Identity In Narrative A Study Of Immigrant Discourse

i
Introduction
The writing of this book was motivated by my involvement in three areas of
interest both in academic and personal life. The first one relates to the ways and means
through which language, and in particular narrative, displays its power to voice
experiences, to bring about shared understandings of life events, to shape and transform
individual and collective realities. The second one relates to migration as a social
phenomenon and as a personal experience. I have migrated more than once during my
adult life and, although I am conscious of the profound differences in motivations,
economic backgrounds, origins, adaptation routes, among those who carry out a journey
that takes them away from their countries to settle somewhere else, I am also convinced
that there are many commonalities, many patterns of behavior and experience that are
shared by all of them. Those commonalities constitute a firm basis for understanding and
solidarity, and an occasion for reflection. Finally, the writing of this book was also
spurred by a deep interest in Mexico since the many years I spent in that country
stimulated in me a great admiration for the richness and complexity of the Mexican
people and of Mexican cultural traditions.
The book is based on interviews and ethnographic observation carried out
between September 1996 and June1997 with 14 Mexican economic immigrants living in
Langley Park, Maryland, who were mostly undocumented at the time. The work responds
to two primary objectives: investigating the constitution, representation and negotiation
of identities among Mexican economic migrants to the United States, and showing in
what ways narrative discourse constitutes a privileged locus for the study of identities.
The focus of the analysis is on the connections between the local expression of identities
in narrative discourse and the social processes that surround migration.
There are two preliminary questions that I would like to discuss in the following
pages. The first one is: why study immigrants? The second one is: what are the
advantages of small-scale discourse analytic studies as opposed to quantitatively based
investigations, in order to gain an understanding of migration and the processes of self
definition and redefinition that immigrants live?

ii
Let us start with the first question: the importance of studying immigrants and
immigration hardly needs stressing given the social relevance of the phenomenon. A
great number of new immigrants enter the U.S. every year and of these immigrants, many
are undocumented Mexicans (Dillon, 1997). The presence of Mexican undocumented
workers in the U.S., already estimated between 1.8 and 3.6 millions in the seventiesi
, has
currently reached, according to the national press, a number between 3 and 4 million of
individualsii
. Quantifications of the immigration flux vary, but figures are high enough to
give an idea of the relevance of a phenomenon, largely unknown, but also unmistakably
part of American everyday life. Mexican immigrants, especially undocumented ones,
become more numerous every year and as the divide between the wealth of the U.S. and
the poverty of its neighbors increases, so does the number of those who are pushed across
the border by the dream of a better life.
However, immigration is not only important because of its numerical significance.
It is also important because of its economic, social, and psychological impact. The
constant debate over this topic in the mass media, in the political arena, in academic
circles, and at dinner tables, is a symptom of the centrality of the role of immigration and
immigrants in the political and social landscape of the country. On the other hand, the
continuous attempts, particularly in the South Western states, to limit and regulate the
rights of immigrant workersiii
show how deeply divided politicians and common citizens
are on the extent to which recent immigrants can be considered a true part of society. In
fact, the constant increase in the number of Hispanic immigrants in particular and their
recent attainment of the status of largest and fastest growing minority in the U.S.iv
, has
raised and continues to raise great anxiety among mainstream Americans since often
these immigrants are seen as taking over the country and imposing their own life style,
language, and customs. In many cases being Hispanic is equated with poor performance
at school, drug abuse, poverty and violence. Images and stereotypes abound, but
information on immigrants is scarce and although a wealth of literature on social and
political aspects of immigration exist, very little is known about who immigrants are,
what they think and what they feel, why they come to the U.S., how they see themselves.
This is particularly true of undocumented Mexican immigrants who are active and
present in innumerable work sites across the country and who lend their workforce for

iii
low skilled jobs in areas such as construction and painting, landscaping, catering and
food serving, agriculture and house cleaning in a large number of American cities and
rural areas. Their language, food, music are gaining increasing popularity, but their voice
is rarely heard. Although visible in the work place, they lead their life in anonymity and
isolation. Thus another reason to study immigrant realities, particularly immigrant
identities, is the need to provide insights into aspects of a phenomenon that is amply
debated but largely under-analyzed.
A focus on immigrants and their identity can also help defeat overgeneralization
and stereotyping and show the complexity of immigrant realities and experiences.
Stereotypes are in fact often the result of a lack of knowledge about immigrants.
Anthropologist Rosaldo (1993) stresses the relationship between stereotyping and
ignorance and argues for the importance of listening to what people say about
themselves. Proposing an analysis of Chicano narratives, he underlines that this kind of
research is a response to our ignorance of members’ self-perceptions and our inability to
answer questions about them. The same can be said about undocumented immigrants and
many other minority groups that are often ignored or largely misunderstood.
The second question proposed at the beginning of this introduction was: why
should we rely on small scale qualitative studies such as the one proposed here in order to
gain a deeper understanding of immigration and of the processes of self definition and
redefinition that accompany it? I argue that a qualitative perspective, particularly one
based on discourse and narrative, is much more insightful than quantitative
methodologies because it helps bring to the surface and understand aspects of the
representation of the self that are not apparent through statistics, questionnaires or sample
interviews.
Qualitative studies on Mexican immigrants are scarce not only in the field of
discourse studies, but also in other research areas. Although sociological and economic
aspects of Mexican immigration have constituted and continue to constitute the focus of
many sociological, economic, social psychological, anthropological studies (see
Cornelius and Marcelli, 2001; Durand, 1991; Gaxiola, 1991; Heer, 1990; Wayne,
Chavez, & Castro, 1982; Bustamante, 1979; Wayne, 1978 a and 1978 b; Gamio, 1969a
and 1969b) questions related to self and other-perception, and self and other-

iv
representation, are relatively neglectedv
. But immigration as a process crucially involves
a continuous definition and redefinition of one’s identity and of one’s membership into
larger communities. Life stories analysts and social psychologists see it as one of the
landmark events in the life of individuals and groups. Thus, it is hardly possible to come
to terms with immigrant realities without understanding these “subjective” processes. In
an investigation of socio-psychological responses to migration among Mexican
immigrants, de la Mora (1983), argued that although many studies on the topic suggest
that the factors influencing the outcome of the process are both subjective and objective,
most mainstream analyses have exclusively focused on objective conditions such as
unemployment, inequality in income distribution, patterns of population growth,
educational levels, work-force qualifications, and so on. This emphasis has resulted in a
lack of understanding of the impact of subjective factors related to migration on the life
of individuals and groups.
The importance of developing knowledge on the self-perception and identity
formation among immigrants has been recognized by anthropologists studying new
immigrant populations (see for example Chavez, 1992 and 1994; Rosaldo, 1993). They
argue that such knowledge may for example, lead to a better understanding of the factors
that help immigrants integrate or that, alternatively, prompt their isolation within the host
society. A comprehensive study on Mexican immigrants in Southern California (Wayne,
Chavez, & Castro, 1982) suggested, for example, that the integration of first generation
Mexican immigrants into American society is minimal, as they tend to see themselves as
outsiders to that society even after many years of residence in the U.S. More recent
qualitatively based analyses challenge this kind of accepted wisdom and suggest in
contrast that generalizations on the way new immigrants adjust to life in the U.S. are ill
founded, since too little is known about their lives and the repertoire of identities that
they might be developing. Lamphere (1992), for example, in the introduction to a
collection of papers about the interrelationships between newcomers (including
Mexicans) and established residents in U.S. cities, argues that stereotypical images about
the way immigrants relate to other ethnic groups are inadequate to describe new urban
realities. Similarly, in a study based on interviews about community membership among
undocumented Mexican immigrants in the San Diego area, Chavez (1992) challenges the

v
assumption that the strong links with their country of origin hinders Mexican immigrants'
development of a new sense of community in the U.S. since:
... while many Mexicans retain ties with their home families and communities,
this does not necessarily undermine their experience in their new communities,
experiences that may isolate them from the larger society or lead to change,
sometimes well thought of and other times unconscious, in their orientation from
sojourners to settlers. (p. 56)
In this process, immigrants may be developing “multiple senses of community
membership.”
In sum, qualitative studies of immigrant communities are important both to assess
and evaluate the ways immigrants fit into the host society, and to provide knowledge
about communities that are often the object of stereotyping and misjudgment. In this
book I argue for the importance of the analysis of identity among Mexican immigrants,
but I also show how such analysis inevitably leads to its expression in discourse. I also
argue that narrative discourse is particularly illuminating of the ways in which
immigrants represent the migration process and themselves in it. Thus, my objective is
not only to describe aspects of the identity of Mexican immigrants, but also to advocate
for a discourse-based approach to identity. Language is central to the expression of
identity because it is not a reflection of our apprehension of reality; it is not a "conduit"
(Reddy, 1979) for thought, but rather a constitutive aspect of our experience of the world.
We cannot understand and share experience if we do not express it linguistically. The
way we express our experiences is as part of those experiences as the material and
psychological processes that prompted our telling of them. Story telling, as other
discourse activities, is seen here as situated discursive practice (Fairclough, 1989) in the
sense that it both obeys and creates social rules, understandings, and roles. It obeys
social rules that dictate how narratives should be constructed, by whom and to whom
they should be toldvi
, what is tellable, and howvii
. Furthermore, storytelling, like other
discursive practices, rests on socially shared meanings, conceptions and ideologies (van
Dijk, 1998), establishing a constant dialogue (Bakhtin, 1981) with them, but also
generates new meanings and new behaviors. Among the central functions of storytelling
is, as I will argue, that of presenting and representing identity. In this framework

