, I had lived a materially comfortable, but

sequestered, existence. I found New York liberating and transforming. Although living

with family, I was the freest I had ever been. At first I struggled with English. I did not

want to sound like an outsider. “Passing”, as Goffman (1968) put it, was my goal. Like

any teenager, I just wanted to fit in (Camilleri and Malewska-Peyre 1997). Going back

was always on the cards, and in fact central to my Dominican identity. But it was not

what I chose—at least, not straight away. I was to spend nine years in the States—a

marriage, a child, a divorce, by which time my country of birth had become haloed by

nostalgia. Only then, in a way defeated, did I return. Looking back, my migrant

identities—Dominican and American—reflected the ambivalence that makes a person

long for a place when living in another, identifying with home when abroad, and with

abroad when home (King, 1995). I knew then that migration changes one, forever,

which is why it is so often associated with loss. Life became complicated and I became

wistful, always lacking, never complete.

My second migration was to Australia, where I now live. It was to accompany my

second partner, a migrant himself, to his adopted country. By now, I was fluent in

English, and although the my accent was never to leave me, I was comfortable with it.

In Australia I embraced my difference. I was Dominican in Australia and so an outsider,

but by now I accepted my difference as a creative resource, not a blemish to be erased. I

played with cultural repertoires, confident in acquiring a new cultural dimension—

133

becoming Australian. Migration gave me a new perspective, a “third space” where I

could explore and extend my cultural repertoire, creatively reconstructing my

Dominican and Australian identities as “hybrid” (Bhabha 1994).

I’m now a “foreigner” who speaks fluent English and lives in Australia. I am in it

but not of it. I am between cultures. It is a role that can be hard to sustain, and has on

occasion been the cause of unhappiness. But it also offers something precious: I am able

to observe Australian culture with a degree of detachment and in this way reflect all the

more on my own life in time and space. At times I feel a bitter-sweet nostalgia for my

country of birth, for what was and never will be. But at other times my nostalgia is like

an illness, which is what Johannes Hofer, the doctor who coined the word ‘nostalgia’ in

1688, believed it to be. It is a rare migrant that has not felt the pain of letting go the life

they once knew.

The migrant story is one of pain. In Australia this is only beginning to be widely

discussed. “In Australia I will”, the posters once trumpeted. Well, some did and some

didn’t; and those who did, often had a hard time of it. Some deplore all this

remembering. Nostalgia would seem to be, after all, the opposite of progress. But

Australia’s migrant experience is not an unmixed march of progress. It also contains

countless stories of heartache. The nation’s identity would be greatly enriched if both

sides of the migrant story were told. In Carey’s (2002) words: “… we are a nation of

immigrants…”.

Having thus disclosed my preconceptions, values, and orientation (following

Stiles, 1993) I now come to my expectations for the present research. As should by now

be apparent, this thesis is autobiographical in two senses: it calls on my participants to

divulge their autobiographical details; equally, it is strengthened by my candid

disclosure of my own autobiographical experience of migration—in general terms if not

134

in detail—in that my participants were all aware that I shared this aspect of life

experience with them. In short, I satisfied all the criteria that I required of my

participants; they knew it, and the knowledge possibly credentialled me in their eyes.

My research took as a framework my own experience of migration; and sharing this

experience with my participants was crucial to establishing the authenticity of my

relationship with them. In this I followed Janesick’s (1998) precept of “researcher as

instrument”, in which I was a participant observer using my own subjectivity to

interpret data.

More specifically, I wanted to give my survey participants the psychological

space to talk freely, and in this way I hoped to encourage them to reflect on, and make

sense of, their bifurcated life stories and possibly disrupted sense of self. I hoped that it

would offer participants constructive outcomes as a result of migration. Accordingly,

my research did not harbour presuppositions about the migration experience, nor did it

define a population of which my participants would constitute a representative sample.

On the contrary, I expected diversity in self-narratives, which made the reported

similarities in emotional valence all the more striking.

135

CHAPTER 6: Study 1 – Identification and Comparison of Migrant Groups

This chapter discusses the first of the three studies conducted as part of this thesis.

All were based on the same data and were complementary in nature. Study 1 was

largely exploratory. Its aim was not to test hypotheses but to reveal the existence of

naturally forming groups among migrants, that is, to determine whether migration

brings into being new I-positions, and if so, whether there exist any patterns among

migrants in their I-positions in country of origin and Australia that can be discerned in

terms of demographics, psychological self-reports and self-confrontation indices.

Four groups of participants were identified by means of cluster analysis. Because

cluster analysis is an exploratory technique that lacks formal tests of statistical

significance, it was only once the groups were identified that hypotheses were tested.

Significant differences were found among the groups, and these differences were found

to be plausible in terms of theory. The four groups identified in Study 1 were

corroborated by Studies 2 and 3. Statistical analyses were performed using SPSS 16.0.

6.1 Preliminary Analysis

Self-report measures of distress and well-being. The distributions of all

measures were inspected for outliers. Preliminary data screening found a single high

outlier on BDI–II, and single low outliers on PWB Personal growth and Purpose in life.

Three low outliers were found on PWB Self-acceptance, two of which were marginal.

Other than the two marginals, all other outliers related to the same participant, which

suggests that they were not due to a coding error but represented a single participant

with mild depression.

Table 6.1 shows the sample means, standard deviations, skewness and kurtosis for

the self-report measures of distress and well-being. Kolmogorov-Smirnov tests showed

136

that normality assumptions were significantly violated at the 5% level for SAD, FNE,

BDI-II and, marginally, PWB Self-acceptance. Although this might cast doubt on later

MANOVAs with these variables, MANOVA is normally robust to such violations

(Tabachnik & Fidell, 2006).

Table 6.1 Means and standard deviations for the self-report measures

Psychological Distress Psychological Well-being Scale

SAD FNE BDI–II PWB1 PWB2 PWB3 PWB4 PWB5 PWB6

Mean 8.4 13.3 9.1 63.1 67.1 63.4 72.7 66.1 64.6

SD 7.2 9.5 9.3 12.1 12.1 11.9 8.0 10.8 12.4

Skewness .831 .374 1.784 -.656 -.568 -.638 -1.018 -1.024 -.876

Kurtosis -.607 -1.261 4.099 .703 -.063 -.672 1.810 2.995 .568

N = 38

Note: SAD = Social Avoidance and Distress Scale; FNE = Fear of Negative Evaluation Scale;

BDI-II = Beck Depression Inventory–II; PWB1 = Environmental mastery; PWB2 = Positive

relations with others; PWB3 = Autonomy; PWB4 = Personal growth; PWB5 = Purpose in life;

PWB6 = Self-acceptance.

Table 6.2 presents bivariate correlations of the self-report measures of distress and

well-being. All were significantly (p < .01) correlated. As can be seen, the

psychological distress scales were significantly positively correlated with each other.

Participants who had a high SAD score tended to have a high FNE score (r = .63);

participants who had a high SAD score also tended to have a high BDI-II score

(r = .45); and participants who had a high FNE score tended to have a high BDI-II score

(r = .57). The PWB subscales, all of which measure various components of

psychological well-being, were strongly positively correlated with each other (r = .63 to

r = .90). Lastly, the PWB subscales were strongly negatively correlated with the three

psychological distress scales (r = –.49 to r = –.80).

137

Table 6.2 Correlations among self-report measures

Psychological Distress Psychological Well-being Scale (PWB)

SAD FNE BDI-II PWB1 PWB2 PWB3 PWB4 PWB5 PWB6

Social Avoidance and Distress Scale (SAD)

Fear of Negative Evaluation Scale (FNE) .628**

Beck Depression Inventory–II (BDI-II) .449** .569**

PWB Environmental mastery (PWB1) -.575** -.668** -.744**

PWB Positive relations with others (PWB2) -.610** -.586** -.592** .725**

PWB Autonomy (PWB3) -.570** -.734** -.561** .718** .638**

PWB Personal growth (PWB4) -.508** -.528** -.623** .664** .661** .725**

PWB Purpose in life (PWB5) -.478** -.544** -.796** .863** .698** .628** .731**

PWB Self-acceptance (PWB6) -.524** -.672** -.750** .903** .677** .820** .759** .815**

N = 38

Note: Values are Pearson’s r correlations, * p < .05 (two-tailed), ** p < .01 (two-tailed).

138

Self-confrontation data. Each participant in the study produced about 40

valuations on average, divided roughly evenly between those relating to country of

origin and those relating to Australia. They were further split across valuations relating

to the past, the present, and the future.

Preliminary data screening found no outliers in the self-confrontation indices (S,

O, P and N) as they describe participants’ valuations relating to country of origin and

Australia. Table 6.5 (Section 5.2, p.116) shows the sample means, standard deviations,

skewness and kurtosis for self-confrontation indices at country of origin and Australia.

Kolmogorov-Smirnov tests showed that normality assumptions were significantly

violated at the 5% level for O for country of origin, and N for Australia. Although this

might cast doubt on later MANOVAs with these variables, MANOVA is normally

robust to such violations (Tabachnik & Fidell, 2006).

Table 6.3 presents bivariate correlations of the indices. For each set of indices

(i.e., for country of origin, Australia, general feeling, and ideal feeling), S, O and P in

the same set were all significantly correlated with each other (p < .01). For country-of-

origin indices, participants who were high on S tended also to be high on O and P

(r = .84 to r = .95). For Australian indices, participants who were high on S tended also

to be high on O and P (r = .84 to r = .95). For indices relating to general feeling,

participants who were high on S tended also to be high on O and P (r = .75 to r = .91).

For indices relating to ideal feeling, participants who were high on S tended also to be

high on O and P (r = .75 to r = .78). Overall, high S and O were statistically associated

both with each other and with high P. The same was also generally true if the indices

came from different sets (although the ideal feeling indices were less strongly correlated

with those of other sets).

139

Table 6.3 Correlations among self-confrontation indices

Country of origin Australia General feeling Ideal feeling

S O P N S O P N S O P N S O P N

Country of

origin

S

O .839**

P .947** .885**

N -.050 .056 -.074

Australia S .764** .691** .711** .017

O .640** .781** .657** .002 .836**

P .760** .734** .771** -.045 .941** .884**

N .047 .057 .065 .662** -.306 -.274 -.323*

General

feeling

S .461** .445** .433** -.116 .732** .522** .600** -.322*

O .461** .599** .505** -.192 .746** .757** .711** -.400** .753**

P .428** .553** .470** -.103 .694** .615** .630** -.314 .910** .818**

N -.139 -.218 -.145 .362* -.435** -.415** -.375** .598** -.657** -.629** -.732**

Ideal

feeling

S .320 .338 .294 .260 .397* .282 .252 .008 .452** .266 .357* -.130

O .514** .601** .530** .121 .538** .496** .481** -.038 .562** .509** .538** -.197 .781**

P .292 .413** .312 .202 .241 .253 .181 .124 .386* .290 .372* -.044 .746** .777**

N .069 .172 .105 .192 .110 .215 .199 .150 -.245 .095 -.122 .322** -.311 -.193 -.197

N = 38

Note: Values are Pearson’s r correlations, * p < .05 (two-tailed), ** p < .01 (two-tailed). S = self-enhancement; O = union-with-the-other; P = positive affect;

N = negative affect.

140

The relationships among S, O and P on the one hand, and N on the other, were

less clear-cut. For country-of-origin indices, N was not significantly correlated with S, O

and P (r = –.07 to r = .06). For Australian indices, N was not significantly correlated

with S (r = –.31) or O (r = –.27), but was moderately negatively correlated with P (r = –

.32, p < .05). For indices relating to the ideal feeling, N was not significantly correlated

with S, O or P (r = –.19 to r = .31). However, for indices relating to the general feeling,

N was strongly negatively correlated with S, O and P (r = –.63 to r = –.73).

In general, correlations were highest between indices within the same set. Lower

(but still significant) correlations were observed between general feeling indices and

those in country of origin. The lowest (mostly not significant) correlations were

between ideal feeling and the other sets.

Relationship between self-report measures and self-confrontation indices.

Table 6.4 presents bivariate correlations of the self-report measures and self-

confrontation indices. In general, self-report measures were not significantly correlated

with self-confrontation indices for the country of origin I-position, and hardly at all with

ideal feeling indices, but were mostly significantly correlated with indices for the

Australian I-position, and even more so with general feeling indices.

As is evident from Table 6.4, for country of origin, self-confrontation indices

were significantly correlated with measures of psychological distress in one instance:

negative affect was positively correlated with BDI-II (r = –.49, p < .01). Likewise, self-

confrontation indices were significantly correlated with measures of psychological well-

being in one case: P was positively correlated with PWB Autonomy (r = –.36, p < .05).

141

Table 6.4 Correlations among self-reports and self-confrontation indices

Country of origin Australia General feeling Ideal feeling

S O P N S O P N S O P N S O P N

SAD -.049 -.063 -.101 .259 -.286 -.206 -.249 .347* -.399* -.449** -.312 .379 -.071 -.166 -.131 .132

FNE -.149 -.109 -.222 .222 -.368* -.133 -.287 .355* -.482** -.399* -.426** .421** -.237 -.254 -.176 .093

BDI-II -.087 -.132 -.160 .491** -.311 -.237 -.335 .647** -.467** -.538** -.517** .654** .047 -.133 .064 .120

PWB1 .267 .267 .289 -.286 .516** .363* .446** -.476** .662** .598** .656** -.608** .181 .328* .163 -.193

PWB2 .179 .233 .201 -.214 .532** .437** .499** -.561** .601** .630** .557** -.638** .249 .372* .227 -.174

PWB3 .291 .216 .358* -.284 .432** .228 .391* -.376* .622** .407* .510** -.377* .320* .448** .384* -.316

PWB4 .194 .247 .274 -.096 .497** .307 .457** -.297 .706** .485** .640** -.490** .293 .374* .173 -.198

PWB5 .263 .294 .315 -.277 .510** .391* .481** -.439** .653** .660** .648** -.585** .091 .272 .122 -.041

PWB6 .224 .160 .255 -.299 .476** .258 .410* -.483** .673** .521** .609** -.538** .158 .279 .103 -.311

N = 38

Note: Values are Pearson’s r correlations, * p < .05, two-tailed, ** p < .01, two-tailed. S = self-enhancement; O = union-with-the-other; P = positive affect;

N = negative affect; PWB1 = Environmental mastery; PWB2 = Positive relations with others; PWB3 = Autonomy; PWB4 = Personal growth;

PWB5 = Purpose in life; PWB6 = Self-acceptance.

142

In the case of the Australian I-position, self-confrontation indices were

significantly correlated with measures of psychological distress in three instances: S was

negatively correlated with FNE (r = –.37, p < .05); and N was positively correlated with

both SAD (r = .35, p < .01) and FNE (r = .36, p < .01). Self-confrontation indices were

also significantly correlated with measures of psychological well-being: S was

positively correlated with all PWB subscales (r = .42 to .53, all at p < .05); O was

positively correlated with PWB Environmental mastery (r = .36, p < .05), Positive

relations with others (r = .44, p < .01), and Purpose in life (r =.39, p < .05); P was

positively correlated with all PWB subscales (r = .39 to .50, most at p < .01) ); and N

was negatively correlated with all PWB subscales except Personal growth (r = –.38 to –

.56, most at p < .01).

