4.2,; a,

$.13.

1287': ‘1‘

:IIC,‘ ,

.. Ll- . :I

..|‘ .7

3

FE

UNI.

1|...“

ll..-

u T 1....

 

 

.wfi.

‘N,

. wndud‘ “14V .

.,v‘1

m-‘knzfiiavxfir

.v

psi 17.
I? 5.1:

4.9.»,

 

.51...
.
r
z.

0.

3

v.3:

...
x, 41.

w

r
v humus Euhmfln‘ha .2
runaufimw
ugh? .ursl
3 .2». .wt ,
Eng
k...

Am

 

 

3w ,

    

.13. t... I
burncx.“ tin . .50..

n _.Ma.nm§ufivvn.....

   

t 'f‘._ .
Mt)l.\

. Lifl
0700:

arm/WM?

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

BILINGUALISM AND IDENTITY: A SOCIOLINGUISTIC
STUDY OF NATIVE SPANISH-SPEAKING CHILDREN IN AN
AGRICULTURAL COMMUNITY IN CALIFORNIA

presented by
CATHERINE L. FLECK

has been accepted towards fulfillment
of the requirements for the

PhD degree in English

AOL/[We 22* Aféwé‘éwd

Debra Hardison

wgfiflcoj
/ 1/
July 8, 2003

MSU is an Affirmative Action/Equal Opportunity Institution

 

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

 

DATE DUE

   
 

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6m cJCIRC/DateDuepss-ms

 

BILINGUALISM AND IDENTITY: A SOCIOLINGUISTIC STUDY OF NATIVE
SPANISH-SPEAKING CHILDREN IN AN AGRICULTURAL COMMUNITY IN
CALIFORNIA
By

Catherine L. Fleck

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

DOCTOR OF PHILOSOPHY
Department of English

2003

ABSTRACT
BILINGUALISM AND IDENTITY: A SOCIOLINGUISTIC STUDY OF NATIVE
SPANISH-SPEAKING CHILDREN IN AN AGRICULTURAL COMMUNITY IN
CALIFORNIA
By
Catherine L. Fleck

This sociolinguistic study focused on the relationship between proficiency in
English as a second language and identity, for native Spanish-speaking children in an
agricultural community in California. The expected results were that there would be a
strong positive correlation between the two main variables in the study—English
proficiency and identity. Data was drawn from 41 subjects, who were generally grouped
into three age groups: six year olds, eight year olds, and ten year olds. The range of
grammatical features that were operationalized as English proficiency reflect deep
structural differences between Spanish and English; these features cover syntactic
configuration (Noun-noun compounds), morphological attachment (tense doubling), as
well as several different functions of a core grammatical element (articles). Identity was
operationalized as a bundling of social factors—a constellation of symbolic information,
and preferences, focusing on the following 5 general categories: Favorite food(s),
favorite music, sports, holiday foods eaten by the family, preference for place—Mexico
or the local California community. In addition, participants were asked about any
English-only speaking associates they have. Scores were derived for both variables,
which were then correlated.

The results obtained from the correlation analysis were not as expected. The

results showed very little correlation: (.l 19). Given these unexpected results, separate

ANOVAs were run on the sub-categories of the identity variable and English proficiency,
with the aim of identifying trends in the data. Not one of these ANOVAs was
significant. However, one of the sub-categories of the identity variable, “number of
English-only speaking associates” showed somewhat significant results, (0.076). This
sub-category, referred to as E0, was reconsidered as an measure of social networks; as
such, it appears to be a good predictor of English proficiency, similar to several other
factors used in this study—age, sex, and SES—-—all of which are large group identity
factors.

Ethnolinguistic identity theory emerges as important to the discussion of the
results of this study; this theory claims that people in a subordinate group will hold on to
their ethnolinguistic identity, if certain conditions exist. One of these conditions,
ethnolinguistic vitality, is a measure that is based on several components, including 1)
demographic strength (group numbers, birthrates, migration, concentration and
distribution in the community), 2) institutional support for the language (in the education
system, the media, the government), and 3) status (social, political, economic, linguistic
prestige). Given the demographic data presented in this study, as well as the descriptions
of the educational support programs available within the school district (bilingual and
migrant programs), combined with the results of this research, it would appear that the
ethnolinguistic vitality of these participants is high, resulting in these participants holding
fast to a Mexican-oriented identity. The question of how this vitality affects whether or

not these children become fully proficient in English is a matter for future research.