vi
narrating is a way of talking about the self, but also a way of practicing certain types of
identity in specific interactional contexts.
The recognition of the structuring power of discourse and of discourse organization
is, therefore, at the heart of the enterprise of studying identity through discourse analysis.
The choice of narratives as the focus of analysis and the centrality of narrative in the
expression and negotiation of identity will be thoroughly discussed in Chapter 1. Here I
only want to note that the focus on the micro-analysis of naturally occurring talk and the
emphasis on the local mechanisms through which identity is expressed and negotiated in
narrative, derive from the conviction, shared by many interactional sociolinguists, that it
is mainly through the analysis of data in painstaking detail and the consideration of the
contextualization cues that speakers use to convey specific meanings (Gumperz, 1982,
1992) that it is possible to generate hypotheses on how members of a community
represent and negotiate their belonging to social categoriesviii
. According to interactional
sociolinguists and other interactionally oriented scholars, in order to understand how
language contextualizes social realities, it is important to combine a close focus on the
details of texts “with a broader conception of meaning” (Basso1992, p. 268). Detailed
discourse analysis is like a magnifying glass in that it illuminates the way linguistic items
and strategies employed by individuals are part of a repertory of resources shared by
communities. It is through the study of situated discourse instances that cultural and
social meanings become apparent to the analystix
.
But why study narratives in particular? Narrative is one of the privileged forms
used by humans to elaborate experience. This is why narratives have been widely studied
as windows into the analysis of human communities and individuals in fields as diverse
as anthropology (Levi Strauss, 1963), ethnography and folklore (Hymes, 1981; Bauman,
1986; Rosaldo, 1986), social history (Griffin, 1993), psychology (Rosenwald & Ochberg,
1992; Bruner, 1990; Polkinghorne, 1988; Mishler, 1986); sociology (Somers & Gibson,
1994). One reason for this popularity is their methodological richness. Narratives have
been used as data in many fields of the social sciences and narrative analysis has
constituted the methodological tool of a revolution in qualitative research that has
become generally identified as the ‘narrative turn’. This generalized interest greatly owes
to the characteristics of narratives as texts. Narratives are highly spontaneous and at the

vii
same time highly organized texts both in the way they are structured, and in the way they
are inserted in conversation (Labov & Waletzky, 1967/1997; Labov, 1972; Jefferson,
1978); for this reason they can be recognized and analyzed as a specific and highly
constrained discourse genre. Furthermore, they are a discourse genre that invites and
promotes involvement and participation. Labov's appreciation of the highly spontaneous
character of narratives led him to use them as a central tool for his study of the vernacular
language, since he thought that when people narrate their experience, they get involved
and become less self conscious of the way they speak. After him, researchers have begun
to use narratives as an alternative to more traditional methods of elicitation such as
questionnaires and formal interviews. In the present study the spontaneity and
involvement that the telling of narratives created within the interview context were an
invaluable aid. I was interested in how immigrants make sense of their immigration
experience and I asked them questions on how they felt, what they thought about it, how
migration had changed them. But a direct reconstruction and reflection on the personal
experience of immigration is difficult to elicit. I anticipated that immigrants would have
difficulties of various kinds in talking about, or reflecting on their experiences explicitly,
while I thought that they would more easily tell stories, whether asked to do so or not.
This turned out to be true, since stories and other kinds of narratives emerged throughout
the interviews as spontaneous answers to questions, as illustrations of argumentative
points, and as recollections of past experience. Narratives were then a central tool for me
as a researcher in that they allowed me to study important aspects of the identity
construction in this group, and for the immigrants as interviewees because they allowed
them to talk more freely about their experiences.
Another important aspect of narratives as resources for studying groups and
communities is their ability to serve as locuses for the keying of experience. Goffman
uses this term to refer to “all strips of depicted personal experience made available for
participation to an audience” (1974, p. 53). In storytelling many linguistic devices, such
as tense, reported speech or pronoun switching, allow narrators to replay their
experiences for an audience as if these were taking place before their eyes. In that sense,
although narratives might occur as a response to a question by the interviewer or they
might be directly elicited, they still largely respond to the expressive needs of the narrator

viii
and therefore are more likely to reveal her/his point of view on events and experiences
than other kinds of talk. Furthermore, narratives are in many cases negotiated, thus their
significance is established interactively by the participants in a speech event. Therefore,
they allow for different participants in an interaction to express their evaluation of events
(Goodwin, 1986). This aspect of storytelling was important to my study since the telling
of narratives constituted an occasion for the discussion of the meaning of personal
experience to members of the community. Participants in interviews expressed collective
values and beliefs either through evaluation of narratives told by others, or through co-
construction of narratives with others. Thus, while answers to questions were most of the
time individual, narratives invited more participation and negotiation of meaning from
participants.
As I have argued, discourse, and narrative in particular, represent the point of
intersection between the expression of individual feelings and representations and the
reflection upon and construction of societal processes, ideologies and roles. The latter
become alive in the arena of talk in a unique way. By analyzing narratives we analyze
not only individual stories and experiences, but also collective social representations and
ideologies.
Overview of the volume
The internal organization of the book mirrors my ideas, detailed in chapter one,
about the relationships between narrative discourse and identity. Except for chapter two,
in which I give background information on Mexican undocumented immigrants and on
the group of immigrant workers interviewed for this study and I discuss some
methodological choices, the rest of the book is centered on the analysis and discussion of
different aspects of the presentation and negotiation of identity in narrative discourse. I
argue that identity is not necessarily expressed at one and the same level since it can be
displayed or given off, but it can also be openly negotiated. The degree of openness may
vary, in the sense that choices as to self-presentation can be more or less explicit
depending on the general interactional function of the narrative itself and the storytelling
context. Identities emerge in my analysis through the establishment of connections

ix
between linguistic choices, interactional worlds and story worlds. My proposal is that in
order to study identity, we need to look at these different aspects and at its different ways
of emergence in discourse.
I focus on the analysis of two basic aspects of the construction and expression of
identity: the projection of the self into specific social roles, and the expression of
membership into groups and communities. The first aspect, the projection of social roles,
is analyzed through the consideration of ways of presenting the self in relation to others,
and of ways of presenting the self in relation to social experiences. I look at the role of
the self with respect to others as expressed in social orientation, while I analyze the role
of the self with respect to social experiences as agency. The linguistic phenomena and
strategies on which I focus are pronominal choices and voicing. Both pronominal choices
and voicing operate at a level where identity is displayed more than openly discussed.
Chapters three and four are devoted to this level of contextualization/expression of
identity. In chapter three I analyze pronominal choice and other linguistic strategies that
are seen as an index of social and cultural meanings related to broad conceptions of the
persona. In chapter four I focus on voicing. The analysis is centered on the use of
constructed dialogue to report events and actions in the particular story worlds connected
with the border crossing. The focus is on the narrators’ presentation of his/her role as
figure in the story world in that the narrators’ choices in terms of reporting forms, types
of acts reported, and attribution of those acts to story characters, is seen as signaling
different degrees of agency and participation in the narrated action.
The second level of analysis of identity deals more, even though not exclusively,
with the explicit construction of self in relation to the member’s community or to external
groups. Basic to membership construction is self and other categorization, which is
studied through identification strategies. When self and other categorization is at stake,
identity is more often negotiated than displayed and in order to analyze it we need to
resort to implicit and explicit references to belief systems and ideologies. This level of
analysis is taken up in chapters five and six. Chapter five discusses the categorization of
self and others. Crucial to such categorization are narrative strategies used to introduce
characters in stories. I argue that the analysis of story orientations reveals that immigrants
use ethnicity as a central identification category for self and others in their stories and

x
that ethnic identification reflects and constitutes different levels of context, from the local
negotiation of positions about self and others and the creation of participation
frameworks in particular interactions, to the articulation of values and beliefs shared in
the community and the contextualization of cultural and social norms. Chapter six
focuses more closely on the narrators’ articulation of social representations and beliefs
through story telling, by looking at the application of the ethnic category “Hispanic” to
self and others in different story-worlds. This chapter focuses on the comparison of the
construction and definition of identity in different story worlds, showing how self and
other representations are based on schematic relationships between actions and identities
that are often encoded in stories. However, I also discuss how even the same categories
for self and other description may acquire different senses depending on the story-world
depicted and/or on the interactional worldx
, and how narrators may display conflicting
stances towards apparently uncontroversial definitions of the self.
Both dimensions of identity are studied interactionally in the sense that the
analysis does not look at story-world organization as such, but at the connections that
speakers establish between their narratives and the discourse in which they are inserted.
However, interactional construction and negotiation is not taken as exhausting the
contextual nature of narrative. The dialogue established by narrators cannot be
exclusively reduced to the exchange with audiences, since participants are also engaged
in dialogue with mainstream discourses about immigrants and immigration. In that sense
again, the analysis of narratives needs to take into account how local contexts interact
with wider contexts such as ideologies, belief systems, and the intertextual dimension.
To conclude, this book proposes an analysis of the narrative construction of
identity by undocumented Mexican immigrants, but also an approach to the study of
identity through narrative. The focus of the analysis is not on the projection of individual
selves, but on the dimension of group identity and therefore particular attention is
devoted to the processes and strategies of identity construction that seem to be common
among members of the group interviewed and on the nexus between the local expression
of identity by particular narrators and the more global processes of collective
representation that frame and interact with such local expressions.