In the case of general feeling, self-confrontation indices were significantly

correlated with all measures of psychological distress in all cases but two: S (r = –.40 to

–.48, mostly at p < .01); O (r = –.40 to –.54, p < .05); P (r = –.43 to –.52, p < .05); and

N (r = .42 to .65, p < .05). Self-confrontation indices were also significantly correlated

with all measures of psychological well-being: positively with S (r = .60 to .71, all at p

< .05); positively with O (r = .41 to .66, most at p.< .05); positively with P (r = .51 to

.66, all at p < .05) ); and negatively with N (r = –.49 to –.64, most at p < .05).

In the case of ideal feeling, self-confrontation indices were not significantly

correlated with measures of psychological distress. However, they were significantly

correlated with measures of psychological well-being in several cases. S was positively

correlated with PWB Autonomy (r = .32, p < .05); O was positively correlated with

PWB Environmental mastery (r = .33, p < .05), PWB Positive relations with others (r =

.37, p < .05), and PWB Autonomy (r = .45, p < .01); and P correlated with PWB

Autonomy (r = .38, p < .05).

143

6.2 Distinguishing I-positions

The theory of dialogical self implies that life-changing events such as migration

can give rise to additional I-positions, and that these I-positions may be accessed

through the SCM. Although not all migrants necessarily develop in this way, the

participants in this study were required to meet the criterion that they perceived

themselves as having “developed two identities as a result of moving to [Australia]”,

where one identity retains aspects of how they were in their country of origin and the

other identity relates to their new country. As part of the data exploration, therefore,

analogous self-confrontation indices pertaining to each participant’s I-position in

country of origin were compared with those pertaining to the I-position in Australia

with the aim of determining if the group means differed significantly.

Comparing self-confrontation indices for country of origin and Australian I-

positions. Paired-samples t-tests were conducted to compare self-confrontation indices

for country of origin and Australian I-positions. Table 6.5 shows means and standard

deviations of self-confrontation indices. Independent samples t-tests were conducted to

compare the self-confrontation indices for country of origin and Australia. Results

revealed that S was significantly greater for Australia than for country of origin,

t(37) = –3.31, p = .002 (two-tailed); and P was significantly greater for Australia than

for country of origin, t(37) = –2.37, p = .023 (two-tailed). No significant differences

were found in the case of O or N. Mean O for Australia did not differ significantly from

that for country of origin, t(37) = –1.12, p = .270 (two-tailed); nor did N for Australia

differ significantly from that for country of origin, t(37) = –0.49, p = .625 (two-tailed).

However, the significance of the difference in at least some self-confrontation indices,

in this case S and P, shows that a new I-position has come into existence, for some

participants at least.

144

Table 6.5 Means and standard deviations of self-confrontation indices

Location

Index

S O P N

Country of origin Mean 10.02 9.54 19.11 10.03

SD 3.91 4.23 7.83 6.14

Skewness -.123 -.211 -.133 .503

Kurtosis -.729 -.997 -.884 -.787

Australia Mean 11.45 10.07 21.17 10.45

SD 3.86 4.52 7.96 6.40

Skewness -.036 -.169 -.014 .378

Kurtosis .561 -1.220 -.873 -.926

N = 38

Note: S and O are scored from 0 to 20; P and N from 0 to 40; S = self-enhancement; O = union-

with-the-other; P = positive affect; N = negative affect.

What these results show is that the differences between means for country of

origin and Australia were small (though nevertheless significant), and that the

variability in self-confrontation indices within groups were large (as revealed in their

standard deviations). There is a simple explanation for both these findings, namely that

the sample contained disparate responses. For some the new I-position reflected a view

of life in Australia as an unalloyed reward, so creating a new I-position with more

positive affect and greater satisfaction of the S- and O-motives; for others, it was the

very opposite; and for others there was little difference. Over the whole sample, then,

the effects may cancel out, so that group mean self-confrontation indices differ little,

when in fact the self-confrontation indices for individual participants differ greatly, but

in opposite directions (a case of the ‘ecological fallacy’). This implies that significant

145

differences in self-confrontation indices may be found when the sample is examined in

more detail, as was done next.

Cluster analysis. Cluster analysis was used as a data reduction method to classify

participants into a small number of relatively homogeneous groups for the purpose of

exploring possible relationships among variables that were masked at the aggregate

level of the whole sample. Participants were characterised according to their I-positions

in country of origin and Australia. Since it was considered that the country-of-origin I-

position might differ from the Australian, two cluster analyses were conducted, one to

classify participants into clusters determined by their self-confrontation indices for

country of origin, the other to cluster participants by their self-confrontation indices for

Australia. Thus, each participant was characterised by membership of two clusters.

Two cluster methods were employed. Hierarchical clustering was selected as the

principal clustering method as it requires no a priori assumptions as to the nature of the

clusters, instead allowing optimal cluster centroids to be determined as part of the

clustering procedure (Clatworthy, Buick, Hankins, Weinman & Horne, 2005). Self-

confrontation indices were clustered according to the CLUSTER procedure in SPSS,

using Ward’s minimum variance method with a squared Euclidean distance metric to

represent dissimilarity between case-pair. This is an agglomerative hierarchical cluster

analysis procedure that joins clusters of cases together in a way that minimises increases

in the within-cluster variance (or fusion coefficient) (Clatworthy et al., 2005). Because

index metrics differed markedly in range and mean, they were converted to z-scores for

clustering purposes so as to avoid bias when computing squared Euclidean distance.

Since clustering techniques lack formal tests of statistical significance,

Beauchaine and Beauchaine (2002) recommend that the robustness of the hierarchical

solution be assessed by re-clustering by k-means clustering. Unlike hierarchical

146

clustering, k-means clustering is an non-hierarchical iterative procedure that assigns and

reassigns cases to the “closest” cluster as measured by Euclidean distance, and requires

that the number and initial centroids of the clusters be specified in advance. When k-

means clustering is used to corroborate hierarchical clustering, it is appropriate to

“seed” the k-means cluster means with those previously established by hierarchical

clustering (Beauchaine and Beauchaine, 2002).

There are no hard-and-fast rules for determining the appropriate number of

clusters, but the cluster dendrogram and agglomeration schedule are taken to be a good

guide. Parsimony favours fewer clusters, while explanatory power favours more

(Aldenderfer & Blashfield, 1984). An abrupt lessening in the incremental reduction in

within-cluster variance in the agglomeration schedule as the number of clusters grows,

and a similar shortening in dendrogram branches, show where further disaggregation is

likely to become unhelpful.

Clusters in country of origin

Hierarchical clustering. The agglomeration schedule (Figure 6.1) and

dendrogram (Figure 6.2) were examined for clustering by self-confrontation indices for

country of origin.

147

Figure 6.1 Agglomeration schedule: cluster in country of origin

Various clustering solutions were tried with the aim of achieving a satisfactory

trade-off between parsimony and explanatory power. A two-cluster solution reduced the

within-cluster variance by about half, a three-cluster solution by about two-thirds, and a

four-cluster solution by about three-quarters (see Figure 6.1). The two-cluster solution

was selected because it produced cluster sizes (21 and 17 respectively) that were

sufficiently large to permit MANOVA, which ideally requires a sample size of 20 in

each cell to ensure “robustness” (Tabachnik & Fidell, 2001).

The dendrogram in Figure 6.2 shows how participants were assigned to clusters in

country of origin. Cluster 1 contained 21 participants and cluster 2, 17 participants. (For

reasons of confidentiality participants are identified by a unique number followed by

their first initial.) The length of the dendrogram branches following the first split

virtually guarantees the validity of at least a 2-cluster solution.

0

50

100

150

0 5 10 15 20 25 30 35

Stage

Wit

hin

-clu

ste

r v

ari

an

ce

(fu

sio

n c

oe

ffic

ien

t)

Two-cluster

solution

148

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 25

Label Num +---------+---------+---------+---------+---------+

Cluster 1

18 E-- 18 ─┐

37 C-- 35 ─┤

15 E-- 15 ─┤

11 D-- 11 ─┤

39 R-- 36 ─┼─────────────┐

30 G-- 29 ─┤ │

02 D-- 2 ─┘ │

06 D-- 6 ─┐ │

34 E-- 32 ─┼─────┐ ├─────────────────────────────────┐

07 F-- 7 ─┤ │ │ │

22 J-- 22 ─┤ │ │ │

21 D-- 21 ─┤ │ │ │

42 K-- 38 ─┘ ├───────┘ │

23 M-- 23 ─┐ │ │

26 D-- 26 ─┼─┐ │ │

01 A-- 1 ─┤ │ │ │

28 G-- 28 ─┘ ├───┘ │

03 Y-- 3 ─┐ │ │

25 N-- 25 ─┤ │ │

17 L-- 17 ─┼─┘ │

13 I-- 13 ─┘ │

│

Cluster 2 │

04 N-- 4 ─┐ │

12 L-- 12 ─┤ │

36 R-- 34 ─┼─────┐ │

05 H-- 5 ─┤ │ │

10 A-- 10 ─┤ │ │

20 S-- 20 ─┤ │ │

40 C-- 37 ─┤ │ │

27 J-- 27 ─┤ │ │

33 F-- 31 ─┘ ├─────────────────────────────────────────┘

09 S-- 9 ─┐ │

35 L-- 33 ─┤ │

16 I-- 16 ─┤ │

19 B-- 19 ─┤ │

08 K-- 8 ─┤ │

24 L-- 24 ─┼─────┘

32 G-- 30 ─┤

14 M-- 14 ─┘

Figure 6.2 Dendrogram of 2-cluster solution: cluster in country of origin

149

K-means clustering. Table 6.6 shows that there was a close match between cluster

membership under both clustering methods. This largely corroborates the hierarchical

clustering procedure.

Table 6.6 Number of participants by cluster: country of origin

K-means cluster Hierarchical cluster

Cluster 1 Cluster 2 Total

Cluster 1 19 – 19

Cluster 2 2 17 19

Total 21 17 38

Two participants were classified differently by the clustering procedures:

participants 01 and 34 were both placed in hierarchical cluster 1 but in k-means cluster

2. It is possible that this was because they lay on the margins of their respective clusters;

a change in the clustering algorithm therefore caused them to switch from one cluster to

another. An examination of the sum of squared differences between each observation

and cluster centroids supports this supposition. Figure 6.3 confirms that k-means

clustering correctly assigned each participant to the cluster with the “nearest” centroid,

as required by definition. However, in some cases participants were almost equidistant

between pairs of clusters, and in such cases hierarchical clustering assigned the

participant to the other cluster of the pair. This is of little consequence for the analysis.

It means that the affected participant has features of both clusters—its hierarchical

cluster and its k-means cluster. This is expected in cluster analysis, where it is normal

for some observations to lie near cluster boundaries (Aldenderfer & Blashfield, 1984).

What is important in this case is that there were few instances of this, which suggests

that cluster identity was mostly unambiguous.

150

Figure 6.3 Sum of squares from centroid: cluster in country of origin

Cluster means. As the clusters were purposely created to be as different as

possible on the selected variables, significant differences between the groups on those

variables were to be expected (Aldenderfer & Blashfield, 1984). In the present study,

the purpose of conducting significance t-tests or ANOVAS using the variables that were

used to form the clusters was to get an indication of whether and how clusters differed

on each clustering variable (Mulry, Kalichman, Kelly, Ostrow, & Heckman, 1997), that

is, to clarify the relative contributions of the different variables to the cluster solution or

the degree of separation of the clusters.

Table 6.7 and Figure 6.4 show mean self-confrontation indices for country of

origin broken down by cluster in country of origin. Independent samples t-tests were

conducted to compare the mean self-confrontation indices for both clusters. Cluster 1

had lower S than Cluster 2, t(36) = –8.74, p < .001 (two-tailed). Cluster 1 also had lower

O than Cluster 2, t(36) = –7.31, p < .001 (two-tailed). Cluster 1 also had lower P than

0

2

4

6

8

10

12

14

16

18

20

01

A--

-

02

D--

-

03

Y--

-

04

N--

-

05

H--

-

06

D--

-

07

F--

-

08

K--

-

09

S--

-

10

A--

-

11

D--

-

12

L--

-

13

I--

-

14

M--

-

15

E--

-

16

I--

-

17

L--

-

18

E--

-

19

B--

-

20

S--

-

21

D--

-

22

J--

-

23

M--

-

24

L--

-

25

N--

-

26

D--

-

27

J--

-

28

G--

-

30

G--

-

32

G--

-

33

F--

-

34

E--

-

35

L--

-

36

R--

-

37

C--

-

39

R--

-

40

C--

-

42

K--

-

Respondent

Su

m o

f s

qu

are

d d

iffe

ren

ces

fro

m

hie

rarc

hic

al

clu

ste

r c

en

trio

d

Cluster 1

Cluster 2

Hierarchical cluster

conflicts with K-means Hierarchical cluster

conflicts with K-means

151

Cluster 2, t(36) = –9.21, p < .001 (two-tailed). However, the clusters did not differ

significantly on N, t(36) = .88, p = .385 (two-tailed).

Table 6.7 Mean self-confrontation indices for country of origin I-position, by cluster in country of

origin

Cluster in

country of origin

Index for country of origin I-position

S O P N

Cluster 1 (n = 21) Mean 7.15 6.64 13.28 10.83

SD 2.55 3.22 4.94 6.79

Cluster 2 (n = 17) Mean 13.55 13.13 26.31 9.06

SD 1.78 1.93 3.43 5.27

Note: S and O are scored from 0 to 20; P and N from 0 to 40. S = self-enhancement; O = union-

with-the-other; P = positive affect; N = negative affect.

Figure 6.4 Mean self-confrontation indices for country of origin I-position, by cluster in

country of origin

Note: S and O are scored from 0 to 20; P and N from 0 to 40.

0

5

10

15

20

25

30

35

S-index O-index P-index N-index

Index

Mean

sco

re

Cluster 1

Cluster 2

152

Clusters in Australia

Hierarchical clustering. The agglomeration schedule (Figure 6.5) and

dendrogram (Figure 6.6) were examined, and a two-cluster solution selected for the

same reasons as applied to clustering by self-confrontation indices for the Australian I-

position.

Figure 6.5 Agglomeration schedule: cluster in Australia

The dendrogram shows how participants were assigned to clusters in country of

origin. Cluster 1 contained 14 participants and cluster 2, 24 participants. The length of

the dendrogram branches following the first split virtually guarantees the validity of at

least a 2-cluster solution.