This work is dedicated to Daniel Vico, who has been with me the whole time.

iv

ACKNOWLEDGMENTS

There are many people to acknowledge in the culmination of this work—many
more than can be named in this brief statement. Those who particularly stand out are the
following: Professor Dennis R. Preston, who directed and co-chaired this research, and
who has been a great source of information, knowledge, learning, and guidance for many
years—throughout my tenure as a graduate student. I am very grateful to have been one
of his students; the support and encouragement he has shown, to me and to his many
other graduate students, has been exceptional. Through working with him, I have learned
what a true mentor is. I would like to also thank Professor Debra Hardison, the other co-
chair of this research, for her consistent support and guidance throughout this long
process of navigating the Ph.D. She has demonstrated time and time again her dedication
to her students, and to their learning, and I very much appreciate her sensitivity to how
arduous the road can be. In addition, I would like to express my appreciation to the other
two professors who served on my dissertation committee. These are Alan Beretta and
Deogratias Ngonyani. Alan has, over the years been a wonderful friend and colleague,
one who has always been willing to engage in critical discussions on many issues; I have
learned a lot from him, and appreciate not only his support as a professor, but his
friendship. Deo very graciously joined my committee late, when another committee
member left the team. I very much appreciate his interest in my project, and have gained
from the conversations we had over the issues involved in this work. And finally, I

would like to thank Professor Anne Violin-Wigent, who was the outside reader for the

defense of this dissertation; I very much appreciate the care with which she read this
research, and the detailed comments she gave me on my work.

Beyond these people, I would also like to thank my colleagues at the Writing
Center. Professors Janet Swenson and Patricia L. Stock have been not only wonderful to
work with—very supportive through good and difficult times—but have become
wonderful friends as well. In addition, there are many graduate students, both colleagues
and clients, that I have appreciated working with over these years; thanks for all the
support and encouragement. And last, but certainly not least, I must mention my husband.
Nevin Leder, who has been my partner in crime for all of these years of graduate work.
Nevin has been a constant source of intellectual stimulation for me, as well as a
wonderfully supportive partner. Words really can’t express my appreciation for his
friendship and love, his critical mind, and his sharing of the love of linguistics with me.

And now together, we turn to our new adventure; adelante siempre.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................... ix
LIST OF FIGURES .............................................................................. xi
INTRODUCTION ........................................................................... 1
RESEARCH QUESTIONS ....................................................................... 7
HYPOTHESIS ................................................................................... 8
ORGANIZATION OF THE DISSERTATION ..................................................... 8
LITERATURE REVIEW .................................................................. 10
BILINGUALISM ................................................................................ 10
IMMIGRATION/MIGRANCY .................................................................. 16
THE COMMUNITY ............................................................................. 22
IDENTITY ...................................................................................... 31
Identity Operational ized .................................................................. 3 8
LANGUAGE FEATURES STUDIED ............................................................ 39
Noun/noun Compounds ................................................................... 42
Articles ....................................................................................... 44
Tense Doubling ............................................................................. 50
METHODOLOGY ........................................................................... 53
PARTICIPANTS ................................................................................. 53
SCHOOL SITE .................................................................................. 57
PRELIMINARY PREPARATION FOR THE STUDY ............................................. 58
METHODS ...................................................................................... 59
Elicitation Methods ........................................................................ 60
Scoring ...................................................................................... 64
RESULTS AND DISCUSSION ........................................................... 67
FREQUENCY DATA: TOTAL IDENTITY AND TOTAL ENGLISH PROFICIENCY .............. 67
ANOVAS: TOTAL IDENTITY SCORE AND FIVE FACTORS ................................. 69
A Closer Look at 3 of the 9 Identity Sub-categories .................................. 72
ANOVAS: TOTAL ENGLISH PROFICIENCY SCORE AND FIVE FACTORS .................... 74
CORRELATION: ENGLISH PROFICIENCY AND SCHOOL/TEACHER SCORES .................. 79

Vii

 

 

CORRELATION: TOTAL ENGLISH PROFICIENCY AND TOTAL IDENTITY .................. 81

Unexpected Results: Assess English Proficiency Variable ........................... 82
Unexpected Results: Assess Identity Variable ......................................... 86
Non-significant Trends in the Data ................................................... 87
Crosstabulation of Total Identity Scores by Grade ................................. 91
Validity of the Identity Variable ...................................................... 93
SUMMARY TABLES ........................................................................... 94
A Closer Look at the E0 Variable ...................................................... 96
Large Group Identity Factors vs Factors Entailing Individual Choice ............ 99
ETHNIC IDENTITY: CAN WE ASSESS THIS? .................................................. 100
IMPRESSIONS: PEOPLE, FOOD, AND VAMPIREs ............................................. 104
RETURN TO MAIN STREET .................................................................... 106
CONCLUDING REMARKS, ON THE GENERALIZABILITY OF THE FINDINGS ............... 1 16
APPENDIX A: WATSONVILLE OCCUPATION CHARACTERISTICS ........................... 122
APPENDIX B: INCOME ELIGIBILITY GUIDELINES FOR FREE LUNCH ....................... 123
APPENDIX C: NOUN/NOUN COMPOUNDING ................................................. 124
APPENDIX D: ARTICLE USAGE ............................................................... 141
APPENDIX E: TENSE DOUBLING .............................................................. 161
APPENDIX F: ARSMA-II (ACCULTURATION SCALE) ..................................... 165

viii

List of Tables

 