xi
i
See "Legal and Illegal Immigration to the U.S.", Report by the Selected Committee on Population. U.S.
House of Representatives, 96 Congress, Second Session, Serial C, Washington D.C., 1978, p.2.
ii
See Allen, M. (2001). Mexico still focused on illegal workers. The Washington Post, September 5, A2.
iii
See the debates over Propositions 185 and 227 in California.
iv
Data released by the U.S. Census Bureau in 2000, show that the Hispanic population has reached 35.3
million in the U.S. thus becoming the largest minority in the country. More than half of these 35 million
individuals declare to be Mexican.
v
Few studies on this topic exist. See Buriel & Cardoza (1993) on ethnic labeling; Chavez (1994) on
perceptions about the place of individuals in communities; De la Mora (1983) on psychosocial factors in
the definition of self among Mexican immigrants.
vi
See Goffman (1981) and his notions of participation frameworks and production formats that explain how
discourse activities are differently organized in terms of production and reception.
vii
See Bauman (1986) on the concept of narrative as ‘performance’, i.e. as a discourse genre governed by
rules dictating how it should be best constructed and presented to an audience.
viii
See Rampton (2001) on Interactional Sociolinguistics as a tool to seek answers to wide social problems.
ix
Sherzel’s observations about anthropological linguistics are illuminating of the way the relationship
between discourse and culture is viewed here. He says: “Increasingly, contemporary research in linguistic
anthropology takes discourse as its starting point, theoretically and methodologically, for linguistic and
cultural analysis. As distinct from viewing texts as metaphors (in the sense of Geertz, 1973), an increasing
number of researchers, each in quite different ways, analyzes discourse, large and small, written and oral,
permanent and fleeting, as not only worthy of investigating in its own right, but as embodiment of the
essence of culture and as constitutive of what the language-culture-society relationship is about” (1987, p.
297).
x
I use the term interactional world to refer to the domain of the interaction in which narratives are told.
The difference in my use of the terms interactional world and storytelling world is that the latter refers to
the immediate context of the telling of a story, while the former refers more in general to the speech
activity of which the telling of a narrative is a particular moment.

1
CHAPTER 1
Identity in narrative: a discourse approach
Introduction
In this chapter I discuss narrative, identity and their relationships. I attempt to
show why narrative is central to the study of identity and which properties of narrative as
a genre make it particularly apt to become the locus of expression, construction and
enactment of identity, but also a privileged genre for its analysis. In the first section, I
present my definition of narrative and review some theoretical models that are basic to
understand both narrative structure and function. In the following section, I examine
some theoretical approaches to identity and to its analysis in narrative discourse. I then
present my own approach to the analysis of identity in narrative. I discuss different levels
and modes of expression of identity in narrative, review linguistic and textual phenomena
that relate to these different levels, and discuss the methodological tools and analytical
levels that I used to analyze identity as a collective phenomenon. In the last section, I go
back to the theoretical question of the relationship between identity, discourse and
context and explain how my approach to narrative identity is informed by a view of
discourse as social practice.
1. Narrative genre and types of narratives
The first question that I want to address is that of the definition of narrative as a
genre and of the kinds of narratives that form the object of my analysis. Among the
criteria proposed to distinguish narrative from non-narrative texts, one dimension is, in
my view, essential to the characterization of a text as narrative. Such dimension is
temporal ordering, or sequentiality. Essentially, narratives are texts that recount events in
a sequential order. Even when sequentiality is conceived in terms of casual connections,
there is a temporal aspect to it since events that generate other events are presented as
preceding them temporally. The idea of temporal ordering as a defining property of
narrative is one of the tenets of literary narratology (Bal, 1985; Genette, 1980), a

2
discipline that has had great influence on linguistic studies of narrative. Prince (1982, p.
4), for example describes narrative as: “the representation of at least two real or fictive
events or situations in a time sequence, neither of which presupposes or entails the other.”
The temporal dimension is viewed by many scholars as inextricably tied with narrative,
both in the sense that time itself cannot be conceived outside its expression through
narrative (Ricoeur, 1984), and in the sense that it is through the weaving of events in time
that narratives realize their meaning making and interpretive functions (Brockmeyer,
2000). Linguists who have studied narrative also give prominence to time as a principle
governing the organization of narrative. In his groundbreaking work, Labov (1972, p.
359) incorporated time in his definition of narrative as the recapitulation of past
experience, and Ochs & Capps (2001, p. 2) recently characterized personal narrative as “a
way of using language or another symbolic system to imbue life events with a temporal
and logic order.”
Aside from temporal ordering, which is usually accepted as basic to narrative,
other criteria have been proposed as distinctive features for narrative, but these are not as
universal, or as applicable to all kinds of narrative texts. In fact, most definitions of
narrative either apply to specific narrative genres, but not to others, or describe
prototypical cases only. The prototype of a narrative, both in literary and conversational
domains, is the story. Stories can be described not only as narratives that have a
sequential and temporal ordering, but also as texts that include some kind of rupture or
disturbance in the normal course of events, some kind of unexpected action that provokes
a reaction and/or an adjustment. Thus linguistic, literary and psychological models of
stories recognize the existence of textual components representing a basic action structure
and progression in these types of texts. Labov (1972) and Labov& Waletzki (1967/1997)
conceived of typical stories as composed of a number of sections:
1) An abstract that summarizes what the story is about
2) An orientation that gives indications about the setting of the story and its protagonists
3) A complicating action that presents the main action of the story
4) An evaluation through which the narrator gives the point of the story

3
5) A result that represents the resolution to the complicating action
6) A coda that signals the closing of the story and bridges the gap with the present
Ochs & Capps (2001, p.173) argue that storylines are articulated in ways that present
explanations of events and propose the following story components:
1) Setting
2) Unexpected event
3) Psychological/physiological responses
4) Object/state change
5) Unplanned action
6) Attempt
In their model, while settings lay the background for understanding unexpected events,
the latter may set in motion a response, a change of state, a random action, or an attempt
to deal with them.
In both these linguistic models the axis around which stories revolve is a
complicating event. Research on psychological responses to stories confirms the
prototypical character of stories that have the kind of structure outlined above. Brewer
(1985. p. 170), who attempted to devise universal properties of stories, hypothesized that
readers enjoy narratives if they produce “surprise and resolution, suspense and resolution,
or curiosity and resolution.” To support his hypothesis, he reports results of a study
conducted with Lichtenstein (Brewer & Lichtenstein, 1980) in which readers who were
asked to rate narratives on the degree to which they were stories or non-stories, did not
consider texts without an “initiating event” or an “outcome” to be stories. Thus the way
we conceive of stories usually reflects a general expectation about their structure: stories
may be told for many reasons including to enjoy, inform, argue, and express feelings, but
they are expected to convey a sense of suspense or surprise and a closure of some kind.
This expectation is related to a further criterion used to distinguish stories from non-
stories: tellability. According to Polanyi (1985), for example, stories are usually
conceived as texts that evolve around events that are ‘tellable’, i.e. interesting, surprising,

4
or unexpected in some way. Thus the idea of tellability is tied to the presence of a
complicating action in the story and so examples of highly tellable stories both in
everyday talk and in literature are those that present dramatic events, out of the ordinary
occurrences, unexpected developments or resolutions. Finally, both Labov (1972) and
Polanyi (1979 and 1985) mentioned the importance for prototypical stories to have a
point, i.e. to convey the narrator’s interpretation and point of view on characters, events,
or state of affairs. Labov talked about evaluation as a main component of stories and a
section destined to carry out the function of responding to a possible “so what?” coming
from a listener. Polanyi expressed a similar view when she argued that conversational
stories need to have a point in order to be successful. To summarize: Prototypical
narratives, or stories, are narratives that tell past events, revolve around unexpected
episodes, ruptures or disturbances of normal states of affairs or social rules, and convey a
specific message and interpretation about those events and/or the characters involved in
them.
However, research in recent years has increasingly pointed to the variability of the
texts that belong to the narrative genre and to the existence of many types of narratives
that do not fit the description given above. Narratives dramatically vary according to
structure, content type, social function, and interactional organization. Thus, while stories
are usually conversational events whose topics are not pre-established, many other types
of narratives develop around topics that have been previously stipulated, such as court
narratives or elicited accounts of personal or social events. While stories have a specific
point, other kinds of narratives, such as autobiographies or historical chronicles, do not
have one single point. While stories depict discrete events, habitual narratives tell events
that used to take place over and over again. While many stories are told to amuse and
entertain, others are told to inform, to accuse, to argue, only to mention some of their
possible functions.
Besides differing in topics, functions, internal structure, narratives greatly differ in
terms of the interactional structures that they create and or reflect. Shuman (1986), Blum-
Kulka (1993), M.H. Goodwin (1990a and b), C. Goodwin (1984), Ervin-Tripp& Küntai
(1997), Schegloff (1997), Ochs & Capps (2001), among others, have shown that
storytelling as an activity may involve a variety of participation formats reflecting the