0

50

100

150

0 5 10 15 20 25 30 35

Stage

Wit

hin

-clu

ste

r v

ari

an

ce

(fu

sio

n c

oe

ffic

ien

t)

Two-cluster

solution

153

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 25

Label Num +---------+---------+---------+---------+---------+

Cluster 1

14 M-- 14 ─┐

35 L-- 33 ─┤

05 H-- 5 ─┤

01 A-- 1 ─┤

24 L-- 24 ─┼───┐

08 K-- 8 ─┤ │

12 L-- 12 ─┤ │

03 Y-- 3 ─┘ ├───────────────────────────────────────────┐

10 A-- 10 ─┐ │ │

27 J-- 27 ─┤ │ │

33 F-- 31 ─┤ │ │

34 E-- 32 ─┼───┘ │

32 G-- 30 ─┤ │

36 R-- 34 ─┘ │

│

Cluster 2 │

26 D-- 26 ─┐ │

28 G-- 28 ─┤ │

19 B-- 19 ─┼───┐ │

13 I-- 13 ─┤ │ │

17 L-- 17 ─┤ │ │

18 E-- 18 ─┘ ├───────┐ │

11 D-- 11 ─┐ │ │ │

39 R-- 36 ─┤ │ │ │

37 C-- 35 ─┤ │ │ │

25 N-- 25 ─┼───┘ ├───────────────────────────────────┘

15 E-- 15 ─┘ │

02 D-- 2 ─┬─────┐ │

21 D-- 21 ─┘ │ │

09 S-- 9 ─┐ │ │

42 K-- 38 ─┼─┐ ├─────┘

06 D-- 6 ─┤ │ │

20 S-- 20 ─┤ │ │

16 I-- 16 ─┘ ├───┘

07 F-- 7 ─┐ │

40 C-- 37 ─┤ │

22 J-- 22 ─┼─┘

30 G-- 29 ─┤

04 N-- 4 ─┤

23 M-- 23 ─┘

Figure 6.6 Dendrogram of 2-cluster solution: cluster in Australia

154

K-means clustering. Table 6.8 and Figure 6.7 show that there was a perfect match

between cluster membership under both clustering methods. This corroborates the

previous hierarchical clustering procedure.

Table 6.8 Number of participants by cluster: Australia

K-means cluster Hierarchical cluster

Cluster 1 Cluster 2 Total

Cluster 1 14 – 14

Cluster 2 – 24 24

Total 14 24 38

Figure 6.7 Sum of squares from centroid: cluster in Australia

0

2

4

6

8

10

12

14

16

18

20

01

A--

-

02

D--

-

03

Y--

-

04

N--

-

05

H--

-

06

D--

-

07

F--

-

08

K--

-

09

S--

-

10

A--

-

11

D--

-

12

L--

-

13

I--

-

14

M--

-

15

E--

-

16

I--

-

17

L--

-

18

E--

-

19

B--

-

20

S--

-

21

D--

-

22

J--

-

23

M--

-

24

L--

-

25

N--

-

26

D--

-

27

J--

-

28

G--

-

30

G--

-

32

G--

-

33

F--

-

34

E--

-

35

L--

-

36

R--

-

37

C--

-

39

R--

-

40

C--

-

42

K--

-

Respondent

Su

m o

f s

qu

are

d d

iffe

ren

ces

fro

m

hie

rarc

hic

al

clu

ste

r c

en

trio

d

Cluster 1

Cluster 2

155

Cluster means. Table 6.9 and Figure 6.8 show mean self-confrontation indices for

country of origin broken down by country-of-origin cluster. Independent samples t-tests

were conducted to compare the mean self-confrontation indices for both clusters.

Cluster 1 had higher S than Cluster 2, t(36) = 8.95, p < .001 (two-tailed). Cluster 1 also

had higher O than Cluster 2, t(36) = 7.79, p < .001 (two-tailed). Cluster 1 also had

higher P than Cluster 2, t(36) = 9.69, p < .001 (two-tailed). Finally, Cluster 1 had lower

N than Cluster 2, t(36) = –2.53, p = .016 (two-tailed).

Table 6. 9 Mean self-confrontation indices for Australian I-position, by cluster in Australia

Cluster in

Australia

Index for country of origin I-position

S O P N

Cluster 1 Mean 15.59 14.51 29.91 7.23

(n = 21) SD 1.79 2.24 3.08 4.77

Cluster 2 Mean 9.03 7.48 16.06 12.32

(n = 17) SD 2.37 3.31 4.78 6.56

Note: S and O are scored from 0 to 20; P and N from 0 to 40. S = self-enhancement; O = union-

with-the-other; P = positive affect; N = negative affect.

156

Figure 6.8 Mean self-confrontation indices for Australian I-position, by cluster in

Australia

Note: S and O are scored from 0 to 20; P and N from 0 to 40.

0

5

10

15

20

25

30

35

S-index O-index P-index N-index

Index

Mean

sco

re

Cluster 1

Cluster 2

157

6.3 Analyses of Participant Type

Definition of participant type. Having characterised participants by their cluster

in country of origin and in Australia, a new variable, termed participant “type”, was

defined as the combination of the two cluster values as follows: participant type 1

combines cluster 1 in country of origin with cluster 1 in Australia; participant type 2

combines cluster 1 in country of origin with cluster 2 in Australia; participant type 3

combines cluster 2 in country of origin with cluster 1 in Australia; and participant type 4

combines cluster 2 in country of origin with cluster 2 in Australia. Table 6.10 is a cross-

tabulation of participants by cluster in country of origin and in Australia, where each

cell represents a different participant type.

Table 6.10 Number of participants, by cluster in country of origin and in Australia

Cluster in

country of origin

Cluster in Australia

Cluster 1 Cluster 2 Total

Cluster 1 3 18 21

Cluster 2 11 6 17

Total 14 24 38

A significant association was found between the two sets of clusters, 2(1,

N = 38) = 10.3, p < .001 (two-tailed). Members of Cluster 1 in country of origin were

likely to be in Cluster 2 in Australia and unlikely to be in Cluster 1, and members of

Cluster 2 in country of origin were likely to be in Cluster 1 in Australia and unlikely to

be in Cluster 2. This carries implications for categorising participants. Were the

statistical association complete, participants would only be found on the diagonal,

implying that a mere two participant types would serve to categorise all participants. In

158

fact a quarter of participants lay in “off-diagonal” cells, implying that more participant

types are needed.

Main analyses. Table 6.11 shows means and standard deviations for self-

confrontation indices by I-position location and participant type. “Location” is a

variable that captures the location of the I-position (country of origin or Australia) to

which the self-confrontation indices relate.

A 2-by-4 mixed-design repeated-measures MANOVA was conducted on the self-

confrontation indices as dependent variables, with location as the within-subject

independent variable and participant type as the between-subject independent variable.

159

Table 6.11 Mean self-confrontation indices for the I-position, by location and participant type

Index Participant type

Location

Country of origin Australia

Mean SD Mean SD

Self-enhancement Type 1 (n = 3) 9.19 3.62 14.92 1.88

Type 2 (n = 18) 6.82 2.29 8.52 2.46

Type 3 (n = 11) 13.83 2.09 15.77 1.81

Type 4 (n = 6) 13.03 0.96 10.57 1.27

Union-with-the-other Type 1 (n = 3) 10.47 3.23 14.49 1.01

Type 2 (n = 18) 6.00 2.82 6.82 3.32

Type 3 (n = 11) 13.24 2.30 14.52 2.52

Type 4 (n = 6) 12.92 1.09 9.47 2.55

Positive affect Type 1 (n = 3) 18.23 4.49 27.38 3.85

Type 2 (n = 18) 12.46 4.61 15.07 4.95

Type 3 (n = 11) 26.46 4.18 30.61 2.64

Type 4 (n = 6) 26.04 1.62 19.03 2.81

Negative affect Type 1 (n = 3) 13.25 7.07 6.52 7.67

Type 2 (n = 18) 10.42 6.87 11.01 6.55

Type 3 (n = 11) 9.01 5.94 7.43 4.20

Type 4 (n = 6) 9.14 4.28 16.25 5.25

160

Main effects of type and location on I-position. A MANOVA on self-

confrontation indices revealed a significant multivariate effect of participant type,

Pillai’s Trace = .88, F(12,99) = 3.42, p < .001, partial η2 = .29, and I-position location,

Pillai’s Trace = .49, F(4,31) = 7.55, p = .001, partial η2 = .50.

Student-Newman-Keuls post hoc comparisons (p < .05) showed that Type 2 was

significantly lower on S (M = 7.67) and P (M = 13.77) than Type 4 (M = 11.80 and

M = 22.54 respectively) and Type 1 (M = 12.05 and M = 22.80), and that both these

types were significantly lower than Type 3 (M = 14.80 and M = 28.53). A post hoc

comparison showed that Type 2 was significantly lower on O (M = 6.41) than Type 4

(M = 11.19), Type 1 (M = 12.48) and Type 3 (M = 13.88). A post hoc comparison

showed no significant differences between types on N.

These main effects were qualified by a significant interaction between location

and type, Pillai’s Trace = .82, F(12,99) = 3.13, p = .001, partial η2 = .28. Simple effects

analyses were conducted comparing self-confrontation indices among the four types at

each I-position location.

161

Interactions of type and location on I-position: self-enhancement. Univariate

comparisons showed significant differences between participant types on self-

enhancement (S). Figure 6.9 compares mean S for country of origin and Australia.

Although S differed significantly according to the location of the I-position,

F(3,34) = 15.10, p < .001, partial η2 = .57, the effect differed by participant type.

A Student-Newman-Keuls post hoc comparison (p < .05) showed that Type 2

(M = 6.82) and Type 1 (M = 9.19) were significantly lower on S for country of origin

than Type 4 (M = 13.03) and Type 3 (M = 13.83). A second Student-Newman-Keuls

post hoc comparison (p < .05) showed that Type 2 (M = 8.52) and Type 4 (M = 10.57)

were significantly lower on S for Australia than Type 1 (M = 14.92) and Type 3

(M = 15.77).

Figure 6.9 Mean self-enhancement for country of origin and Australian I-position, by

participant type

0

2

4

6

8

10

12

14

16

18

Origin Australia

Location

S-i

nd

ex

Type 1

Type 2

Type 3

Type 4

162

Interactions of type and location on I-position: union-with-the-other.

Univariate comparisons showed significant differences between participant types on

union-with-the-other (O). Figure 6.10 compares mean O for country of origin and

Australia. Although O differed significantly according to the location of the I-position,

F(3,34) = 9.06, p < .001, partial η2 = .44, the effect differed by participant type.

A Student-Newman-Keuls post hoc comparison (p < .05) showed that Type 2

(M = 6.00) was significantly lower on O for country of origin than Type 1 (M = 10.47),

Type 4 (M = 12.92) and Type 3 (M = 13.24). A second Student-Newman-Keuls post

hoc comparison (p < .05) showed that Type 2 (M = 6.82) and Type 4 (M = 9.47) were

significantly lower on O for Australia than Type 1 (M = 14.49) and Type 3 (M = 14.52).

Figure 6.10 Mean union-with-the-other for country of origin and Australian I-position,

by participant type

0

2

4

6

8

10

12

14

16

Origin Australia

Location

O-i

nd

ex

Type 1

Type 2

Type 3

Type 4

163

Interactions of type and location on I-position: positive affect. Univariate

comparisons showed significant differences between participant types for positive affect

(P). Figure 6.11 compares mean P for country of origin and Australia. Although P

differed significantly according to the location of the I-position, F(3,34) = 21.89,

p < .001, partial η2 = .66, the effect differed by participant type.

A Student-Newman-Keuls post hoc comparison (p < .05) showed that Type 2

(M = 12.46) was significantly lower on P for country of origin than Type 1 (M = 18.23),

which was significantly lower on P for country of origin than Type 4 (M = 26.04) and

Type 3 (M = 26.46). A second Student-Newman-Keuls post hoc comparison (p < .05)

showed that Type 2 (M = 15.07) and Type 4 (M = 19.03) were significantly lower on P

for Australia than Type 1 (M = 27.38) and Type 3 (M = 30.61).

Figure 6.11 Mean positive affect for country of origin and Australian I-position, by

participant type

0

5

10

15

20

25

30

35

Origin Australia

Location

P-i

nd

ex

Type 1

Type 2

Type 3

Type 4

164

Interactions of type and location on I-position: negative affect. Univariate

comparisons showed significant differences between participant types for negative

affect (N). Figure 6.12 compares mean N for country of origin and Australia. Although

N differed significantly according to the location of the I-position, F(3,34) = 10.14,

p < .001, partial η2 = .47, the effect differed by participant type.

A Student-Newman-Keuls post hoc comparison (p < .05) showed no significant

differences on N for country of origin. A second Student-Newman-Keuls post hoc

comparison (p < .05) showed that the subset comprising Type 1 (M = 6.52), Type 3

(M = 7.42) and Type 2 (M = 11.01) was significantly lower on N in Australia than the

subset comprising Type 2 (M = 11.01) and Type 4 (M = 16.25).

Figure 6.12 Mean negative affect for country of origin and Australian I-position, by

participant type

0

2

4

6

8

10

12

14

16

18

Origin Australia

Location

N-i

nd

ex

Type 1

Type 2

Type 3

Type 4

165

Type by location interaction for self-confrontation indices. Figure 6.13 shows

mean self-confrontation indices for both the country of origin and Australian I-positions

broken down by participant type (“Difference in index” is defined as the index for the

Australian I-position less the analogous index for the country of origin I-position).

Paired t-tests were used to compare analogous self-confrontation indices for country of

origin and Australia broken down by participant type. The purpose was to determine

which means were significantly different, and if so, to what degree.

Figure 6.13 Differences in self-confrontation indices for country of origin and

Australian I-positions, by participant type

In the case of participant Type 1, S for Australia (M = 14.92, SD = 1.88) was

significantly higher than for country of origin (M = 9.19, SD = 3.62), t(2) = 5.44,

p = .032 (two-tailed); and P for Australia (M = 27.38, SD = 3.85) was also significantly

higher than for country of origin (M = 18.23, SD = 4.49), t(2) = 7.21, p = .019 (two-

tailed). O and N did not differ significantly by location.

-8

-6

-4

-2

0

2

4

6

8

10

S-index O-index P-index N-index

Index

Dif

fere

nc

e i

n i

nd

ex

Type 1

Type 2

Type 3

Type 4

Solid markers show

significant changes

(p < .05)

166

In the case of participant Type 2, S for Australia (M = 8.52, SD = 2.46) was

significantly higher than for country of origin (M = 6.82, SD = 2.29), t(17) = 3.94,

p = .001 (two-tailed); and P for Australia (M = 15.07, SD = 4.95) was also significantly

higher than for country of origin (M = 12.46, SD = 4.61), t(17) = 3.20, p < .01 (two-

tailed). O and N did not differ significantly.