Table 1: Continuum of Generalizability .......................................................................... 15
Table 2: 2000 Census Data on Race/Ethnicity for Watsonville, CA, and US .................. 36
Table 3: 2000 Census Data on Employment Status .......................................................... 37
Table 4: 2000 Census Data on Class of Worker ............................................................... 38
Table 5: 2000 Census Data on Income for Watsonville, CA, US ..................................... 39
Table 6: 2000 Census Data on Educational Attainment for Watsonville, CA, US .......... 40
Table 7: Possible Range of Responses for Tense Doubling Data ..................................... 62
Table 8: Participant Age and Sex, by Grade ..................................................................... 64
Table 9: Birthplace of Participants and Parents ................................................................ 65
Table 10: Migrant Status ................................................................................................... 65
Table 11: Free/Reduced Lunch Status .............................................................................. 66
Table 12: Home Language ................................................................................................ 67
Table 13: School Site Demographics ................................................................................ 68
Table 14: Scoring Rubric for Language Features Studied ................................................ 75
Table 15: Scoring Rubric for Identity Data ...................................................................... 76
Table 16: Total Identity Score—Frequency ..................................................................... 78
Table 17: English Proficiency Score—Frequency ............................................................ 79
Table 18: ANOVA, Total Identity Score and Sex ............................................................ 80
Table 19: ANOVA, Total Identity Score and Age ........................................................... 81
Table 20: ANOVA, Total Identity Score and Lunch Status ............................................. 81
Table 21: ANOVA, Total Identity Score and Home Language ........................................ 82
Table 22: ANOVA, Total Identity Score and Migrant Status .......................................... 82
Table 23: Favorite Soccer Team—Frequency .................................................................. 84
Table 24: Have Favorite American Football Team—Frequency ..................................... 84
Table 25: Prefer Soccer or F ootball—F requency ............................................................. 85
Table 26: ANOVA, English Proficiency Score and Sex .................................................. 85
Table 27: ANOVA, English Proficiency Score and Age .................................................. 86
Table 28: Crosstabulation Of Teacher’s Overall English Score ........................................ 86
Table 29: Crosstabulation of School Initial English Scores ............................................. 87

Table 30: Crosstabulation of School Initial Spanish Scores ............................................. 88
Table 31: ANOVA, English Proficiency Score and Lunch Status ................................... 88
Table 32: ANOVA, English Proficiency Score and Home Language .............................. 89
Table 33: ANOVA, English Proficiency Score and Migrant Status ................................. 89
Table 34: Correlation—Total Identity Score/three sub-categories of Engl. Prof. Score .. 94
Table 35: Articles—Frequency Data ................................................................................ 94
Table 36: Noun-noun Compounding—Frequency Data ................................................... 95
Table 37: Tense-doubling—Frequency Data .................................................................... 96
Table 38: Mean Scores for Identity Sub-scores with trends opposite expected ............... 98
Table 38: ANOVA—Better/Me'xico or Watsonville and English Proficiency Score ....... 98
Table 39: ANOVA—Favorite Soccer Team and English Proficiency Score ................... 99
Table 40: Mean Scores for Identity Sub-scores with non-linear trends ............................ 99
Table 41: ANOVA, What family ate for Thanksgiving and English Proficiency Score 100
Table 42: ANOVA, Pref. for Soccer or Amer. Football and English Proficiency Score 100
Table 43: ANOVA, Preference for living in Mexico or Watsonville ............................. 100
Table 44: Mean Scores for Identity Sub-scores with expected trends ............................ 101
Table 45: ANOVA, Favorite singer and English Proficiency Score .............................. 101
Table 46: ANOVA, Favorite American football team and English Proficiency Score .. 102
Table 47: ANOVA, Favorite food and English Proficiency Score ................................. 102
Table 48: ANOVA, English-only Friends and English Proficiency Score ..................... 102
Table 49: Crosstabulation of Total Identity Scores by Grade Level .............................. 103
TableSO-a, b, c: Crosstabulation Of English Proficiency by Grade Level ...................... 103
Table 51 : Summary—ANOVA Results for Identity/English Prof. + Five Factors ........ 106
Table 52: Summary—ANOVA results for Engl. Prof. + 9 sub-categories of Identity... 107