5
power and social relationships among interactants. From monologic narratives, to
polyphonic ones, from spontaneous narratives to elicited ones, from finished to
unfinished tellings, from tellings that take place once to retellings, from disputed to
undisputed tales, the interactional formats that narratives create and in which they are
inserted are innumerable. For all these reasons, although we may look at stories as
prototypes and as a basic genre from which the others are derived, characterizing
narratives in terms of stories is reductive and may lead to neglect storytelling as a process
and to focus exclusively on stories as products.
In this book, I use as data two types of narratives: Stories of personal experience and
accounts of the border crossing that I call chronicles. While stories of personal
experience exhibit the characteristics of prototypical narratives as described above,
border crossing chronicles are usually longer narratives told by the informants in
response to a question on how they managed to get to the United States and are centered
around the telling of the journey between Mexico and the United States, or simply around
the crossing. I describe the characteristics of these narratives in more detail in chapter 4
where I compare them to stories of personal experience.
2. Identity and narrative
Identity is an extremely complex construct and simple definitions of what the
term refers to are difficult to find as there is no neutral way to characterize it. Definitions
of identity, especially within social psychology, often refer to a sense of belonging to
social categories. According to Tajfel (1981, p. 255), for example, identity is ‘that part of
an individual’s self-concept which derives from his knowledge of his membership in a
social group (or groups) together with the value and emotional significance attached to
that membership.” Linguists and linguistic anthropologists focus on the role of language
in the process. Thus Kroskrity (2000) talks about identity as “the linguistic construction
of membership in one or more social groups or categories” (p.111), and stresses the fact
that although identity is not necessarily expressed through linguistic means, language
plays a central role in its construction. These definitions reflect some of the terms of the
current debate on identity since while Tajfel describes identity as ‘self-concept’,

6
Kroskrity talks about it as a ‘construction’. Thus, on the one hand we have a description
that implies something stable and definite like a concept, but on the other hand we have
the characterization of a process. Furthermore, while Tajfel ascribes identity to an
individual, Kroskrity does not attribute the process to any specific agent. Another point of
contention that is apparent in these two definitions and that elicits opposing views in
contemporary debate over identity, is the contrast between a process firmly situated in the
individual and a process grounded within social interactions and institutions in which and
with which individuals and groups are engaged.
Post-structuralist and social constructivist positions developed in the 60’s and
70’s have profoundly influenced recent reflections on identity in linguistics. Francophone
post-structuralist thinkers have contributed to modern conceptions of identity through
their reflections on ‘the subject’ in language, pointing to the irreducible link between the
constitution of subjectivity itself and language. Benveniste’s equation between the
subject and the subject of speech (Benveniste, 2000 [orig.1958]) and Derrida’s claim that
the subject is ‘inscribed in language, is a function of language’ (Derrida, 2000, p. 91
[orig.1972]) both point in this direction. Another tenet of post-structuralist thinking has
been the idea that subjectivity only exists as an effect of social practices and cultural
templates. Such is the sense of Althusser’s claim that the subject is an ideological effect
since individuals become subjects only by virtue of their ‘interpellation’ through ideology
(Althusser, 1971). This is also the direction of Foucault’s theory that social practices are
responsible for creating specific social subjects (Foucault, 1984).
Social constructivist theorists in the social sciences, on the other hand, have
contributed to a notion of identity based on the premise that social realities are
constructed and not given (Berger & Luckman, 1967, p.84) and therefore need to be
regarded as accomplishments to which human beings arrive through social work
(Zimmermann & Wieder, 1970). These ideas have been instrumental to the recent turn in
identity studies away from a notion of identity as the prerogative of a subject and a
function of her/his beliefs and feelings and a conception of subjectivity itself as a stable
and coherent ensemble of characteristics defining groups or persons. Postmodern ideas
about identity reject the notion of the ‘subject’ as a Cartesian unit encompassing
rationality and freedom of choices. They have led to the substitution of the single term,

7
‘identity’ with alternative formulations, such as its plural ‘identities’ - reflecting the
notion that individuals and groups have access to a repertoire of choices socially
available to them- or the term ‘identification’ - referring to a construction and a process
never completed that requires discursive work (Hall, 2000, p.16). This turn has had
important consequences for discourse studies since researchers have turned to the
investigation of ways in which fragmented and ‘polyphonous’ (Barrett, 1999) identities
coexist within the same individual, ways in which identities change and evolve according
to situations, interlocutors and contexts, ways in which identities are created, imposed,
enjoined, or repressed through social institutions and interactions.
With respect to narrative studies, this new focus on identity as a social
construction has taken a number of different routes. Among them we can distinguish two
dominant paradigms: On the one hand the tradition centered on autobiography and based
on psychological theories of identity, and on the other hand, the conversation analytic and
ethnomethodological tradition. In the first approach, the relationship between narrative
and the expression of identity has been widely conceived in terms of the relationship
between the self and the act of narrating, positing the act of narrating as an act of
constitution of identity. A great deal of work on autobiography has followed this route
and many scholars in psychology have been interested in the connection between ‘the
self’ and narrative. Bruner (1990) noticed that between the 70's and the 80's psychologists
increasingly started to see the self as a storyteller. As a result, narrative studies have
grown exponentially adopting as their methodological tool the investigation of the
narrative construction of the self by individuals and groups. Bruner was among the first
scholars to embrace a view of the self not as a static and fixed entity, but a social
construction that emerges mainly in narrative form. Another psychologist, Polkinghorne
(1991), suggested that narrativization is a process basic to the constitution of the self in
that it allows humans to make sense of experience and to grasp the self as a whole. He
argued that narrative helps build a sense of self by providing temporal organization,
which in turn produces coherent self-understanding. In his view, narrative configuration
is a process that takes place through emplotment, "a procedure configuring temporal
elements into a whole by grasping them together and directing them towards a conclusion

8
or sequence of disconnected events into a unified story with a point or theme" (1991,
p.141). In a similar vein, the philosopher Kerby has argued that:
Narratives are a primary embodiment of our understanding of the world, of
experience, and ultimately of ourselves. Narrative emplotment appears to yield a
form of understanding of human experience, both individual and collective, that is
not amenable to other forms of exposition or analysis. (1991, p.3).
For these authors then, narrative is central in the encoding of human experience because
it is based on temporal sequence and because experience itself becomes intelligible to
humans only when they narrate it. Studies of autobiographical narrative (see for example;
Rosenwald and Ochberg, 1992; Gergen and Gergen, 1988; Bruner, 1991and 1993;
McAdams, 1993; Lieblich, Lieblich,Tuval-Mashiach & Zilber 1998; Mishler, 1999;
Brockmeier & Cardbaugh, 2001) have stressed the postmodernist conception of the self
as “a reflexive construction” (Brockmeier, 2000, p. 53) and as a process in flux.
According to this approach stories reflect an inner reality, but also shape it and therefore
identity cannot be seen as a product or a given, but needs to be seen as an ever changing
process. Recent developments in this field have stressed the role of interaction in
autobiographical self-construction through the concept of ‘positioning’: a process of
identity construction involving both the storyteller and the audience (see Wortham, 2001;
Bamberg, 1997; Davies & Harré, 1990; Harré & Van Langenhove, 1999). However,
many scholars working within this tradition have focused on the concept of ‘self’ as the
expression of individual, mainly monologic, processes of construction and reconstruction
of personal experience.
At the opposite end, the tradition of narrative studies inspired by Ethno-
methodology and Conversation Analysis looks at identity mainly as emerging in
interactional circumstances, thus a process in itself, constituted in ‘performance’,
negotiated and enacted, not internalized in any way, and with no substantial existence
outside the local interactional context. Bauman’s (2000, p. 1) description of identity
applies to this approach:

9
In this perspective identity is an emergent construction, the situated outcome of a
rhetorical and interpretive process in which interactants make situationally
motivated selections from socially constituted repertoires of identificational and
affiliational resources and craft these semiotic resources into identity claims for
presentation to others.
Within this paradigm, identity is defined in terms of members’ orientation to the context
at hand, and as a process activated in relation to different contexts of interaction. A basic
construct to analyze processes of identification in such approach has been the study of
categorization processes, since people are not seen as having an identity, but rather as
being “cast into a category with associated characteristics or features.” (Antaki &
Widdicombe, 1998, p.3). Categories defining people’s identity are seen as locally
occasioned and made relevant through specific orientations displayed by interactants in
interactional contexts and negotiated with their interlocutors. This approach to identity
has spurred some interesting developments in narrative studies as well. Following
remarks by conversation analysts on the importance of incorporating interactional
contexts and members’ orientations within the study of narrative, (see for example
Schegloff 1997; Goodwin, 1997) scholars in this tradition have criticized the prevalence
of narrative studies centered on monologic stories produced within interviews. They have
focused instead on the co-construction and negotiation of identities as accomplishments
within talk- in- interaction. These accounts underscore the role of interviewers (Lucius-
Hoene and Deppermann, 2000; McKenzie, 1999) or other interactants (Kyratzkis, 1999)
in the co-contruction of the identity displayed by narrators and emphasize that identity is
a strategic construct sensitive to local occasioning and circumstances.
This brief summary of positions on identity is useful to point to some conflicts in
the theoretical-methodological choices of scholars who study the intersection between
identity and narrative language. The first one opposes a focus on the individual as the
receptor and articulator of social meanings or of conflicting personal images, to the focus
on context as the shaping force determining individual identities. The second one opposes
the study and analysis of ‘naturally occurring’ narratives to the study of elicited or
‘solicited’ narratives.