In the case of participant Type 3, S for Australia (M = 15.77, SD = 1.81) was

significantly higher than for country of origin (M = 13.83, SD = 2.09), t(10) = 2.97,

p = .014 (two-tailed); and P for Australia (M = 30.61, SD = 2.64) was also significantly

higher than for country of origin (M = 26.46, SD = 4.18), t(10) = 3.88, p = .003 (two-

tailed). O and S did not differ significantly.

In the case of participant Type 4, S for Australia (M = 10.57, SD = 1.27) was

significantly lower than for country of origin (M = 13.03, SD = 0.96), t(5) = –9.32,

p < .001 (two-tailed); O for Australia (M = 9.47, SD = 2.55) was significantly lower

than for country of origin (M = 12.92, SD = 1.09), t(5) = –3.87, p = .012 (two-tailed);

and P for Australia (M = 19.03, SD = 2.81) was also significantly lower than for the

country of origin (M = 26.04, SD = 1.62), t(5) = –8.19, p < .001 (two-tailed).

Conversely, N for Australia (M = 16.25, SD = 5.25) was significantly higher than for

country of origin (M = 9.14, SD = 4.28), t(5) = 5.17, p = .004 (two-tailed).

Naming of participant types. Figure 6.14 shows the self-confrontation indices

both in country of origin and in Australia relating to each participant type. Each of the

four participant types was accorded a mnemonic name on the basis of its pattern of self-

confrontation indices and the differences between valuations pertaining to the country

of origin I-position and to the Australian I-position. Since self-enhancement, union-

with-the-other, and positive affect were highly correlated, and negatively correlated

167

with negative affect, participant types were described by whether their S, O, and P were

relatively higher or lower for country of origin as compared with Australia.

Cluster 1 in Australia Cluster 2 in Australia

Type 1 (Progressive) Type 2 (Low-stable)

Clu

ster

1 i

n

cou

ntr

y o

f ori

gin

n = 3 n = 18

Type 3 (High-stable) Type 4 (Regressive)

Clu

ster

2 i

n

cou

ntr

y o

f ori

gin

n = 11 n = 6

Note: Solid markers show significant differences (p < .05).

Figure 6.14 Mean self-confrontation indices, by participant type

Participant type 1: “Progressive”. For this group (n = 3; 8%) the country of

origin I-position was characterised by a low S, O and P, and high N; but the Australian

I-position by high S, O and P, and a low N. When the two I-positions were compared,

the Australian was found to have significantly higher S and P. In summary, these are

0

5

10

15

20

25

30

S index O index P index N index

Origin

Australia

0

5

10

15

20

25

30

S index O index P index N index

Origin

Australia

0

5

10

15

20

25

30

S index O index P index N index

Origin

Australia

0

5

10

15

20

25

30

S index O index P index N index

Origin

Australia

168

people whose Australian I-position is perceived in a more positive light than their

country of origin I-position.

In Gergen & Gergen’s (1986) typology, a “progressive” narrative links events “in

such a way that one steadily progresses towards a goal” (p. 27). Consequently this

group is termed “Progressive” as their well-being is higher for their Australian I-

position, suggesting greater fulfilment of their S and O motives through migration.

Participant type 2: “Low-stable”. For this group (n = 18; 47%) the country of

origin I-position was characterised by low S, O and P, and high N; and the Australian I-

position also by low S, O and P, and high N. When the two I-positions were compared,

the Australian was found to have significantly higher S and P, though the differences

were small. In summary, these are people who perceive both their I-positions more

negatively than do the other participant types.

In Gergen & Gergen’s (1986) typology, a “stability” narrative links events “in

such a way that the protagonist remains essentially unchanged with respect to evaluative

position” (p. 27). Consequently this group is termed “Low-stable” as their well-being

remains low for their Australian I-position, suggesting largely unchanged fulfilment of

their S and O motives through migration.

Participant type 3: “High-stable”. For this group (n = 11; 29%) the country of

origin I-position was characterised by high S, O and P, and low N; and the Australian I-

position also by high S, O and P, and low N. When the two I-positions were compared,

the Australian was found to have significantly higher self S and P, though the

differences were small. In summary, these are people who perceive both their I-

positions more positively than do the other participant types.

In Gergen & Gergen’s (1986) typology, a “stability” narrative links events “in

such a way that the protagonist remains essentially unchanged with respect to evaluative

169

position” (p. 27). Consequently this group is termed “High-stable” as their well-being

remains high for their Australian I-position, suggesting largely unchanged fulfilment of

their S and O motives through migration.

Participant type 4: “Regressive”. For this group (n = 6; 16%) the country of

origin I-position was characterised by high S, O and P, and low N; but the Australian I-

position by low S, O and P, and high N. When the two I-positions were compared, the

Australian was found to have significantly lower S, O and P, and higher N. In summary,

these are people whose Australian I-position is perceived in a more negative light than

their country of origin I-position.

In Gergen & Gergen’s (1986) typology, a “regressive” narrative is one in which

one “is continuously moving away from the valued state” (p. 27). Therefore their self-

narrative should reflect relative decrease in well-being, not only by location but also

through time. Consequently this group is termed “Regressive” as their well-being is

lower for their Australian I-position, suggesting reduced fulfilment of their S and O

motives through migration.

Patterns of change. Two overall patterns—one absolute, one relative—are

observed when comparing country of origin and Australian I-positions. In absolute

terms the Australian I-positions of most participant types—84% of the sample—had a

higher S and P (though not O) than their country of origin I-positions; and one

participant type, Type 4 (Regressive), showed the opposite. But in relative terms the

Australian I-positions of most participants—the two “stable” types, totalling 76% of the

sample—stayed relatively the same; that is, the ranking of S and P for country of origin

and Australia, as compared with those of other participant types, did not differ.

170

6.4 Comparison of Participant Types

This section examines how the four participant types differed on demographic

characteristics, on measures of psychological distress and of well-being, and on the self-

confrontation indices.

Demographic characteristics. Table 6.12 shows the nominal demographic

variables cross-tabulated by participant type. Chi-square analysis revealed no

differences across participant types for gender, χ2(3, N = 38) = .49, p = .92 (two-tailed),

marital status, χ2(6, N = 38) = 2.47, p = .87 (two-tailed), education, χ

2(12,

N = 38) = 8.46, p = .75 (two-tailed), or language, χ2(6, N = 38) = 9.46, p = .15 (two-

tailed). However, participant types did differ on ethnic identification, χ2(3,

N = 38) = 13.11, p < .01 (two-tailed). Those who identified with their ethnic group were

more likely to be found among Type 1 (Progressive) (38% as compared to an expected

8%) and less likely to be among Type 4 (Regressive) (0% as compared to an expected

16%). Thus it seems that ethnic identification predisposes to higher S and O, and higher

P for the Australian I-position (though it is noted that this finding is based on a very

small sample).

171

Table 6.12 Tabulation of demographic characteristics, by participant type

Participant type Total

Type 1

(Progressi

ve)

Type 2

(Low-

stable)

Type 3

(High-

stable)

Type 4

(Regressi

ve)

N %

Gender Male 1 6 5 2 14 37%

Female 2 12 6 4 24 63%

Marital

status

Single 5 3 2 10 26%

Married 3 10 6 3 22 58%

Divorced 3 2 1 6 16%

Education Before yr 10 1 1 2 5%

Year 10 1 1 3%

Year 12 2 1 1 4 11%

Undergraduate 5 3 1 9 24%

Postgraduate 2 9 7 4 22 58%

Language

at home

English only 1 8 5 1 15 39%

Bilingual 2 1 2 1 6 16%

Mother tongue 9 4 4 17 45%

Ethnic

identifica-

tion

Belongs to

group 3 3 2 8 21%

No belonging 15 9 6 30 79%

Total N 3 18 11 6 38

% 8% 47% 29% 16%

Table 6.13 shows the continuous demographic variables cross-tabulated by

participant type. ANOVA revealed no differences across participant types for years in

Australia, F(3,34) = 2.34, p = .09, age, F(3,34) = 1.26, p = .30, and individualism,

F(3,34) = .31, p = .82.

172

Table 6.13 Mean and standard deviation of demographic characteristics, by participant type

Type

Type 1

(Progressi

ve)

Type 2

(Low-

stable)

Type 3

(High-

stable)

Type 4

(Regressi

ve)

All

Years in Australia Mean 31.0 16.4 19.6 24.3 19.8

Standard deviation 11.0 10.6 8.9 9.6 10.5

Age Mean 50.7 39.4 42.8 42.5 41.8

Standard deviation 7.5 9.9 8.8 11.3 9.8

Individualism Mean 42.3 44.3 39.8 35.0 41.4

Standard deviation 31.7 31.7 31.7 31.7 31.7

Self-report measures of distress and well-being.

Psychological distress. Table 6.14 shows means and standard deviations for the

measures of psychological distress by participant type. Participant types did not differ in

their psychological distress: a single-factor between-subjects MANOVA with

participant type as the independent variable was non-significant, Wilks’ λ = .738,

F(3,32) = 1.150, p = .339. Univariate comparisons also showed no significant

differences in SAD, F(3,32) = 1.02, p = .40; in FNE, F(3,32) = 1.76, p = .17; or in BDI,

F(3,32) = 1.02, p = .40.

173

Table 6.14 Mean psychological distress, by participant type

Participant type

Psychological distress

SAD FNE BDI–II

Type 1 (Progressive) Mean 2.00 10.00 11.33

(n = 3) SD 1.73 10.15 9.07

Type 2 (Low-stable) Mean 9.11 15.72 10.61

(n = 18) SD 7.66 10.45 11.65

Type 3 (High-stable) Mean 7.82 8.45 4.91

(n = 11) SD 6.15 6.39 4.74

Type 4 (Regressive) Mean 10.33 16.33 10.83

(n = 6) SD 8.73 9.33 7.33

N = 38

Planned comparisons were performed to compare each participant type with every

other participant type in respect of the three psychological distress measures. In respect

of SAD, Type 1 differed significantly from Type 2 (t = –3.45, df = 16, p < .01), and

from Type 3 (t = –2.76, df = 12, p < .05). Type 1 achieved a borderline difference from

Type 4 (t = –2.25, df = 6, p = .067)—a consequence, no doubt, of these groups’ small

sample sizes. No other planned comparisons were significant. This indicates that Type 1

(Progressive) participants scored generally lower on SAD than the other types. In

respect of FNES, Type 2 differed significantly from Type 3 (t = 2.32, df = 27, p < .05):

Type 2 (Low-stable) participants scored higher than Type 3 (High-stable). No

significant differences were found between participant types on BDI-II. In conclusion,

the incongruent results between the mean comparisons with the t-tests for the Type 4

(Regressive) participants may well be attributed to the small sample size.

Psychological well-being. Table 6.15 shows means and standard deviations for

the measures of psychological well-being by participant type. A single-factor between-

174

subjects MANOVA with participant type as the independent variable was conducted

and found to be marginally significant, Wilks’ λ = .40, F(18,83) = 1.77, p = .044.

Univariate comparisons showed that the difference between participant types was

confined to PWB Positive relations with others, F(3,32) = 4.64, p = .008.

A post hoc comparison showed that Type 4 (Regressive) scored significantly

lower on PWB Positive relations with others than did Type 1 (Progressive) and Type 3

(High-stable) (Student-Newman-Keuls, p < .05). The mean score for participants in

Type 2 (Low-stable) fell between the means for Type 4 (Regressive) and Type 1

(Progressive), and between the means for Type 4 (Regressive) and Type 3 (High-

stable), but was not significantly different from them.

Table 6.15 Mean psychological well-being, by participant type

Participant type

Psychological Well-being Scale

PWB1 PWB2 PWB3 PWB4 PWB5 PWB6

Type 1 (Progressive) Mean 60.00 75.00 64.33 76.33 63.67 62.33

(n = 3) SD 8.66 7.00 9.50 4.16 7.64 10.97

Type 2 (Low-stable) Mean 59.44 63.94 59.50 70.89 62.33 62.17

(n = 18) SD 13.26 12.32 12.66 9.00 11.60 12.68

Type 3 (High-stable) Mean 71.00 75.00 70.73 76.18 73.09 71.45

(n = 11) SD 10.71 8.92 10.48 6.91 9.12 12.47

Type 4 (Regressive) Mean 61.17 57.83 61.00 69.67 65.83 60.17

(n = 6) SD 5.49 8.47 8.05 6.44 7.14 9.11

N = 38. Note: PWB1 = Environmental mastery; PWB2 = Positive relations with others;

PWB3 = Autonomy; PWB4 = Personal growth; PWB5 = Purpose in life; PWB6 = Self-

acceptance.

175

Self-confrontation indices.

Country of origin. Table 6.16 shows means and standard deviations for the

country of origin I-position by participant type. A single-factor between-subjects

MANOVA with participant type as the independent variable was significant, Wilks’

λ = .21, F(12,82) = 5.49, p < .001. Univariate comparisons revealed significant

differences across participant types for S, F(3,34) = 27.73, p < .001); for O,

F(3,34) = 23.62, p < .001; and for P, F(3,34) = 32.25, p < .001; but not for N,

F(3,34) = .42, p = .74.

Table 6.16 Mean self-confrontation indices for country of origin I-position, by participant type

Participant type

Self-confrontation indices

S O P N

Type 1 (Progressive) Mean 9.19 10.47 18.23 13.25

(n = 3) SD 3.62 3.23 4.49 7.07

Type 2 (Low-stable) Mean 6.82 6.00 12.46 10.42

(n = 18) SD 2.29 2.82 4.61 6.87

Type 3 (High-stable) Mean 13.83 13.24 26.46 9.01

(n = 11) SD 2.09 2.30 4.18 5.94

Type 4 (Regressive) Mean 13.03 12.92 26.04 9.14

(n = 6) SD 0.96 1.09 1.62 4.28

A post hoc comparison showed that Type 1 (Progressive) and Type 2 (Low-

stable) scored significantly lower on S than Type 3 (High-stable) and Type 4

(Regressive) (Student-Newman-Keuls, p < .05). A post hoc comparison showed that

Type 2 (Low-stable) was significantly lower on O than Type 1 (Progressive), Type 3

(High-stable) and Type 4 (Regressive) (Student-Newman-Keuls, p < .05). A post hoc

176

comparison showed that Type 2 (Low-stable) scored significantly lower on P than did

Type 1 (Progressive), and that both of these types scored significantly lower than

Type 3 (High-stable) and Type 4 (Regressive) (Student-Newman-Keuls, p < .05).

Australia. Table 6.17 shows means and standard deviations for the Australian I-

position by participant type. A single-factor between-subjects MANOVA with

participant type as the independent variable was significant, Wilks’ λ = .18,

F(12,82) = 6.24, p < .001. Univariate comparisons showed significant differences across

participant types for S, F(3,34) = 30.22, p < .001; for O, F(3,34) = 18.63, p < .001; for

P, F(3,34) = 36.58, p < .001, and for N, F(3,34) = 3.46, p < .05.