10
The emphasis of this book is on group identity and on the expression of identity
as a process that is shaped and at the same time shapes collective social and discursive
practices. In this sense, identity is not primarily conceived as the expression of an
individual’s definition of self, since the ‘self’ is “never more than a part in a social
relation, and the subject is, as they say, social even in his or her solitude." (Hanks, 1990,
p.7). The approach taken here stresses the fact that narrators construct and articulate a
variety of meanings that go beyond the manifestation of their individual selves to
encompass their multiple ties to social groups and practices. Narrating is seen as a
discursive practice, i.e. a form of social practice centered on discourse (Fairclough, 1989)
that both reflects social beliefs and relationships and contributes to negotiate and modify
them. Through narratives people create and negotiate understandings of social realities,
but they also continuously modify the social relationships that exist among them and also,
potentially, with others who are not present in the interaction.
The relationship between narrative and identity is here seen as operating at
different levels:
a) At one level, identity can be related to narrators’ adherence to cultural ways of
telling through the articulation of linguistic and rhetorical resources. Narrators
draw on, and creatively build upon, shared narrative resources such as story
schemata, rhetorical and performance devices, styles, that identify them as
members of specific communities
b) At another level, identity can be related to the negotiation of social roles (both
local and global) that conform or oppose the ones attributed to narrators by
communities and individuals. Narrators use stories as stages for the enactment,
reflection or negation of social relationships and concretely contribute to
perpetrate or modify them
c) Yet at another level, identity can be related to the expression, discussion and
negotiation of membership into communities. Central to such process is the
categorization of self and others and the negotiation of beliefs and stances that
help narrators identify themselves as members of groups or distinguish
themselves from members of other groups

11
The first aspect of narrative identity has to do with specific ways of telling related to
the use of shared linguistic, rhetorical and interactional narrative resources. Groups
defined in terms of nationality, gender, or ethnicity have been shown to use narrative
resources in specific ways that set them aside from other groups. Scholars that have
carried out systematic comparisons between groups of speakers belonging to different
nationalities for example have been able to demonstrate that differences between national
groups exist in terms of story topics and of storytelling strategies. Tannen (1980 and
1989) showed differences in the choice of evaluation devices and use of detail between
Greeks and Americans; Blum-Kulka (1993) reported differences between Americans and
Israelis in the type of narratives that were told, in their topics, and in the participation
frameworks enacted when the narratives were told. Still other scholars attempted to show
how ethnically defined groups exhibit specific patterns in telling strategies, topics, or
narrative organization. For example, Michaels (1981) investigated differences in topic
organization between Caucasians and African Americans children and reported that black
children use a topic-associating style that is different from the hierarchical organization
of narratives told by white children. Similarly, Heath (1983) found important
dissimilarities in the way narratives were organized among the population of two towns
in the U.S. and related them to specific socialization processes. Johnstone (1990), who
studied the cultural content of narratives told by Midwestern Americans, found
differences between men and women in the type of worlds evoked and the way
protagonists were depicted in narratives. More recent studies have proposed other aspects
that may differentiate groups (defined by gender or ethnicity) in the way they tell
narratives, such as the degree of discourse integration of stories (Sawin, 1999), and the
choice of language or language varieties (Bukholtz, 1999; Holmes, 1997; Barrett, 1999).
Adherence to cultural ways of telling has also been interpreted as the adoption of
particular telling styles. Bauman (1986 and 2000) and Hymes (1981) studied narratives as
cultural speech acts with specific performance rules respectively in South Western
communities and among Native Americans. Other scholars (Maryns & Blommaert, 2001)
have analyzed how shifts in narrative style connect the same speaker to polyphonous
identities related to different ethnic or national groups. Anthropologists and linguists have
also argued that linguistic devices used in narratives can only be understood against the

12
background of wider cultural frameworks for the organization of experience in specific
speech communities. Scollon & Scollon (1981), for example, described Athabaskan
narratives as having both a rhythmic organization that is different from Western
narratives, and a content that reflects values and beliefs related to Athabaskan
philosophy. All these studies show the many ways in which speakers use the narrative
resources available to them in ways that are in some sense ‘typical’ of their communities.
The second level of identity construction that I have identified is the negotiation
of personal and social roles that takes place through stories. Researchers investigating the
representation of self in story worlds have pointed to different kinds of positions that
narrators attribute to themselves as figures in the story-world by looking at linguistic
choices indexing social or personal roles in both story and in interactional worlds.
Schiffrin (1996, 2000), for example, discusses how stories told by Jewish women about
relatives reflect and shape their stances with respect to their identities as women and
family members. O' Connor (1994) shows how narratives allow prisoners to position
themselves in non agentive terms towards their past actions, and to propose a new sense
of self. Hamilton suggests (1998) that personal narratives told by patients contribute to
their collective construction of an identity as "survivors" and fighters. Studies
investigating gendered identities have also greatly contributed to the analysis of agency
and role representation in stories and in storytelling. Ochs and Taylor (1995) demonstrate
how narratives become occasions for the reproduction and replay of family roles, Capps
(1999) illustrates how women are constructed as agoraphobic through narratives, Holmes
(forthcoming) shows how both men and women manage anecdotes to project and build
individual and collective work-related roles. Finally, research in anthropological
linguistics shows how narrators emplot important social events in ways that help
strengthen or underscore the roles that they impersonate within their communities
(Briggs, 1997).
The third level at which narratives can become a locus for the enactment and
reflection of identities is the expression and negotiation of membership into communities.
Such sense of belonging is expressed through processes of categorization and labeling
and is often defined by the adherence to values, beliefs and behaviors. Stories provide a
powerful occasion for narrators to classify and evaluate characters and their actions

13
against implicit or explicit norms and values. Since stories typically deal with violations
of expected courses of actions, narrators are able to present moral stances that confirm or
refute generally held positions and values and therefore to evaluate themselves and/or
others as members of groups holding or rejecting moral values and social norms. Polanyi
(1985) underscores this aspect of storytelling arguing that the analysis of the evaluation
of unexpected events in stories provides important insights into the values and beliefs of
specific groups or cultures. She proposed a "cultural reading" of the stories that she
collected, and was able to show how their topics and the evaluations of events and
characters in them, reflected values widespread at all levels of American society. Ochs &
Capps also (2001) discuss the fact that experiences are framed within the limits of
stereotypes and socially accepted conventions through cultural templates or conventional
images of people and events. Life-story analysts also illustrated the connection between
stories and beliefs. Linde, (1993), for example argued that individual life stories are
constructed according to coherence principles which, in turn, reflect systems of beliefs
held by members of certain social groups. Luborsky (1990) showed how personal life
stories, far from constituting raw data, were highly processed according to situational,
professional and cultural norms such as narrative sequencing, metaphors used to
represent experience, and cultural "templates" representing overarching personal
meanings. These devices in turn reflect ways of thinking about the individual and human
life that are shared by a given community. Values and beliefs are related to characters in
stories and characters are evaluated according to categories such as the morality or
immorality, normality or abnormality, adequacy or inadequacy of their actions. Thus
narratives allow narrators to relate identities with acceptable and unacceptable behaviors.
However, when we look at narratives as a kind of discourse practice it becomes
clear that narratives do not merely “evaluate” actions and identities, they also contribute
to change or maintain them. It has been stressed how narratives are important in the
diffusion and strengthening of social prejudice (see van Dijk, 1987). Narrators create,
circulate and contest images about in-groups and out-groups by stressing similarities and
differences, by building interpretations on common contexts of experience. Many studies
underline the role of narratives in negotiating attitudes towards social categories such as
race (Bukholtz, 1999), gender (Kiesling, forthcoming) ethnicity (De Fina, 2000) and their

14
centrality in the creation of expectations and or myths about social experiences and stress
how it is often through them that individuals and groups construct their group
memberships.
To summarize the arguments developed in this section: story telling is a type of
discourse practice that involves the reflection, negotiation and constitution of identities at
three levels:
a. through styles of telling that derive from common uses of narrative resources;
b. through the projection, representation, and re-elaboration of social roles and
relationships;
c. through the negotiation of membership into communities that are seen as holding
common beliefs and values and behaving in specific ways.
In this book I focus on the latter two levels: the representation and elaboration of
social roles, and the presentation and negotiation of membership into communities.
However, I also explore the extent to which common uses of linguistic resources may
allow us to speak of a style of telling about the self that is either culturally specific or
typical of this particular group of speakers. I study identity in relation to social roles
through the investigation of linguistic choices and strategies that reflect ways of
presenting the self in relation to others, and ways of presenting the self in relation to
social experiences. For the analysis of the presentation of the self in relation to others I
look at social orientation. For the analysis of the projection/construction of roles with
respect to social experience, I look at agency as the represented degree of activity and
initiative that narrators attribute to themselves as characters in particular story worlds.
The focus of the analysis is on the linguistic mechanisms and strategies that help speakers
construct these of roles. Specifically, I focus on pronominal choice and voicing devices.
Through pronominal choices narrators express personalized or depersonalized views of
experience and construct themselves in stories as socially or personally oriented
individuals. Through use of voicing devices (reported dialogue and reported actions
within it) narrators convey degrees of initiative that they attribute to themselves and
others within crucial experiences such as the border crossing. At the same time, I explore
the extent to which common uses of narrative resources among members of this group of

15
speakers points to the existence of shared ways of telling and of constructing personal
experience.
I analyze the second level of identity construction: the negotiation of membership into
particular social groups through the study of categorization and identification strategies
used by narrators to introduce themselves and others in narratives. The questions asked
are: what are the salient categories for self and other description for this particular
community? What relationships do narrators establish between identities, actions and
reactions? Stories where categorization is prominent are typically argumentative in that
they re-present and evaluate adherence to or violation of social norms and the analysis of
identity at this level leads more directly than in other types of narratives, to implicit
mental representations and ideologies.
Thus, the expression of particular identities is tied in my analysis to the use of
linguistic elements and communicative and rhetorical strategies both in the representation
of characters within worlds of experience and in its negotiation with interlocutors. Such
linguistic phenomena and strategies belong to different (although interdependent) levels
of analysis: lexical, textual/pragmatic, and interactional. The lexical level refers to the use
of specific words or expressions. The textual pragmatic level refers to textual logical and
argumentative relationships both explicit and implicit. The interactional level refers to the
devices and strategies used by narrators to index their stances and attitudes both towards
their own texts and other interlocutors. Among other discursive mechanisms and
strategies and the linguistic elements that I have focused on in the analysis are the
following:
Lexical level
a. Pronouns, verbs and syntactic constructions indicating different degrees of
responsibility, engagement and activity both in relation to the story-world and
the storytelling world
b. definite descriptions, referential terms, pronouns used to identify self and others
Textual/Pragmatic level
a. different types of implicatures, implicit propositions, and presuppositions
b. relationships of consequence, cause or effect

16
c. oppositions between terms, actions or descriptions
d. relations between identifying descriptions and actions
e. cohesive devices and coherence relationships between textual segments and
between the text and the discourse surrounding it
f. argumentative relations between parts of the text
Interactional level
a. Devices and strategies encoding shifts between the story world and the
interactional world
b. Performance devices such as reported speech, tone, tempo, rhythm, repetition
conveying implicit stances towards characters or events
c. Devices and strategies indicating involvement or distancing with respect to
interlocutors and-or narrated events
Focus on linguistic phenomena does not imply that identities are directly related to
linguistic choices. Rather, identities emerge through the interplay between linguistic
choices, rhetorical and performance strategies in the representation of particular story
worlds, and the negotiation of such representations in the interactional world.
Negotiations involving not only narrators and interviewers but also other participants in
the interaction often show how the construction of particular identities is subject to
conflicts and reformulations. Identities are “achieved” not given, and therefore their
discursive construction should be seen as a process in which narrators and listeners are
constantly engaged. The analysis of identity as discursive work requires therefore
consideration of the discursive mechanisms through which narrators convey, negotiate,
contest, discuss, certain identities, of the ways in which such identities are negotiated
with other interactants, and of the relationships between identities and particular contexts
of experience, represented in and through story worlds. Thus, in the present study, for
example, I show how depersonalization in the representation of experience does not
merely result from the choice of individual or collective pronouns in the representation of
story characters, but also from differences between the pronominal choices suggested by
the interviewer in her questions (singular you) and the pronominal choices adopted by the

17
narrators in the telling (plural us). Similarly, the analysis considers for example how
particular descriptions for the identification of self or others as characters in the story
world are negotiated with the interviewer or with other interactants, but also
problematized through the use of performance devices that allow narrators to convey
implicit stances towards their characters. Thus, for example, narrators may identify
themselves as Hispanic in story worlds, but may at the same time convey conflicting
attitudes towards such categorization through the use of performance devices such as
voice, tempo, laughter, etc. They may also convey conflicting and contradictory identities
as they shift from one self-description to another in connection with different worlds of
experience. These shifts are apparent when we relate changes in pronominal choices or
identification devices to differences in the story worlds evoked. As story worlds represent
different life domains, narrators ascribe, contest and negotiate varying inventories of
identities.
Besides recognizing the existence of different levels of expression and
construction of identity in narrative, we need to also acknowledge the existence of a
variety of modes of emergence of identities within discourse. Identity can be given off,
conveyed, enacted, performed, discussed, contested by narrators. For example, when
narrators use particular linguistic devices such as first person singular or plural pronouns
to refer to themselves, employ or switch between linguistic codes, choose certain styles
such as topic association or “franqueza” (Farr, 2000), they may convey, or give off, their
identities simply by adhering to telling norms and styles that are shared by other members
of their communities. On the other hand, when narrators use particular accents,
impersonate, ventriloquate (Bakhtin, 1986), imitate, different voices, or employ other
kinds of devices that allow them to express footings (Goffman, 1981), they may be
“performing” identities. Finally, when narrators adopt identificational strategies for
themselves and others as characters in the story-world and/or as participants in the
interactional world, or when they critically present characters as breaking social rules,
they may be openly accepting, contesting and discussing identities. These ways of
impersonating, presenting, re-presenting identity are not necessarily exclusive of each
other, but appear to different degrees according to the objectives, topics and moments of
the interactional contexts and call for different tools of analysis. For example, identities

18
that are conveyed or given off can be related to representational choices such as the
choice of pronouns or type of verbs to depict the action in the story world. On the other
hand, identities that are negotiated and discussed can be related to the argumentative
attribution of certain actions to characters and to the use of explicit external evaluation
devices in stories.
Contexts are crucial not only for the kinds of identities presented, but also for the
ways these are presented. In the case of storytelling taking place within interviews, the
level of explicit negotiation of identities is important because tellers are often invited to
reflect on who they are and how they are defined by society and therefore they use stories
to accomplish socially acceptable self-presentations. Interview contexts encourage for
example both long monologic tellings in which little negotiation takes place and most of
the identity work is done by the narrator, and the telling of argumentative stories told to
support images of the in group or the out-group. For undocumented immigrants
storytelling within interviews represents an important occasion for the negotiation of their
presentation of self, since their opportunities to be heard by social actors who don’t
belong to their group are limited. Interviews also represent interactional events where
interviewers and interviewees often try to make sense of social reality through explicit
analysis of social circumstances and roles. This does not imply that identities are always
openly discussed since narratives told in interviews are also often performed or arise
spontaneously in connection with points that are being discussed, but simply that
narrators rely more heavily than in in-group conversation on explicit discussions about
their identity. In conclusion, the choice of an interview context as a site for the
investigation of identity makes relevant the analysis of explicit, argumentative modes in
the presentation of self, and of their negotiation with an interviewer who, although
sympathetic, is not a member of the group. The question of the interviewer’s role in the
kinds of identities that emerge in this context is also, therefore, highly salient.
3. Local and global contexts
The arguments discussed in the previous sections converge on the idea that
identities are situated in historical, social, and interactional contexts. Looking at identity

19
in social constructionist terms, I have argued that identities are the result of ‘discursive
work’ (Hall, 2000), and that there can be no single identity, but a constellation of
identities often conflicting with each other, a repertoire that is available to individuals and
from which they draw when presenting and representing who they are. I have also
discussed the fact that selection within the repertoire of possible identities within and
outside story worlds crucially depends on the context. But which context is pertinent for
the construction and analysis of specific identities? Social constructionist approaches
stress the plurality of identities that may be displayed and their context sensitivity, but
often leave open the question of how local and global identities interact with each other
and what kinds of contexts are pertinent for their analysis. Identities constructed through
narratives may be related to a multiplicity of contexts.
The local context situates narratives within the interaction at hand. Conversation
analysts have pointed to the fact that narratives are told by speakers to audiences and that
narrators introduce or close their stories following the constraints imposed by other
interactants (Jefferson, 1978 ; Polanyi, 1985) through clear displays of relevance and
adequacy of their content to the rest of the interaction. They have shown how these texts
develop according to the presence or absence of audience reactions (Goodwin, 1986;
Duranti, 1986; Polanyi, 1985) and are built based on evaluation of audience expectations
(Sacks, 1992 a, b). By the same token, identities are locally produced since narrators
position themselves and enact specific identities that are at least partly the product of
ongoing negotiation processes and therefore create or refute particular alignments and
participation frameworks with other speakers and listeners (Goodwin, 1986; Goodwin,
M., 1993). At the local level of interactional positioning (Bamberg, 1997) narrators may
engage for example in discursive work aimed at projecting their moral identity as
collaborative, or mature or knowledgeable individuals. At another local level, they may
stress their dependence on the sympathy of the listener, or conversely, their independence
and individuality. Participants, including interviewers (Wortham, 2001), may in turn be
oriented towards the construction or contestation of such identities. Furthermore, as we
have seen, narrative structure and development, story content and therefore also the
identities enacted in specific interactions display clear links with the interactional
practices in which they are inserted and with the roles of the participants in them. The