Table 6.17 Mean self-confrontation indices for the Australian I-position, by participant type

Participant type

Self-confrontation indices

S O P N

Type 1 (Progressive) Mean 14.92 14.49 27.38 6.52

(n = 3) SD 1.88 1.01 3.85 7.67

Type 2 (Low-stable) Mean 8.52 6.82 15.07 11.01

(n = 18) SD 2.46 3.32 4.95 6.55

Type 3 (High-stable) Mean 15.77 14.52 30.61 7.43

(n = 11) SD 1.81 2.52 2.64 4.20

Type 4 (Regressive) Mean 10.57 9.47 19.03 16.25

(n = 6) SD 1.27 2.55 2.81 5.25

A post hoc comparison showed that Type 2 (Low-stable) and Type 4 (Regressive)

scored significantly lower on S than did Type 1 (Progressive) and Type 3 (High-stable)

(Student-Newman-Keuls, p < .05). A post hoc comparison showed that Type 2 (Low-

stable) and Type 4 (Regressive) were significantly lower on O than were Type 1

177

(Progressive) and Type 3 (High-stable) (Student-Newman-Keuls, p < .05). A post hoc

comparison showed that Type 2 (Low-stable) and Type 4 (Regressive) scored

significantly lower on P than did Type 1 (Progressive) and Type 3 (High-stable)

(Student-Newman-Keuls, p < .05). A post hoc comparison showed that Type 1

(Progressive) and Type 3 (High-stable) scored significantly lower on N than did Type 4

(Regressive) (Student-Newman-Keuls, p < .05). The mean score for participants in

Type 2 (Low-stable) fell between the means for Type 1 (Progressive) and Type 3 (High-

stable) and the mean for Type 4 (Regressive) but was not significantly different.

6.5 Summary

In Study 1, participants were first characterised in terms of demographic

descriptors, psychological self-reports and self-confrontation indices. Cluster analysis

was then employed as a data reduction technique. The participants were grouped into

four types: Progressive, Low-stable, High-stable and Regressive. Following that, mixed-

design repeated-measures MANOVAs were conducted on the self-confrontation indices

as dependent variables, with location of the I-position (country of origin or Australia) as

the within-subject independent variable and participant type as the between-subject

independent variable.

Because the main effects were qualified by significant interactions between

location and participant type, separate paired t-tests were conducted to compare self-

confrontation indices pertaining to each participant’s I-position in country of origin with

those pertaining to the I-position in Australia. Having established that participants could

be grouped in types, and that the types differed in terms of self-confrontation indices,

MANOVA was used to explore where and how the participant types differed in terms of

demographics and psychological self-reports.

178

Perhaps the most significant finding concerned the stability of I-positions.

Participants belonging to Types 2 (Low-stable) and 3 (High-stable), together comprising

76% of the sample, evinced little difference between their Australian and country-of-

origin I-positions. True, their self-enhancement (S) and positive affect (P) were

significantly higher for Australia than for country of origin, but only marginally so.

Only Types 1 (Progressive) and 4 (Regressive), together comprising 14% of the sample,

showed much difference.

Another finding concerned the self-confrontation indices of participant types.

Participants belonging to Types 1 (Progressive), 2 (Low-stable) and 3 (High-stable),

together comprising 84% of the sample, showed on average more self-enhancement and

positive affect for their Australian valuations than for their country-of-origin valuations.

Type 4 (Regressive), comprising 16% of the sample, was the only group not to

benefit in this way: not only did they experience less self-enhancement and union-with-

the-other motives but their well-being declined as a result of reduced positive affect and

increased negative affect. Interestingly, for Types 1, 2 and 3 the self-enhancement

motive (S) was significantly higher for Australia than for country of origin but the

union-with-the-other motive (O) was not significantly different.

A third finding concerned differences in psychological distress and well-being. As

measured by SAD, Type 1 (Progressive) participants were in general less socially

anxious than the other types, though the difference from Type 4 (Regressive) was of

borderline significance due to the smallness of the sample. As measured by FNE, Type

2 (Low-stable) participants were more fearful of negative evaluation than Type 3 (High-

stable). Again, the smallness of the sample appears to have prevented significance

differences being found where otherwise they would have been expected: between, say,

179

Type 1 (Progressive) and Type 4 (Regressive). Lastly, no significant differences were

found between participant types on depression as measured by BDI-II.

Turning now to psychological well-being, significant differences were found in

one of these: Positive relations with others. Type 4 (Regressive) participants had

significantly worse ‘relations with others’ than did Type 1 (Progressive) and Type 3

(High-stable). Type 2 (Low-stable) participants on average rated above Type 4

(Regressive) and below Type 1 (Progressive) and Type 3 (High-stable), but the

differences were non-significant.

In conclusion, this chapter documents the first of three studies performed as part

of thesis. Like the second study, which is discussed next, it is nomothetic in nature, but

differs from it in the way the data were organised at the case level.

180

CHAPTER 7: Study 2 – The Dialogical Self as a Probability Distribution

This chapter discusses the second study conducted as part of this thesis. (A

version of this material has appeared in the International Journal for Dialogical Science

(Sanchez-Rockliffe & Symons, 2010.) It characterises the dialogical self as a

hypothetical probability distribution of themes of valuations, wherein the self-

confrontation method is a technique for generating a sample of themes from which the

properties of this distribution can be inferred, as in standard sampling theory. Since the

dialogical self emerges from a probability distribution of valuation themes, a sampling

interpretation is a fruitful way to view this technique. It was in this context that

hypotheses concerning stability of the self were developed and formally tested.

Study 2 corroborated Study 1 in a broad sense. Unlike Study 1, which used cluster

analysis to group individuals, Study 2 classified them by comparing their valuations in

their country of origin and Australia. But despite the difference in analytical approach,

Study 2 also found differences in migration experience that in key ways resembled

those of Study 1. In particular, both studies identified broadly similar patterns of

stability and change in I-positions.

7.1 The Dialogical Self as a Probability Distribution of Themes

This section presents a conceptual model of the dialogical self that is grounded in

statistical sampling theory. Such an approach is uniquely appropriate because the

dialogical self is by nature not one thing but many, and so demands measurement

methods that recognise the fact. Unlike most psychological variables, which describe a

single characteristic of the individual, the dialogical self describes a multitude of semi-

autonomous selves and their inter-relationships (Hermans & Kempen, 1993; Hermans &

Hermans-Jansen, 1995). In this sense, therefore, it is more like a community than an

181

individual. Communities are conveniently described not by considering a single

member but by considering a sample and expressing the results as frequency

distributions and correlations. This approach makes clear the basis by which statistically

testable hypotheses are generated.

In this model, every individual can generate a potentially infinite number of

valuations, and it is through these valuations that the individual’s dialogical self is

expressed. Each valuation furthermore has a “theme”, by which is meant its vector of S,

O, P, N variables as measured by the SCM. It follows therefore that the dialogical self is

described by a hypothetical multidimensional probability distribution of valuation

themes.

Graphical representation. Figure 7.1 shows the probability distribution of

themes for a hypothetical individual. Each point on the horizontal plane denotes a

particular theme, and the height of the graphical surface at each point represents the

probability with which the individual will generate valuations having that theme.

Alternatively the figure can be regarded as showing the relative proportion of themes in

the total set of valuations. The distribution of themes, considered as a complex but well-

defined mathematical entity, is thus a characteristic of the dialogical self. As with any

model, it is a simplification—for instance it does not capture inter-relationships between

I-positions—but it does provide an objective and useful description of the dialogical self

by conveying the relative weight of various themes in the life of the individual.

182

Figure 7.1 Hypothetical probability distribution of valuation themes: oblique view and plan view (inset)

Thematic dimension 2

Fre

qu

en

cy

Thematic dimension 1

Global mean

Mode 1

Mode 3

Mode 2

Mode 3

Mode 2

Mode 1

Note: Because the figure is

necessarily limited to three

dimensions for illustrative purposes,

the four thematic dimensions (S, O,

P and N) have been reduced to the

two horizontal dimensions, and the

third (vertical) dimension represents

the frequency (or alternatively

probability) with which a given self-

narrative theme occurs.

183

The properties of the distribution need to be inferred from a sample of valuation

themes. This sample is provided by the SCM: participants generate a sample of

valuations, which are then scored to establish their themes. The modes of the

distribution are of particular interest as they correspond to recurrent themes in

valuations. For a multi-modal distribution, the mean is less informative than the modes.

It is, however, easy to calculate, and the average thematic value of a set of valuations is

clearly a useful statistic. In general, inferring the detailed structure of a multivariate,

multimodal probability distribution is difficult with modest sample sizes, as here. Hence

a variety of ad hoc procedures are employed in the current research to study these

distributions.

I-positions. The I-position is conceived of as the hypothetical author of a

valuation. This study focuses on the themes of valuations, as it is these that are directly

measurable. Participants were required to perceive themselves as having two separate

cultural identities—that is, they should think of themselves as having achieved bi-

cultural identification, one identity belonging to their country of origin and the other

identity belonging to Australia. Later, in the SCM, participants were invited to focus

first on their cultural identity of their country of origin, then on their cultural identity in

Australia, and to tell about important experiences and circumstances from these

particular perspectives.

Thematic modes. Themes that regularly occur in valuations are of particular

interest, as they reflect psychological invariance over I-positions. These valuations will

be revealed as modes of the probability distribution of themes (see Figure 6.18). In a

sample of valuations, clusters of roughly similar theme appear as modes. Strongly

recurrent themes will tend to produce larger clusters and thus larger hills around the

184

mode. Figure 7.1 is drawn with three modes. In this case three sets of similar themes

would tend to re-occur in the valuations of the particular individual.

Themes and migration. This section explores what happens to the distribution of

themes when a person experiences a dramatic change in life circumstances, as for

example occurs after migration. A requirement of the study was that participants

acknowledge both an Australian identity and a continued identity associated with the

country of origin. This implies that living as a migrant brings into being new I-

positions. However, it may be that these new I-positions are thematically similar to

corresponding I-positions in the country of origin, at least in some cases. For instance,

the themes of domestic I-positions (I-as-parent, I-as-spouse etc.) might change little in a

migrant who brings his or her family, but those for work I-positions might change

greatly. The probability surface would then shift, more in some places than others. The

study gauged the extent to which migration changed the thematic distribution for this

sample.

7.2 Hypotheses

The aim of Study 2 was to explore how the dialogical self is affected by a radical

change in the social environment, in particular by migration to a new country. However,

since the temporal stability of the self is one of its abiding characteristics (James, 1890),

an explanation is required as to how such change might occur. According to Hermans

(1987a, b, 1988, 1992), valuations have both latent and manifest content. The latent

content is represented by the S and O motives; the manifest, by the P and N affects. In

this theoretical framework, the S and O motives represent deep psychological drives; the

P and N affects represent the extent to which these motives are gratified in a particular

social environment, and hence manifest as affects.

185

For an individual placed in two different environments, one can therefore consider

four possibilities.

Hypothesis 1: Latents constant, manifests constant. This is the case of complete

thematic stability. It would occur when the individual’s fundamental motives are

constant and each environment offers an equivalent opportunity for gratification.

Hypothesis 2: Latents constant, manifests change. This is perhaps the most natural

case. The new environment offers different possibilities for gratification but the

individual’s fundamental motives remain the same.

Hypothesis 3: Latents change, manifests constant. This could arise from

psychological development over the life cycle. As an individual progresses through life,

it is natural to expect fundamental motives to change. This may not be accompanied by

affective change.

Hypothesis 4: Latents change, manifests change. In this case the fundamental

motives of the individual change, as does the extent of gratification. This could come

about if the new environment creates new I-positions on the one hand, or simply for

developmental reasons on the other.

7.3 Analysis Based on Major Themes

Two similar but comparable analyses were performed in Study 2. In the first

analysis, participants were grouped on the basis of the profiles of major themes

manifested by their valuations. This was done separately for valuations pertaining to the

country of origin and to Australia. A transition matrix was then generated to determine

which participants were in the same group for both sets of valuations, and which

participants changed groups. Formal tests of statistical significance were then performed

on the transition matrices to test the null hypothesis that participants remained in the

same groupings, that is, that migration produced no significant change in I-position.

186

Thematic nature of I-positions. In Study 1, valuations (and hence participants)

have been characterised in terms of self-confrontation indices. In this study, valuations

were classified according to the six “major themes” defined by Hermans and Hermans-

Jansen (1995) rather than by indices. The rules for classifying valuations are set out in

Table 7.1 and are largely those used by Hermans and Hermans-Jansen (1995, p.254). As

some latitude is permitted, however, the definitions of “high” or “low” self-

enhancement and union-with-the-other were set at 17 and 4 respectively for this study.

This typology is not exhaustive, as valuations can fail to meet the criteria for inclusion

under any of the major themes; by setting the rules in this way, the 1565 valuations

obtained from all migrants were reduced a more manageable 570. The analysis was

based on these.

Table 7.1 Rules for assigning valuations to major themes

Major theme

Criteria for assignment

P – N S – O S O

Strength and unity (+HH) ≥ 12 … > 17 > 17

Unity and love (+O) ≥ 12 ≤ –6 … …

Unfulfilled longing (–O) ≤ –12 ≤ –6 … …

Powerlessness and isolation (–LL) ≤ –12 … < 4 < 4

Aggression and anger (–S) ≤ –12 ≥ 6 … …

Autonomy and success (+S) ≥ 12 ≥ 6 … …

Key: S = Self-enhancement; O = Union-with-the-other; P = Positive affect; N = Negative affect.

Participant types were characterised by the major themes that their valuations

revealed, yielding a thematic profile of each type. Table 7.2 tabulates those valuations

by participant type, by I-position location (that is, country of origin or Australia), and by

major theme. Each row represents the profile of the stipulated I-position of each

187

participant type. Since the absolute number of valuations in each row is arbitrary, it is

the proportion assigned to each theme that is meaningful.

Table 7.2 Number of valuations, by participant type and major theme

Participant

type

Major theme

Strength

and

unity

(+HH)

Unity

and

love

(+O)

Unfulfil-

led

longing

(–O)

Power-

lessness

and

isolation

(–LL)

Aggres-

sion and

anger

(–S)

Auto-

nomy

and

success

(+S)

Total

Type 1

(Progressive)

Origin 6 4 6 7 3 26

Australia 11 1 1 5 18

Total 17 4 6 8 1 8 44

Type 2

(Low-stable)

Origin 4 24 15 53 4 53 153

Australia 9 8 5 39 8 48 117

Total 13 32 20 92 12 101 270

Type 3

(High-stable)

Origin 52 7 7 10 2 28 106

Australia 40 3 1 2 17 63

Total 92 10 8 12 2 45 169

Type 4

(Regressive)

Origin 24 4 3 9 8 48

Australia 4 1 5 8 11 10 39

Total 28 5 8 17 11 18 87

Grand total 150 51 42 129 26 172 570

Procedure for grouping participants. Participants were classified according to

the following procedure:

Participants were represented by the proportions of all valuations in each major

theme. Two profiles were produced for each participant: one pertaining to all valuations,

termed the “global profile”, and one for valuations at country of origin.

188

Proportions were regarded as “important” if they were 20% or more of the total

number of valuations; otherwise they were regarded as unimportant.

Participants were assigned to a common group if they had important proportions

in precisely the same categories.

For this purpose, an important proportion in strength and unity (+HH) was

regarded as equivalent to important proportions in both autonomy and success (+S) and

unity and love (+O).

Likewise, an important proportion in powerlessness and isolation (–LL) was

regarded as equivalent to important proportions in both aggression and anger (–S) and

unfulfilled longing (–O).

Table 7.3 shows the resulting classification (names of participants in this table and

subsequently were changed to preserve anonymity). Each row of the table gives the

proportion of each participant’s valuations expressing each major theme. Proportions

pertaining “important” major themes are bolded and underlined. The resulting vector of

proportions is termed the participant’s “thematic profile”.

Ideally the country-of-origin profile for each participant would have been

compared with the corresponding Australian one. This was not done because some

participants had too few Australian valuations with major themes (two or fewer) to

render comparison meaningful; they would have had to be dropped from the sample. By

comparing the country-of-origin profile with the global profile the problem was

mitigated, but at the cost of creating another one: the null hypothesis was now harder to

reject since the global profile was contaminated with country-of-origin valuations. In

the event this did not matter as this null was still rejected, for rejection of the one

hypothesis entails rejection of the other. Under this approach only nine of the 38 in the

sample needed to be dropped, limiting the test sample to 29.

189

On the basis of this table, participants were classified into two groups: “univalent”

and “multivalent”. Univalent participants are those who expressed only a single affect,

either positive or negative (but not both) in their valuations. Multivalent participants

were those who expressed a mix of both positive and negative affects. The distinction is

crucial as it is the basis for determining the stability of thematic profiles. Multivalent

participants are of particular interest because they have unambiguously multimodal

major themes and hence exhibit instability in their thematic profiles. These individuals

are thus much in line with Hermans’ view of the dialogical self. It is noteworthy that

this instability is not merely an artefact of migrants’ differential states of mind at origin

and destination: of the 15 multivalent participants, 12 were multivalent at country of

origin.

Of the univalent respondents, one expressed only –O; three expressed both –S and

–O; six expressed only +S; and 13 expressed both +S and +O. Of the multivalent

respondents, three expressed a mix of most major themes (+S, +O, –S and –O), two

expressed +S but also –LL, two expressed +O but also –LL, one expressed +HH but

also –O, one expressed +HH but also –S, and six expressed +S but also –LL. The three

most common profiles are those whose valuations expressed only autonomy and success

(+S); those that expressed both motives (S and O), both with positive affect; and those

that expressed autonomy and success (+S), but combined with powerlessness and

isolation (–LL).

To judge whether country-of-origin profiles were importantly different from

global profiles, participants were subjectively classed as either “stable” or “unstable” on

the basis of the degree of difference between their respective profiles. Of the 29

participants classed in this way, 23 were considered stable and six unstable.

190

Table 7.3

Thematic profiles showing “important” major themes, by participant and location

All valuations

Valuations at country of origin

Positive affect Negative affect Positive affect Negative affect Stability Distance

+HH +S +O

–LL –S –O

+HH +S +O

–LL –S –O

Univalent

–O only 17 Ljubica

0.10 0.20

0.10

0.60

0.10 0.20

0.10

0.60

X n.a.

–S and –O 07 Frida 0.11 0.06 0.17

0.33 0.33 0.00

0.10 0.10 0.10

0.50 0.20

S 0.23

03 Yoana 0.20 0.10 0.10

0.40

0.20

0.17

0.50

0.33

U 0.29

02 Dragan

0.04

0.86 0.04 0.07

1.00

S 0.17

+S only 28 Gonzalo 0.11 0.50 0.07

0.18

0.14

0.05 0.53 0.11

0.11

0.21

S 0.12

15 Etsuko

1.00

1.00

X n.a.

09 Sohrab

1.00

1.00

X n.a.

39 Rhonda 0.05 0.76

0.19

0.92

0.08

S 0.19

30 Germaine

0.53 0.18

0.18 0.12

0.50 0.10

0.30 0.10

S 0.15

36 Robert 0.13 0.75

0.13

0.13 0.73

0.13

S 0.02

191

All valuations

Valuations at country of origin

Positive affect Negative affect Positive affect Negative affect Stability Distance

+HH +S +O

–LL –S –O

+HH +S +O

–LL –S –O

+S and +O 13 Irene 0.11 0.21 0.53

0.11

0.05

0.00 0.18 0.73

0.09

S 0.24

34 Emilia 0.50

0.13

0.17 0.04 0.17

0.35

0.18

0.24

0.24

S 0.19

20 Sancho 0.56 0.19

0.13

0.13

0.67 0.08 0.00

0.17

0.08

S 0.16

24 Leo 1.00

X n.a.

14 Maryam 0.93 0.03

0.03

1.00

S 0.08

32 Griselda 0.50 0.25 0.25

0.50 0.25 0.25

X n.a.

27 Janet 0.33 0.33 0.17

0.08

0.08

0.00 0.40 0.40

0.20

U 0.44

08 Kate 0.73 0.27

0.71 0.29

S 0.02

10 Alicia 0.67 0.22

0.11

0.50

0.50

X n.a.

12 Laura 0.57 0.22

0.17 0.00 0.04

0.47 0.26

0.21

0.05

S 0.11

05 Hendrik 0.56 0.22

0.11 0.11

0.60

0.20 0.20

U 0.26

35 Luisa 0.52 0.29 0.16

0.03

0.39 0.39 0.17

0.06

S 0.16

01 Azzam 0.30 0.70

1.00

U 0.42

192

All valuations

Valuations at country of origin

Positive affect Negative affect Positive affect Negative affect Stability Distance

+HH +S +O

–LL –S –O

+HH +S +O

–LL –S –O

Multivalent

S and O, and

both positive

and negative affect

19 Beatriz 0.25 0.31 0.13

0.25

0.06

0.29 0.29 0.14

0.29

S 0.09

23 Marta

0.50 0.25

0.25

0.50 0.50

X n.a.

18 Eva

0.50 0.25

0.25

0.50

0.50

X n.a.

+S and -LL 42 Keira

0.43 0.14

0.36

0.07

0.00 0.22 0.22

0.56

U 0.31

04 Nadine

0.33

0.67

1.00

X n.a.

+O and -LL 25 Nicole 0.06 0.11 0.28

0.50

0.06

0.10

0.40

0.40

0.10

S 0.20

06 Danilo 0.15 0.08 0.23

0.31 0.08 0.15

0.33

0.33 0.11 0.22

S 0.22

+HH and -O 33 Farid 0.31

0.15

0.15

0.38

0.27

0.09

0.18

0.45

S 0.10

+HH and -S 40 Carla 0.37 0.11

0.14 0.29 0.09

0.71 0.06

0.18

0.06

X 0.45

+S and -LL 26 Damon

0.78

0.22

0.67

0.33

S 0.16

193

All valuations

Valuations at country of origin

Positive affect Negative affect Positive affect Negative affect Stability Distance

+HH +S +O

–LL –S –O

+HH +S +O

–LL –S –O

11 Damara

0.72

0.28

0.60

0.40

S 0.17

22 Juan

0.60

0.33

0.07

0.50

0.33

0.17

S 0.14

37 Cristiano

0.55

0.45

0.38

0.63

S 0.24

21 Diana 0.09 0.27

0.45 0.09 0.09

0.07 0.36

0.50

0.07

S 0.13

16 Irina 0.13 0.25 0.19

0.25 0.06 0.13

0.20 0.30 0.30

0.10

0.10

S 0.22

Note: Names have been changed to preserve anonymity.

Key: Bold cells indicate “important” major themes. Stability is coded as: S = stable; U = unstable; X = unclassified.

194

Tests of statistical significance. The null hypothesis was that thematic profiles

are stable as between country of origin and Australia. Tests of the null are based on the

differences between the profiles measured separately with respect to country of origin

and Australia. Due to occasional lack of classified data at origin or in Australia, five

participants in the sample of 38 were perforce dropped from the analysis. Table 7.4

reports averages, standard deviations, and t-statistics for the remaining 33, for the null

hypothesis that mean differences are zero.

For two-tailed tests, only –LL is significant at the 5% level: thus on average,

participants told fewer –LL valuations of Australia. It is, however, notable (though of

borderline significance) that they also told more +HH and +S valuations and fewer +O

valuations of Australia.

Table 7.4 Average differences in proportions of major themes

Major theme M SD t df p

Strength and unity (+HH) 0.08 0.27 1.80 32 0.081

Unity and love (+O) 0.09 0.27 1.92 32 0.064

Unfulfilled longing (–O) –0.05 0.21 –1.54 32 0.133

Powerlessness and isolation (–LL) –0.11 0.29 –2.33 32 0.026

Aggression and anger (–S) 0.03 0.13 1.59 32 0.122

Autonomy and success (+S) –0.04 0.16 –1.42 32 0.165

The question is whether the six average differences are jointly different from zero.

This requires a test of a mean, complicated by the fact that the mean is here a vector in

six dimensions. Such a test can be based on the statistic:

12 )/( NVT

195

where μ is the measured vector of means (of differences), V the covariance matrix

of the sample of differences, and N the sample-size.1

In computing this statistic, account must be taken of the fact that the underlying

variables are proportions that sum to unity. The resulting redundancy is solved by

arbitrarily eliminating one of the six categories: –O was chosen. If this is done the

statistic is distributed as a chi-square with five degrees of freedom under the null of no

differences in means (in large samples).

There is strong evidence against the hypothesis of complete stability of the

thematic profile, 2(5, N = 33) = 26.2, p < .002. In other words, for some individuals at

least, the themes of Australian valuations tended to differ significantly from those of

their country of origin valuations.

7.4 Analysis Based on Self-confrontation Indices

The second analysis of Study 2 resembled the first except that participants were

grouped on the basis of the self-confrontation indices of their valuations—whether

above or below the mean. As with in the first analysis, this was done separately for

valuations pertaining to the country of origin and to Australia. A transition matrix was

then generated to determine which participants were in the same group for both sets of

1 If the five selected variates were normally and independently distributed, this statistic would be the sum

of squares of the t-statistics in Table 3, which is chi-square with five degrees of freedom in large samples.

The presence of the term V/N in the formula is to correct for possible cross-tabulations between the

variates. This approach has the advantage that it follows on naturally from the t-tests. It is thus preferable

to a contingency table analysis, which would be in any case vitiated by low expected frequencies in many

cells.

196

valuations, and which participants changed groups. Formal tests of statistical

significance were then performed on the transition matrices to test the null hypothesis

that participants remained in the same groupings, that is, that migration produced no

significant change in I-position.

Procedure for grouping participants. The preceding analysis was based on

Hermans’ typology of valuation themes. His classification has the advantage of clinical

experience but results in large attrition of valuations since only about one third of

valuations are so classified; diagnostic certainty is gained at the cost of potential loss of

information. The following analysis therefore considers the full set of valuations. A

typology of participants exploiting the full set is required. To reduce dimensionality,

positive and negative affect indices are represented by a single index, well-being (see

Chapter 4).

Next, a version of Hermans’ typology is constructed from each participant’s

average score on the three indices S, O and W, both at country of origin and in Australia.

An index value is considered high (or low) if it is higher (or lower) than the full sample

mean in the country of origin. This gives an eight-way classification of individuals at

the origin and in Australia. Using this typology it is possible to construct an 8 × 8

transition matrix to compare the themes expressed by migrants concerning the country

of origin and Australia. This procedure extends the Hermans classification of valuations

to a classification of individuals.

Note that under this typology it is possible for an individual to score low on S and

O and high on W, as well as high on S and O and low on W. Such cases are ignored in

the Hermans approach but the proposed classification procedure throws up a few

examples, termed +E and –E, respectively. The transition matrix is given in Table 7.5.

Row-totals and column-totals give the numbers of participants classified in each of the

197

eight nominated categories for country-of-origin and Australian valuations respectively.

The values in each cell of the matrix give the number of people who pass from the

corresponding category of country-of-origin valuation to the corresponding category of

Australian valuation. For example, in the top left cell, eight participants pass from +HH

for country-of-origin to +HH for Australia.

Table 7.5

Thematic transitions

Major theme in Australia

Major theme at country of origin +HH +S +O +E -LL -S -O -E Total

Strength and unity (+HH) 8

1 1 1 1 2 14

Autonomy and success (+S) 1 2

3

Unity and love (+O)

1

1

Low S, O, high W (+E) 3

3

Powerlessness and isolation (–LL) 1 1 1

10 1

1 15

Aggression and anger (–S)

0

Unfulfilled longing (–O)

1 1 2

High S, O, low W (–E)

0

Total 13 3 2 1 11 2 2 4 38

Key: Shaded cells show stable participants. Boxed cells show participants whose affects

changed from positive to negative well-being or vice versa.

In respect of their country of origin I-positions, 14/38 = 37% of individuals are

classified as +HH and 15/38 =39 % as –LL, compared to 34% and 29% with regard to

that of Australia, respectively; 21/38 = 55% show positive affect in the country of

origin, 19/38 = 50% show positive affect in Australia. The entries along the diagonal of

the matrix show individuals with stable average themes with respect to country of origin

and Australia. Thus 22/38 = 55% exhibited similar themes in valuations about the

country of origin as about Australia. The bottom left portion of the matrix indicates

198

individuals with negative affect at the country of origin but positive affect in Australia,

3/38 = 8% in total; 6/38 = 16% (the top right portion) were the opposite.

In summary, there was considerable thematic stability across valuations about the

country of origin and Australia. There was some evidence of change, contained in the

off-diagonal portion of the transition matrix. Five participants exhibited positive affect

with regard to Australia and negative affect with regard to the country of origin; three

participants exhibited the opposite. This confirms the general picture emerging from the

preceding section of small but measureable changes in a context of considerable

stability.

Tests of statistical significance. Analogously to Table 7.4, Table 7.6 reports

statistics for the differences in numerical index scores between origin and Australia.

Notably, average theme is higher in Australia along all four dimensions. Thus, the

themes in Australia on average showed stronger sense of self-enhancement (S), stronger

sense of union with the other (O), stronger positive affect (P), and stronger negative

affect (N). The average increase in S and P were significant at the 5% level, the other

two are not. The ensemble test of the four parameters together yields a value of 42.1,

which has a p-level of 1.59×10-8

. This is compelling evidence against the null

hypothesis of unchanged themes.