20
identity work done through stories told in interviews may differ dramatically from that of
stories told in conversation because of the distance in the relationship between
interactants. Similarly, stories told within other discursive practices such as sermons
(Ochs & Capps, 20001) or educational discussions (Moita Lopes, forthcoming) may
differ as the identities that speakers and audiences produce and reproduce crucially relate
to the circumstances of production. In other words, identities are bound to interactional
contexts through their connections with participant frameworks and speech events.
At another level, the context of narrative identities is given by much wider social
circumstances which constitute the broad framework for the attribution to self and others
of membership into ethnic, social, economic categories. Yet at another level, the story
world in which both interactional frameworks and worlds of experience are re-presented
provides a further, represented, context.
Although it is true that the expression and negotiation of identities may connect
these different contexts, it is also true that the analysis of the way contexts interact with
identities is necessarily selective and interpretive. Focus on local identity displays may
take the analyst deeper and deeper into the dynamics of a specific interaction and of the
participation framework of an event while focus on the narrators’ management of
represented identities may take her further and further away from the local context and
deeper into the relationship between self representation and experience of the world. As a
result, the analysis of identities may rely more or less heavily on the local or global
context as explanatory constructs. Reliance on the local context to explain and frame
identities is typical of ethnomethodologically and C.A. oriented social constructionist
approaches (see Antaki & Widdicombe, 1998). These methodologies posit that the
pertinent level of analysis is the local construction of identities as signaled by the
orientation of interactants. In the analyses inspired by those approaches, global identities
only become pertinent as they are signaled, enacted, or negotiated in the interactional
context and identities are constructed and accomplished in the process by speakers and
other interactants.
However, a reduction of the context to sequentially and locally accomplished
actions does not allow a full appreciation of the links between locally expressed identities
and global phenomena of identity formation since their complex relationships are

21
mediated through wider discursive and social practices that may not necessarily be
apparent in individual interactions, or signaled by speakers’ orientation towards them.
Thus, for example, the frequent switches between yo (I) and nosotros (we) exhibited by
Mexican immigrants when describing themselves as actors in story worlds, may go
unnoticed within the local interaction, but acquire significance when analyzed at a more
general level as a strategy of positioning vis a vis life experiences that constitute a threat
to their integrity and sense of self. Similarly, the recourse to specific ethnic
identifications to refer to oneself or others in story orientations, may evoke no specific
reaction among participants and no orientation signaling their significance. However, an
analysis of a number of stories and background information on the role of ethnicity in
American society and in immigrant life, may shed light on the significance of those story
identifications for Mexican immigrants. Thus, it is argued in this book that the analysis of
group identity in stories cannot rely exclusively on the local context, but needs to take
into account its complex relationships with the wider context of social and discursive
practices and their dynamic connections with the discourse of specific actors.
Story telling is a discursive practice marked by its insertion within certain
conditions of production and reception. The sociologist Bourdieu (1982) explains context
dependence in terms of markedness. Using as an example the words in a language, he
says : "The dictionary word has no social existence: in practice it only exists as immersed
in certain situations," (p. 16). In the same way as words become socially charged as soon
as they are uttered, so do utterances and longer stretches of talk since they are inserted
within social and interactional practices, and therefore within other contexts. One way in
which we may connect specific discourse instances to macro social circumstances is
through analysis of the "conditions of production and reception" of discourses (see
Pecheux, 1969 and Pecheux & Fuchs, 1975). These are not something external to
discourses, but something that shapes them. Conditions of production include the
institutional framework, the ideological apparatus within which certain discourses are
produced, mental representations, the political situation and force relationships among
social groups, intended effects and strategies. The former are not simply 'circumstances'
that exert constraints on discourse, rather, they constitute it and characterize it (Gardin,
1976). These wider social factors are contextualized in storytelling through the use of

22
linguistic elements and strategies that connect for example specific instances of discourse
to wider ideologies and mental representations, social behaviors and social relations.
Again, an understanding of the role that narratives have in conversation and of the
meanings that are transmitted and negotiated through them would not be possible without
reference for example to implicit and explicit beliefs and values held by most members in
the community, even if they have not been brought to bear in the particular interaction, or
participants do not orient to them. The analysis of stories, and particularly the analysis of
identities in stories, cannot avoid incorporating an analysis of ideologies (van Dijk, 1998)
and beliefs. Thus, in this book, the ways in which identities are related to actions in
stories is connected to schematic representations about self and others that appear to be
shared by group members, and implicit evaluations of actions are studied against
common moral stances. These representations and stances are often discussed in the
interviews and constitute a frame of reference for the evaluation of characters.
Another way in which local contexts connect local identity displays with wider
group relations and mental representations is through intertextuality (see Kristeva, 1980).
Beyond the interaction at hand narrators establish intertextual connections not only with
other stories such as other narratives about migration, but also with other “discourses”,
such as dominant images about immigrants circulated through institutions and media.
While responding to interlocutors, narrators also respond to discourses that are not
necessarily uttered in their presence, but that are being socially circulated. In brief, texts
produced in specific circumstances are also part of a discursive chain that links together
texts produced at different moments and by different people. Thus, when immigrants
present certain images of themselves or apply definitions to others, they are often reacting
to what the media, or other social actors say about them. Their stories are often designed
to counter negative images or to incorporate commonly held prejudice about competing
groups. Therefore, interactional negotiations about identity cannot be explained without
reference to these external voices.
Because the focus of my work is on the connection between local expression of
identities and group representations about identities, the local context is taken as an
explanatory and constitutive frame for the expression of identities in so far as it connects
to wider social contexts. For this reason, not much attention is paid to the personal

23
dynamics between interviewer and interviewees, or between interviewees, which
certainly belong to the level of interactional positioning. Furthermore, phenomena are
seen as significant if they show patterns that occur in different stories precisely because
the emphasis is on shared processes of construction and representation.
In chapters three, four, five, and six, I explore the connections between local and
global identities in detail. However, I devote the next chapter to the description of some
aspects of the migration of Mexican workers (particularly of undocumented ones) to the
United States and to a presentation of the subjects, data and methodology of this study.

1
Chapter 2
The social phenomenon: Mexican migration to the U.S.
Introduction
In this chapter, I discuss some aspects of what I have called “the conditions of
production” of the narrative discourse of the Mexican immigrants who were interviewed
for this study. I present an overview of the social phenomenon of Mexican undocumented
migration to the United States looking at its size, origins, and motives in order to explain
the position of undocumented immigrants within U.S. society. I introduce my informants
and their specific social economic background, and give some information on the nature
of the discourses on migration circulated by the media and in the political arena since, as
I argue, these discourses constitute an intertextual domain with which immigrants (and
the interviewer) establish connections in their narratives and arguments. Finally, I
explain some of my choices in terms of fieldwork, data collection procedures and
analysis.
1. Mexican undocumented immigrants to the United States
The migration of millions of workers from Mexico to the United States has been a
recurrent phenomenon in the history of the two countries and a focus of concern, debate,
and conflict on both sides of the border. Mexican workers started migrating to the United
States to work in the agricultural sector in the 19th century, shortly after the signing of
the Guadalupe Hidalgo Peace Treaty (1848), which sanctioned for Mexico the loss of a
great part of its Northern territories. The flux of workers has never stopped since then.
Scholars of Mexican migration (Chavez, 1992; Rouse, 1991; Gaxiola, 1991)
suggest that because of the geographical proximity of the two countries and because of
this long history of migration, Mexican workers have a different attitude to moving into
the U.S. than migrants from other countries. First, they see the possibility of crossing the
border as an opportunity that has been exploited again and again by generations of people
in their own family or village, as something that has a precise historical tradition, and as a
resource that is always present in moments of economic difficulty. Secondly, and as a

2
consequence of this, they see themselves as forming part of a transnational labor market,
not as a labor force whose place of employment is restricted to their own national
boundaries.
Another interesting aspect, also related to the historical and traditional character
of Mexican migration, is the fact that migration is largely a social process, much more so
than an individual one. Mexicans hear about life in the United States from returning
migrants, they usually discuss their decision to leave with members of their family or
friends, and they often leave for their journey in groups.
The geographical proximity of the two countries also facilitates migration, since
Mexicans, unlike other Latin Americans, only need to cross one border in order to get to
their destinations, and many of them cross it more than once in their lifetime. Migration
to the United States is, in sum, a widespread process in the history of Mexico as a country
and part of the shared experiences of people from particular villages, cities, and states
within it.
1.1 Number and origin of Mexican undocumented workers in the U.S.
It is difficult to estimate what proportion of the Mexican immigrants who work in
the United States are undocumented. Estimates vary greatly and often are based on
conjecture. Gaxiola (1991) reviewed a number of studies conducted both in the United
States and in Mexico on the volume of the undocumented population between 1970 and
1980, and reported that estimates varied between 3 and 13 millions of undocumented
workers in general, with varying proportions of Mexicans. Her review points to the fact
that data on the presence of undocumented workers are in many cases unreliable since
they often respond to the political aims of those who provide them.
A study conducted by Lesko and Associates in 1975 that had been requested by
Chapman, then INS commissioner, estimated for example that the number of illegal
Mexican workers in the United States was 5,204,000. This estimate was widely criticized
because of the unreliability of the methods used to calculate it (Heer, 1990). Research
conducted between 1977 and 1979 by the Mexican Centro Nacional de Información y
Estadística del Trabajo (CENIET) (published in 1982) provided more reliable data. The

3
research was conducted among Mexicans who were being expelled from the United
States, and Mexicans who were back in Mexico but had recently been living in the United
States. The study concluded that the number of undocumented Mexican workers in the
United States could be estimated at 990,719. Heer (1990, p. 51), who surveyed 10 studies
on the presence of Mexican undocumented workers between 1973 and 1980, proposed a
figure of 1,781,000 for 1980.
The figures are probably higher today, since immigration seems to have increased
in the eighties at both global (Papail & Arroyo, 1996, p.16-17) and local levels. If we
look at reports of immigration in individual states like California, the immigration from
Mexico to the United States has constantly increased in the last thirty years (Wayne,
Chavez, & Castro, 1982, p.13). According to Chavez (1994, p.52) in the eighties it was
calculated that between 200,000 and 300,000 undocumented workers from all countries
stay in the U.S. each year. The number of immigrants crossing the border is also directly
related to the economic situation in both countries and the variations in the real salaries
(Hanson & Spilimbergo, 1997, p.7). Thus, given the dramatic fall in the real salaries of
Mexican workers in the nineties, it is also reasonable to suppose that the flux of
immigrants from Mexico has increased in the same period.
Studies of Mexican migration (Gamio, 1969 a and b; Bustamante, 1979; Morales,
1981; Gaxiola, 1991) also agree on the fact that most of the Mexican undocumented
workers traditionally have come, and still come, from a limited number of states of the
Mexican Republic, namely: Guanajuato, Jalisco, Chihuaua, Zacatecas, Michoacan. Other
states where migration is a significant phenomenon are Durango, San Luis Potosí, Baja
California. This means that the existence of a tradition of migration is a strong factor in
the diffusion and establishment of the process.
Most of these states are not Border States but occupy the central region of the
country. It is also interesting to stress that they are not the poorest states in Mexico. This
shows how the decision to migrate is not only determined by economic factors, but also
by the presence of relatives and friends on the other side of the border, and the existence
of a local tradition. On the other hand, surveys of Mexican undocumented population in
the United States agree on the fact that most of the undocumented workers choose as
their destination the South-Western United States. Heer (1990, p.53-54) reports data from

4
the 1980 Census, according to which it was calculated that 67% of all undocumented
Mexican workers were in California, while another 13% were in Texas. Thus, these two
states constituted together the destination of 80% of all undocumented workers in the
U.S. He also quotes the CENIET (1982) study as confirming these data for documented
and undocumented workers, since it reported that the states where most of the Mexicans
residing in the U.S. were found were respectively: California (49.2%), Texas (22%),
Illinois (8.6%), New Mexico (2.0%), Colorado (1.9%) and Arizona (1.8%).
1.2 Reasons for migrating and sociocultural characteristics of Mexican immigrants
According to the CENIET study (1982), about 78% of the undocumented workers
interviewed had a job in Mexico before emigrating. This seems to corroborate the
hypothesis proposed by different authors that one of the main reasons for migrating is not
unemployment, but the desire to improve one's economic situation and the need to get a
better salary. Many undocumented workers report, in fact, that their salaries in Mexico
are insufficient to provide for their basic needs, while in the United States their income is
more substantial. Even though they sometimes earn less than the minimum wage, they
can still send money to their family back home. Wayne (1978 a) reported a difference of
up to 13 to 1 in the salary earned by an immigrant in the United States and in Mexico.
Chavez (1992, p.29-33) also quotes other reasons for migrating reported by the
undocumented workers he interviewed in California. Among them are the desire to
follow "the immigrant dream" of getting a better life socially and economically,
overcoming family conflicts, or wanting to satisfy a need for adventure. On the whole,
nonetheless, most authors agree that the main motive for migrating is, in the case of
Mexican undocumented workers, economic need (Gamio, 1969a; North & Houston,
1976; Morales, 1981; Gaxiola, 1991; Chavez, 1992).
What is the social profile of Mexican undocumented migrants? In a study
conducted by the Consejo Nacional de Población (CONAPO) (1987, p. 73-77), which
surveyed 9,631 Mexicans who were sent back to Mexico from border cities because of
lack of proper documentation, it was found that the larger groups of immigrants were
composed of people between 15 and 29 years of age, steadily decreasing after that age.

5
According to the same study, more men migrate than women, hardly surprising
information if we think that women whose age is between 15 and 29 are in their
childbearing years, and therefore have less mobility than men.
According to Morales (1981), who bases her conclusion on a survey of several
studies of Mexican undocumented migration, another characteristic of migrant workers is
that their level of education is low. The majority of the workers interviewed in the
CONAPO (1987) study had completed only elementary school. Chavez (1992) also found
that most of the workers he interviewed at different campsites around San Diego had little
education. Gaxiola (1991) reports that 45% of the 200 hundred undocumented workers
detained at the border that she interviewed in Laredo, Texas, had completed elementary
school, while another 20% had completed between 3 and 4 years of primary education.
Data on the occupation of migrant workers in the United States are more difficult
to compare, since most studies have been conducted in the South of the United States and
their results do not necessarily represent the situation in other areas. Different studies
found that the majority of the Mexican undocumented workers are employed in the
agricultural sector (Bustamante, 1979; North & Houston, 1976; CONAPO, 1987). The
CONAPO study also found that the most common occupation after agriculture was
industry. According to Wayne, Chavez, & Castro (1982, p.29), in Southern California
Mexican immigrants can be found holding unskilled and semi-skilled jobs in virtually
every sector of the region’s economy. These authors also suggest that although
agriculture was still an important area of employment for Mexican workers, there was a
trend towards moving from agriculture to different types of jobs and that most workers
were employed in small firms.
These studies confirm that Mexican undocumented workers are employed as
unskilled workers in most sectors, and that their earnings are low and their work
conditions often worse than the ones that American workers would accept. It is
nonetheless likely that the sectors of employment will vary with the areas to which these
workers migrate.
To summarize, most studies agree on the fact that undocumented workers are
mostly young, between 20 and 29 years old, that their educational level is low, and that
they are mostly occupied as unskilled workers.

6
1.3 The migration process
We have already seen that migration is usually a social process, in the sense that it
often involves contact with people who are (or were) in the United States, and it is also
often based on an established local tradition. Another aspect of the migration process that
has been underscored by many authors is its temporal nature.
According to Morales (1981, p.182-183), the majority of the Mexican workers
interviewed in those studies stayed in the United States for less than one year. This is due,
according to the author, to the cyclic nature of agricultural work. These conclusions are
again largely based on the situation of Mexican immigrants in the South of the U.S. The
case of those who manage to reach the northern areas of the country and get a job in the
industrial sector is different. The trip back to Mexico is more expensive and difficult, and
therefore they probably stay longer there and some of them even bring their families.
Not much is known on the percentage of Mexican undocumented workers who
stay in the United States, but it is generally accepted that their proportion is smaller than
the proportion of those who come and go. This has led to a vision of Mexican
undocumented workers as homing pigeons who do not develop any ties with the host
country. Chavez (1994) strongly argues against this vision saying that many Mexicans
stay in the United States and that the migration process has an inevitable effect on the
people who undertake it, whether they go back to their country or not, since they
ultimately develop multiple senses of community membership.
2. The subjects of the study
The data for this study come from sociolinguistic interviews with 14 Mexican
immigrants living in Langley Park, Maryland. Most of them came from the same village
in Mexico, lived in houses that were not too far apart, and often visited each other on
weekends. Twelve of the fourteen people that I interviewed were born in El Oro, Estado
de México, while two of them (César and Sixtoi
), were born respectively in Mexico City

7
and San Luis Potosi. Their age varied, but most of them were young since 9 were in their
twenties, 3 were in their thirties, and 2 were in their fifties.
The immigrants belonged to 4 different households, or ‘domestic groups’
(Chavez, 1992, p.129). Chavez describes domestic groups as houses where people live
together but do not necessarily constitute a family. This arrangement is common among
immigrants for two reasons: first, because newly arrived immigrants often are housed by
friends or relatives who are already living in the country, and secondly, because it allows
them to share expenses related to rent and utilities.
Domestic groups have different types of compositions; in the case of the
immigrants I interviewed, there were 4 domestic groups which were all formed either by
members of the same extended family, or by relatives and friends. Leo, for example,
lived with his wife and brother. Silvia lived in an apartment with 7 other guests: Omar,
her brother, Raquel and Lourdes, her cousins, who in turn were sisters, and 4 other young
people who were unrelated to them. So the domestic groups that I visited were
combinations of family members living together, and family members and friends.
Among these immigrants, the general level of education was higher than the one
reported in other studies since more than half the people I interviewed had studied
beyond elementary school; in fact 2 had started university before coming to the United
States, 2 had completed high school and 4 had studied a technical career after high
school, while another 2 had started, but not completed, high school.
This higher level of education reflected the fact that most of my informants did
not belong to the poorest layers of society. Most of them could be classified as middle
class or lower middle class. Among the women: Silvia had worked as a computer
specialist in a firm in Mexico, Laura as a receptionist, Raquel as an employee of the court
house, María had been owner of a restaurant, while Virginia had not worked outside the
house. Willi had worked at the Nissan plant in Mexico City, Cesar had worked as a
waiter, Oscar had been employed as a shop assistant. Not all these informants had been
employed before coming to the United States; in fact Leo had never worked because he
had left Mexico when he was very young, and Juan attended school before he came to the
U.S. with his mother. Ciro told me that he and his wife were not well off in Mexico and
that they did not own a house, but that they were not poor. He said that he had worked for

Identity In Narrative A Study Of Immigrant Discourse

Identity In Narrative A Study Of Immigrant Discourse

Recommended

Recommended

More Related Content

More from Erin Torres

More from Erin Torres (20)

Recently uploaded

Recently uploaded (20)

Identity In Narrative A Study Of Immigrant Discourse