Table 7.6

Average differences in thematic scores (Australia minus country of origin)

Index M SD t df p

Self-enhancement (S) 1.43 2.67 3.31 32 0.002

Unity and love (O) 0.53 2.91 1.12 32 0.271

Unfulfilled longing (P) 2.05 5.35 2.37 32 0.024

Autonomy and success (N) 0.41 5.17 0.49 32 0.627

199

7.5 Summary

Two separate but comparable analyses were performed in Study 2. In both,

participants were grouped on the basis of the types of valuations they produced, though

according to different rules. One analysis grouped participants according to the

combinations of major themes manifested by their valuations; the other grouped them

according to the self-confrontation indices of their valuations. This was done separately

for valuations pertaining to the country of origin and to Australia. Transition matrices

were then produced to determine which participants were in the same group for both

sets of valuations, and which participants changed groups. Finally, tests of statistical

significance were performed on the transition matrices to test the null hypothesis that

participants remained in the same groupings, that is, that migration produced no

significant change in I-position.

There were two significant findings. Under both analyses, results were as

hypothesised: that is, I-positions were found to differ as between those pertaining to

country of origin and those pertaining to Australia. The other significant finding comes

from Analysis 1 and relates to locational differences in the major themes of valuations.

On average, participants told significantly fewer –LL valuations for Australia than for

country of origin.

In conclusion, this chapter documents the second of three studies performed as

part of thesis. Like the second study, which was discussed in the preceding chapter, it is

nomothetic in nature, but differs from it in the way the data were organised at the case

level. The third study, which is discussed next, was idiographic, analysing case studies

representing each of the participant types identified by the first study.

200

CHAPTER 8: Study 3 – Case Studies

In Study 3, an idiographic approach was adopted to complement the nomothetic

approach of Studies 1 and 2. Four case studies are discussed, one of each migrant type:

Progressive, Low-stable High-stable, and Regressive. The aim was to exemplify and

corroborate the nomothetic principles previously identified. The idiographic analyses in

this chapter furnish a more complete understanding of the individual than do the

nomothetic analyses of the preceding two chapters.

The chapter begins with a discussion of the Euclidean distance between each

participant’s valuation profiles pertaining to country of origin and to Australia. This

assisted in the selection of case studies by showing how they are positioned in relation

to other study participants. By comparing and contrasting pairs of valuations, one from

country of origin and the other from Australia, the case studies illustrate how I-positions

differed, or conversely were stable, between the two locations. Importantly, the case

studies demonstrate at an individual level that while most migrants are stable in terms of

their I-positions, a minority are not.

8.1 Euclidean distance

An objective corroboration of stability of the self was provided by the Euclidean

distance between valuation profiles, defined as the square root of the sum of squared

differences between them (Deza & Deza, 2009). Figures 7.1 and 7.2 both rank

participants by Euclidean distance, shown by the length of the bars along the horizontal

axis. The figures differ in that Figure 8.1 uses self-confrontation indices as the measure

of Euclidean difference, while Figure 8.2 uses major themes. The dark blue and orange

bars in both charts denote participant Type 1 Progressive and Type 4 Regressive, that is,

the two migrant types that exhibited most instability in I-position.

201

An interesting, but unsurprising, finding is that participants of the unstable

types—that is, Types 1 and 4—also tended to exhibit the greatest Euclidean distance

between their profiles for country of origin and Australia. The tendency is more

pronounced in Figure 8.1 than 7.2 because the former is based on self-confrontation

indices, which were directly used to derive the participant types, while the latter is based

on major themes, which are only indirectly linked to self-confrontation indices, and

thence to major themes.

202

Figure 8.1 Euclidean distance between profiles for country of origin and Australia

based on mean self-confrontation indices of valuations

,0 ,2 ,4 ,6 ,8 ,10 ,12

28 Gonzalo

07 Frida

08 Kate

14 Maryam

15 Etsuko

26 Damon

36 Robert

02 Dragan

23 Marta

24 Leo

25 Nicole

33 Farid

30 Germaine

17 Ljubica

06 Danilo

22 Juan

11 Damara

32 Griselda

21 Diana

37 Cristiano

13 Irene

09 Sohrab

42 Keira

39 Rhonda

10 Alicia

12 Laura

35 Luisa

27 Janet

34 Emilia

04 Nadine

20 Sancho

19 Beatriz

16 Irina

01 Azzam

40 Carla

05 Hendrik

18 Eva

03 Yoana

Type 4

Type 3

Type 2

Type 1

203

Figure 8.2 Euclidean distance between profiles for country of origin and Australia

based on proportions of major themes of valuations

,0.0 ,0.2 ,0.4 ,0.6 ,0.8 ,1.0

08 Kate

36 Robert

19 Beatriz

22 Juan

11 Damara

02 Dragan

14 Maryam

30 Germaine

21 Diana

28 Gonzalo

35 Luisa

39 Rhonda

25 Nicole

26 Damon

07 Frida

13 Irene

05 Hendrik

16 Irina

01 Azzam

10 Alicia

12 Laura

20 Sancho

34 Emilia

33 Farid

06 Danilo

18 Eva

23 Marta

04 Nadine

03 Yoana

27 Janet

42 Keira

40 Carla

37 Cristiano

Type 4

Type 3

Type 2

Type 1

204

8.2 Case study: The ‘Progressive’ Migrant (Type 1)

Yoana (this and other names are pseudonyms) is representative of the Progressive

migrant type. She is a Colombian who came to Australia at 23. At the time of the

interview she was married with two young children, a girl and a boy, both still in

primary school. Although she had obtained a degree in journalism in her native country,

she was not working at the time of the interview and she listed her occupation as “house

duties”. Two key points come across from her valuations about her youth in Colombia.

Her parents divorced when she was seven and she felt abandoned by her much-admired

father, as well as sympathy for her mother’s predicament. This sense of abandonment

was compounded when her sister migrated to Spain. The second theme concerns her

distress at the quality of life in Colombia, which she perceives as “a world full of

materialism, alcohol, sex, drugs, violence, etc.” In Australia she has made a happy

marriage. Her valuations emphasise her devotion to her children. She is a convert to the

Baha’i religion and aspires to “a very pure and spiritual life”.

Yoana’s measures of psychological distress and well-being corroborate this

picture of a determined and caring person who does not let adversity get the better of

her. At the time of the interview she was confident in social situations (SAD = 1, cf.

sample mean of 8.4); had little concern at giving a negative impression to others (FNE =

1, cf. sample mean of 13.3); and was not depressed with respect to the sample mean

(BDI-II = 10, cf. sample mean of 9.1). Her psychological well-being was above average

on all subscales except ‘Purpose in life’, where she was slightly below (62 cf. sample

mean of 66.1).

Major themes. Figure 8.3 shows the mean thematic profile of Type 1 participants

(Panel 1) and for Yoana, the case-study selected to exemplify them (Panel 2). The mean

profile of valuations of Type 1 participants differed significantly according to whether

205

they pertained to country of origin or Australia, χ2(5, N = 44) = 16.56, p < .01 (two-

tailed). Valuations located in country of origin were high on the themes of Unfulfilled

longing and Powerlessness and isolation, but those in Australia were high on the

themes of Strength and unity and Autonomy and success.

Yoana’s thematic profiles correlated highly with the mean profiles of Type 1

participants both for country of origin ( r = .67) and for Australian valuations (r = .90),

indicating that she is representative of the Progressive migrant type.

206

Panel 1 All Progressive migrants

Panel 2 Yoana

Figure 8.3 Proportion of valuations, by major theme and location: Progressive migrants

and Yoana

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

207

Self-confrontation indices. Figure 8.4 shows Yoana’s mean self-confrontation

indices. On self-enhancement, her valuations of country of origin scored significantly

less (M = 5.2, SD = 5.5) than those of Australia (M = 13.0, SD = 5.2), t(22) = –3.57,

p < .01 (two-tailed). On union-with-the-other, there was no significant difference

between her valuations of country of origin (M = 9.1, SD = 6.6) and those of Australia

(M = 13.6, SD = 5.8), t(22) = –1.77, p = .09 (two-tailed). On positive affect, her

valuations of country of origin were significantly less (M = 13.7, SD = 12.3) than those

of Australia (M = 25.0, SD = 12.5), t(22) = –2.24, p < .05 (two-tailed). On negative

affect, her valuations of country of origin scored significantly more (M = 13.5, SD =

8.5) than those of Australia (M = 2.5, SD = 4.0), t(22) = 4.07, p < .001 (two-tailed).

Figure 8.4 Mean indices of valuations, by location: Yoana

0

5

10

15

20

25

30

35

40

S-index O-index P-index N-index

Origin

Australia

General

Ideal

208

Yoana is representative of the Type 1 (Progressive) migrant. According to Gergen

& Gergen (1986) a “progressive” narrative links events “in such a way that one steadily

progresses towards a goal” (p. 27). Corroboration of this is found in Figure 8.5, which

shows the temporal dimension of her self-confrontation indices.

Yoana’s S , O and P rise, and her N falls, in her future-oriented country-of-origin

valuations—a clear temporal progression. But no such temporal progression is observed

in her Australian valuations, which are at a uniformly high level. This suggests that

Yoana is basically affectively positive both about Australia (in all periods) and about

the future (in both locations). What makes her “progressive” is the fact that her past-

and present-oriented valuations are relatively affectively negative for her country of

origin but not for Australia.

Figure 8.5 Mean indices of valuations, by location and period: Yoana

0

5

10

15

20

25

30

35

Past Present Future Past Present Future

Origin Australia

S

O

P

N

209

Self-narrative. Yoana’s self-narrative is characterised by progression towards a

desired goal. She espouses the twin goals of mother and spiritual being. The following

contrasting valuations show how she is attaining these goals.

She has turned away from her own mother’s goal of focusing on her husband

(Yoana’s father) and instead focuses on her children:

Country of origin: “… my mother used to see my father as the centre of her

universe” (–LL).

Australia: “In my present existence the most important thing are my children …”

(+HH).

She has successfully repudiated the drinking culture of her country of origin:

Country of origin: “I look at alcohol as a destructive drug” (–O).

Australia: “I was so impressed by the fact that the youth were not drinking alcohol

…” (+HH).

... and has likewise decided to be a spiritual person, and has achieved that aim:

Country of origin: “I decided to live a very pure and spiritual life”.

Australia: “… I closed my eyes and I experience God” (+HH).

... and has successfully fought a familial tendency towards depression:

Country of origin and Australia: “Another influence has been my aunty who

suffered from depression when I was a little girl. When … I suffered my own

depression I decided never to be like her, therefore I had the strength to fight it

and overcome it” (–LL).

Yoana’s is a self-narrative of setting new and very different goals, and achieving

them—truly a progression to better things.

210

Table 8.1 Valuations with major themes, by location and period: Yoana

Valuation S O P N G I Theme

Country of origin: Past

1 A mark in my life is that at that time my mother used to see my father as the

centre of her universe. Now that I am married I refuse to see my husband the

same way. Instead, my kids are the centre of my life.

0 0 0 19 -0.69 -0.78 -LL

2 An episode I remember is my father visiting us. During that time my mother

used to prepare us the best possible way, we had to wear our best clothes and

our best behaviour before he arrived. But to our great disappointment he will

come and sit there for the whole afternoon reading newspapers. He never

talked to us, never took us to the park or cinema, etc. My mother made us sit

there with him, just watching him. From that moment I always experience a

sense of not being good enough.

2 2 3 24 -0.71 -0.78 -LL

4 When I was 14 I clearly remember a scene like a scene from a movie where

we were in a party with very close friends. The son of one of those friends

was only 4 and I recall that alcohol was offered to him and was encouraged to

taste the delicious flavour and to my amazement Leonardo took it. From that

moment I look at alcohol as a destructive drug.

0 13 0 18 -0.30 -0.33 -O

Country of origin: Present

9 In my opinion my extended [Colombian] family is very significant. There

was an episode in my life that confronted me to this reality. My son was

operated on as an unexpected emergency and my daughter was left at school

with no one to look after her while I was with her brother.

0 14 0 18 -0.27 -0.34 -O

10 My children were watching TV and I just learned on the phone that my mum

was flying to Australia to help us during my sickness. The nice surprise was

that my sister told me on the phone that my mother was just one hour away

before she landed in Melbourne.

5 13 26 5 0.48 0.46 +O

11 The same month I happened to be diagnosed with three lumps in my breast.

When I learned about it I was very upset and the doctor had to comfort me.

There I felt more than ever the ongoing importance of having an extended

family.

3 0 2 14 -0.36 -0.46 -LL

Australia: Past

1 When I was 24 here in Melbourne I went to a Youth Baha’i party and there

were boys and girls all about the same age. I was so impressed by the fact

that the youth were not drinking alcohol and yet they were enjoying

themselves and having a good time. It impressed me a lot the clear

environment that surrounded me that night.

13 6 29 0 0.78 0.77 +S

Australia: Present

5 Another influence has been my aunty who suffered from depression when I

was a little girl. When [this year] I suffered my own depression I decided

never to be like her, therefore I had the strength to fight it and overcome it.

0 0 0 13 -0.55 -0.61 -LL

6 Classical music plays a very important role in my life. One day I was

watching my daughter in a concert. She was playing a beautiful piece by

Bach and I closed my eyes and I experience God.

19 19 35 0 0.90 0.96 +HH

7 I have a sign on my fridge that said “the child’s destiny lies in her/his parents

hands” and I have made a commitment to my children to try to develop all

their talents and reach their full potential.

16 19 35 0 0.85 0.93 +HH

8 In my present existence the most important thing are my children—they are

my centre, totally.

19 19 35 1 0.95 0.98 +HH

211

8.3 Case study: The “Low-stable” Migrant (Type 2)

Juan is representative of the “Low stable” migrant type. He was born and brought

up in El Salvador and emigrated to Australia as a young adult. His childhood was in

some ways difficult. “[N]aïve and trusting” as a child, he nevertheless felt rejected by

his mother; and though his father was “fair”, he was also “inflexible”. Having “learned

to be with people who feel ambivalent about me”, he would “accept blame for

anything” and became obedient in order to please those around him. One salient event

he remarked on: his sister’s apparent suicide, for which he “admired” her. At school he

was “insecure”, “a loner” and “shy”, and his academic performance poor. Later he

would drop out of university in Australia and suffer an extended bout of depression. He

had married in El Salvador, but his marriage broke up in Australia. Having left El

Salvador to live in a “peaceful place”, has no desire to return. On the plus side, he is

“happy” in Australia and expresses guarded optimism about the future and what it

holds.

Juan’s measures of psychological distress and well-being complete the picture of

a well-adjusted person who accepts life as it comes. At the time of the interview he felt

confident in social situations (SAD = 2, cf. sample mean of 8.4); had little concern at

giving a negative impression to others (FNE = 2, cf. sample mean of 13.3); and was not

depressed (BDI-II = 0, cf. sample mean of 9.1). His psychological well-being was about

average on all subscales except Self-acceptance, where it was somewhat higher (PWB6

= 70, cf. sample mean of 64.6).

Major themes. Figure 8.6 shows the mean thematic profile of Type 2 participants

(Panel 1) and for Juan, the case-study selected to exemplify them (Panel 2). The mean

profile of valuations of Type 1 participants differed significantly, but only slightly,

between country of origin and Australia, χ2(5, N = 270) = 14.09, p = .015 (two-tailed).

212

Valuations located in country of origin were high on Autonomy and success,

Powerlessness and isolation, and Aggression and anger whereas those in Australia,

though still high on Autonomy and success and Powerlessness and isolation, were

higher on Aggression and anger and lower on Unity and love.

Juan’s thematic profiles correlated highly with the mean profiles of Type 2

participants both for country of origin ( r = .89) and for Australian valuations (r = .97),

indicating that he is representative of the Type 2 migrant type.

213

Panel 1 All Low-stable migrants

Panel 2 Juan

Figure 8.6 Proportion of valuations, by major theme and location: Type 2 participants

and Juan

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

214

Self-confrontation indices. Figure 8.7 shows Juan’s mean self-confrontation

indices. On self-enhancement, there was no significant difference between his

valuations of country of origin (M = 7.4, SD = 4.9) and those of Australia (M = 9.2, SD

= 5.2), t(33) = –1.07, p = .29 (two-tailed). On union-with-the-other, there was no

significant difference between his valuations of country of origin (M = 6.4, SD = 3.6)

and those of Australia (M = 4.8, SD = 3.2), t(34) = 1.36, p = .18 (two-tailed). On

positive affect, there was no significant difference between his valuations of country of

origin (M = 14.9.7, SD = 9.4) and those of Australia (M = 16.8, SD = 10.5), t(32) = –.58,

p = .57 (two-tailed). On negative affect, there was no significant difference between his

valuations of country of origin (M = 20.5, SD = 12.3) and those of Australia (M = 13.4,

SD = 11.5), t(34) = 1.79, p = .08 (two-tailed).

Figure 8.7 Mean indices of valuations, by location: Juan

0

5

10

15

20

25

30

35

40

S-index O-index P-index N-index

Origin

Australia

General

Ideal

215

Juan is representative of the Type 2, “Low stable” migrant type. According to

Gergen & Gergen (1986) a “stability” narrative links events “in such a way that the

protagonist remains essentially unchanged with respect to evaluative position” (p. 27).

Corroboration of this is found in Figure 8.8, which shows the temporal dimension of his

self-confrontation indices.

Juan’s self-confrontation indices show a slight temporal progression, which is,

however, unrelated to location. That is, his past-oriented valuations for country of origin

resemble his past-oriented valuations for Australia, and likewise for present- and future-

oriented valuations. This explains why his valuations appear to be stable when the

temporal dimension is ignored: they are stable with respect to location. Juan does show

a slight progression in a temporal sense—his future-oriented S and O are slightly higher,

and well-being (defined as P minus N) is substantially higher; but that progression is not

noticeably linked to the act of migration.

Figure 8.8 Mean indices of valuations, by location and period: Juan

0

5

10

15

20

25

30

35

Past Present Future Past Present Future

Origin Australia

S

O

P

N

216

Self-narrative. Juan’s self-narrative is characterised by stasis. He struggles; his is

not an easy life. The following paired valuations show this.

He was, and is, an outsider:

Country of origin: “At school I felt insecure. I was a loner …” (–LL).

Australia: “I felt that I did not belong to this society because people did not know

who I was”. “I felt that I did not belong even to the groups I was participating in

(e.g. University) (–LL).

His reaction to what to most people would be a life-changing event, is almost

laconic: his sister’s suicide did not even qualify for a major theme under his affective

rating. Later, in Australia, he suffers from occasional depression and a negative self-

image:

Country of origin: “When [my sister] apparently killed herself I admired her for

it.”

Australia: “I was very depressed and I had very low self-esteem.” (–LL).

But his life is not without its small victories. He is less shy now:

Country of origin: “I was shy and I learnt not to show off because I thought

people would not like me” (–LL).

Australia: “I am able to express my feelings and thoughts more easily now” (+S).

... and is learning, and relishing, some independence:

Country of origin: “I learnt from [my father] that rules were not to be bent or

broken” (–LL). “I learnt to accept blame for anything and to be obedient …”

Australia: “For the first time in my life I feel I am working for myself as an

individual” (+S).

… though apparently not enough to prevent his being classified as “Low-stable”.

Juan’s is a self-narrative of mental struggle and survival—stable, but

compromised by underachievement.

217

Table 8.2 Valuations with major themes, by location and period: Juan

Valuation S O P N G I Theme

Country of origin: Past

1 Throughout my upbringing I became naïve and trusting. My family

emphasised on me being good.

3 12 10 24 -0.40 -0.57 -O

2 When I was 11, my brother and I went to Los Angeles to study English for a

year. This event set me apart from my peers.

16 8 29 9 0.79 0.65 +S

3 My father is fair but inflexible. I learnt from him that rules were not to be

bent or broken.

2 2 12 35 -0.25 -0.82 -LL

8 At school I felt insecure. I was a loner and was constantly failing subjects. I

was shy and I learnt not to show off because I thought people would not like

me.

2 1 3 33 -0.34 -0.88 -LL

11 At university I discovered the humanistic subjects and for the first time I

enjoyed schooling and studying.

20 6 33 5 0.68 0.77 +S

Country of origin: Present

30 I don’t feel the need to communicate with my family. If I had a problem I

would talk to a friend here.

10 4 21 5 0.53 0.64 +S

Australia: Past

18 I felt that I did not belong even to the groups I was participating in (e.g.

University).

2 0 1 30 -0.16 -0.82 -LL

19 During the first 6 year in Australia I did not do much for myself and

therefore achieved very little. I felt I was shutting down one aspect of my life

after another.

0 0 0 39 -0.37 -0.96 -LL

22 After 5 years of studying I was asked to withdraw from university because I

was failing, and I withdrew from my course and gave up. I was very

depressed and I had very low self-esteem.

1 1 5 24 -0.41 -0.85 -LL

24 Not having a partner I felt that I could make my own decision and I started to

live my own life and not a family group life.

15 6 21 8 0.56 0.70 +S

25 Getting my first professional job was a self-esteem boost for me. I felt I

finally belonged to a group of people that were similar to me.

15 7 23 3 0.62 0.83 +S

Australia: Present

26 I have a job. I’m still learning but I’m where I want to be. 13 4 25 7 0.75 0.71 +S

27 For the first time in my life I feel I am working for myself as an individual. 15 6 26 6 0.71 0.82 +S

29 I am able to express my feelings and thoughts more easily now. 12 4 29 1 0.57 0.84 +S

Australia: Future

33 I want to go back to study to expand my professional horizon. 13 4 24 4 0.66 0.78 +S

218

8.4 Case study: The “High-stable” Migrant (Type 3)

Luisa is representative of the “High stable”, migrant type. She was born in

Mexico and emigrated to Australia as a mature adult of 30. Her childhood was largely

happy, but not without the frustration of “not being recognised … because [she] was a

girl”, and because being “the only girl among four brothers suppressed [her] character.”.

Nevertheless, she saw herself very much as the equal of the men in her family,

describing herself as “competitive” with her brother and the “soul mate” of her husband.

She was also adventurous, with a “thirst for independence” and a fascination for distant

lands, and Australia in particular. Though her relationship with her father was “not very

close”, she states, “I treasure the memory where my father went to my school to drop

the book he had bought for me - which he covered with beautiful paper and labelled

with my name. I felt so special.”

Luisa’s measures of psychological distress and well-being complete the picture of

an optimistic personality who gets the most she can from life. At the time of the

interview she felt confident in social situations (SAD = 1, cf. sample mean of 8.4); had

some, but still below-average, concern at giving a negative impression to others (FNE =

7, cf. sample mean of 13.3); and was not depressed (BDI-II = 3, cf. sample mean of

9.1). Her psychological well-being was well above average on all subscales (PWB1 =

77, cf. sample mean of 63.1; PWB2 = 83, cf. sample mean of 67.1; PWB3 = 79, cf.

sample mean of 63.4; PWB4 = 84, cf. sample mean of 72.7; PWB5 = 83, cf. sample

mean of 66.1; and PWB6 = 70, cf. sample mean of 64.6).

Major themes. Figure 8.9 shows the mean thematic profile of Type 3 participants

(Panel 1) and for Luisa, the case-study selected to exemplify them (Panel 2). The mean

profile of valuations of Type 3 participants did not differ significantly according to

whether they pertained to country of origin or Australia, χ2(5, N = 169) = 7.21, p = .21

219

(two-tailed). Valuations located in both country of origin and Australia were high on the

themes of Strength and unity and Autonomy and success; and, for Luisa but not Type 3

in general, moderately high on Unity and love.

220

Panel 1 All High-stable migrants

Panel 2 Luisa

Figure 8.9 Proportion of valuations, by major theme and location: Type 3 participants

and Luisa

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

0%

10%

20%

30%

40%

50%

Strength and unity

Unity and love

Unfulfilled longing

Powerlessness and isolation

Aggression and anger

Autonomy and success

Origin

Australia

221

Luisa’s thematic profiles correlated highly with the mean profiles of Type 3

participants both for country of origin ( r = .85) and for Australian valuations (r = .96),

indicating that she is representative of the Type 3 migrant type.

Self-confrontation indices. Figure 8.10 shows Luisa’s mean self-confrontation

indices. On self-enhancement, there was no significant difference between her

valuations of country of origin (M = 14.2, SD = 5.6) and those of Australia (M = 17.2,

SD = 3.2), t(39) = –2.25, p = .03 (two-tailed). On union-with-the-other, her valuations

of country of origin scored significantly less (M = 12.8, SD = 7.3) than those of

Australia (M = 17.3, SD = 3.4), t(36) = –2.71, p < .01 (two-tailed). On positive affect,

her valuations of country of origin scored significantly less (M = 28.5, SD = 12.2) than

those of Australia (M = 34.8, SD = 5.7), t(36) = –2.27, p < .05 (two-tailed). On negative

affect, there was no significant difference between her valuations of country of origin

(M = 5.3, SD = 6.1) and those of Australia (M = 4.6, SD = 4.9), t(43) = .45, p = .66

(two-tailed).

Figure 8.10 Mean indices of valuations, by location: Luisa

0

5

10

15

20

25

30

35

40

S-index O-index P-index N-index

Origin

Australia

General

Ideal

222

Luisa is representative of the Type 3, “High stable” migrant type. According to

Gergen & Gergen (1986) a “stability” narrative links events “in such a way that the

protagonist remains essentially unchanged with respect to evaluative position” (p. 27).

Corroboration of this is found in Figure 8.11, which shows the temporal dimension of

her self-confrontation indices.

Luisa’s self-confrontation indices show little temporal or other progression. For

the most part, the self-confrontation indices of all her valuations are quite similar,

whether differentiated by location, by period, or both.

Figure 8.11 Mean indices of valuations, by location and period: Luisa

0

5

10

15

20

25

30

35

Past Present Future Past Present Future

Origin Australia

S

O

P

N

223

Self-narrative. Luisa’s self-narrative is characterised by stasis. She brims with

self-confidence and the achievements, real and perceived, that such confidence

engenders. The following paired valuations show this.

From the outset she feels special and valued—a sentiment that stands her in good

stead later in life:

Country of origin: “I treasure the memory where my father went to my school … I

felt so special” (+HH).

Australia: “This provided me with an interest and the freedom to be with John

when we had time off work” (+HH).

She anticipates, and plans for, a marriage of equals, and achieves it:

Country of origin: “I came back to Mexico. John followed me a few months after

and we were married” (+HH).

Australia: “This was an encounter with a soul mate; we felt so happy in each

other’s company from the beginning and we fell in love” (+HH).

She has a childhood dream, and achieves that too:

Country of origin: “Since my childhood I felt a special fascination with anything

to do with Australia”.

Australia: “I had already become an Australia citizen and I felt totally at home”

(+HH).

Her two dreams—of a soul-mate and a spiritual home—are captured again in

these valuations:

Country of origin: “I finally came to visit Australia. I met my future husband

during this trip. It was on the cards”.

Australia: “My first arrival in Australia was almost like a homecoming.

Everything was more than I ever expected” (+HH).

And she does it all by complementing her country-of-origin self, not by

repudiating or changing it:

224

Country of origin: “The joy of participating closely with the family is hard to

describe. The warmth of the people in general” (+HH).

Australia: “The early years of my children were the happiest and the most

rewarding of my life. I was completely fulfilled and did not want for anything

more” (+HH).

Luisa’s is a self-narrative of building on a strong sense of herself and achieving

more—stability crowned by emotional reward.

225

Table 8.3 Valuations with major themes, by location and period: Luisa

Valuation S O P N G I Theme

Country of origin: Past

3 Whilst in the U.S. (for a period of 2 and a half years) I had to conform with

my grandmother’s ideas of discipline, and that involved performing a number

of responsibilities in the household which did not befit a fourteen year old.

8 0 0 19 -0.44 -0.46 -S

4 School life - year 7 and 8 was a fantastic experience, both academic and

social.

15 8 32 0 0.76 0.76 +S

5 Childhood years- enjoyed school enormously and always excelled in my

grades.

15 9 36 1 0.79 0.80 +S

6 Social life revolved around our cousins and our neighbourhood friends. It

was fun.

9 16 31 10 0.69 0.69 +O

7 I treasure the memory where my father went to my school to drop the book

he had bought for me- which he covered with beautiful paper and labelled

with my name. I felt so special.

20 20 35 0 0.93 0.94 +HH

8 At the age of 15 I returned home to Mexico. Because of the education system

of the time I could not go to high school and therefore the opportunity of

going to university was closed to me. My parents chose to send me to a

secretarial course instead. This was quite disappointing for many years.

11 5 27 3 0.56 0.57 +S

9 When I finished secretarial school I applied for work in the airline industry. I

was very lucky to get a job straight away with American Airlines. From there

I moved on to a better position with Qantas airways. I was 19 at the time.

This was to be an experience that opened my horizons and changed my life.

19 9 38 0 0.86 0.87 +S

10 Being the only girl among four brothers suppressed my character. And this

promoted in me a great thirst for independence. I became very competitive

with Eduardo, my brother, who was closest in age to me and I felt quite

happy when I could earn a better salary [than him, working as a secretary].

16 2 25 4 0.49 0.50 +S

11 My job with Qantas lasted 10 years. It offered me a fabulous wide experience

of interesting work, very rewarding in every sense: intellectually, socially

and financially.

19 12 37 0 0.88 0.89 +S

12 During those same 10 years I was able to travel extensively to places I never

dreamed of.

19 8 35 0 0.80 0.81 +S

13 After 10 years of working for Qantas I finally came to visit Australia. I met

my future husband during this trip. It was on the cards.

16 14 37 0 0.93 0.94

14 [After John’s marriage proposal] I came back to Mexico. John followed me a

few months after and we were married.

20 20 39 3 0.99 0.99 +HH

17 On the other hand I had a close relationship with my mother. As I have got

older I have discovered many of her traits in my personality.

19 20 30 14 0.52 0.54 +HH

18 As my children reached the age of 10 and 12 years we felt we would like to

move to Mexico for them to absorb the Mexican culture; for all of us to live