List Of Figures

Figure 1: Correlation of Scores—School English and English Proficiency .................... 91
Figure 2: Correlation of Scores—Teacher English and English Proficiency ................... 92
Figure 3: Correlation Of Scores—Total Identity and English Proficiency ...................... 93
Figure 4-a: Watsonville Plaza ......................................................................................... 1 19
Figure 4-b: Trash Bin—Spanish & Figure 4-c: Trash Bin—English ........................... l 19
Figure 5: Bilingual Watsonville City Banner ................................................................ 120
Figures 6-a & b: El Vaquero, and Child on Pick-up Truck ............................................ 121
Figure 7: Social Security Administration Office and J oyeria ........................................ 122
Figure 8: Super Grullense Taqueria Atotonilco ............................................................. 123
Figure 9: Stevie G’s Meats ............................................................................................ 125
Figure 10: Taqueria on Main Street ............................................................................... 126

xi

INTRODUCTION

 

Languages do not recognize political boundaries. With fewer than 200 countries
in the world today, and an estimated number of languages between 6,000 and 10,000, it is
inevitable that different languages will come into contact. In fact, one might argue, given
such pressures in the world as increasing population movements, due to immigration and
a globalized world economy, that the phenomenon of languages in contact is
fundamental to the human experience. One obvious result of language contact is
bilingualism/multilingualism. While the estimate for the number of bilingual or
multilingual speakers in the world today varies according to one’s definition of
bilingualism, the general consensus is that there are many more bilingual people than
monolingual. Given this, it would not be unreasonable to argue that understanding
bilingualism is a key component of any comprehensive view of the linguistic nature of
human beings.

While there are multiple approaches to the study of bilingualism, most researchers
argue that this phenomenon can only be understood in relation to social context. It is
understood that bilingualism is quite complex, a result of the interaction of many factors,
social and linguistic; bilingualism is seen to be a behavioral pattern of linguistic practices
that are modified by, and relative to, many aspects of the Speakers’ lives. For the most
part, bilingual individuals live within communities of languages in contact—bilingual
communities that are such because they are composed of people who have reasons for
being bilingual. The complicated interplay of social and linguistic factors involved in
bilingualism makes this a fascinating area of research, but also, an extremely difficult

one.

One area of focus within the study of bilingualism is the formation and re-
formation of ethnic identity within bilingual communities. This research considers
ethnicity and identity to be tightly bound, generally regarding ethnicity as a construct
within which social identities are created. Another social factor that is key in the
development of bilingual communities is migration, as noted. This also has much
bearing on issues of identity formation. As the immigrants move between worlds,
whether to permanently remain in the new (host) country, or whether to stay only for a
particular work cycle, as migrant workers do, and then to return “home,” how are their
identities impacted? Do these people display patterns of cultural persistence, or of shift?

The close relationship between ethnicity and identity, aS mentioned, is well
established. Adding dimension, however, to this relationship is the role language plays in
the mix; it can be argued that one’s social identity is primarily established and maintained
through language. At this point, the questions become even more interesting, and
complicated: How is language related to any patterns of persistence or shift? Are there
aspects of language shifl that are related to identity (re)formation? It is this added
complexity that makes the relationship between language and identity a particularly rich
area of research. And a bilingual community, with well established migrant patterns,
becomes a prime location for investigating the relationship between proficiency in a
second language, and identity.

Watsonville, the community this study took place in, is agricultural, drawing
many migrant workers to the area from Mexico. Generations of migration have resulted
in a community that is largely Hispanic. This study draws its data from participants who

are children, and who are bilingual, with Spanish as their dominant language and English

as their second. The majority of the 41 participants in this research were classified as
migrants by the school district they were enrolled in—they were the children of migrant
workers. All of the children, with the exception of one, were born in Watsonville, and
the majority of their parents were born in México. Migrancy affects these participants’
social lives; as migrants, they make annual trips to Me’xico from Watsonville, during the
school year, and they participate in migrant academic support programs provided by their
school. Presumably, these children’s social identities are shaped by their movement
between places, across national borders. All of these factors combined drove the central
focus of this research, which was the relationship between these children’s proficiency in
English and their identities.

It is not clear to what extent we can generalize the questions or the results related
to language proficiency and identity. At some level, it is almost certain that there is some
generalization possible—we just don’t know where. It may be that at this point the only
meaningful questions that can be raised, and answers that can be obtained, are fairly
specific. If it turns out that this relationship is highly idiosyncratic, then it may be that
the only sort of question that can be asked is how some particular group of subjects’
identity is related to their second language proficiency, at some particular time and place.
But such extreme idiosyncracy is probably not the case. Perhaps at the least, we could
ease the restriction on ‘specific group of subjects’ and ‘specific time’ and safely ask the
question generalized to the point of considering similar groups in the same community,
but at different historical times, and so on. We can conceive of this as a sort of

continuum:

Table 1: Continuum of Generalizability

 

Worst Case Specific Generalized Further Best Case
Community Community Generalized
X X X X X
Everything Keeping Might be able to As we attempt to Can broadly
idiosyncratic demographic abstract features of generalize further, generalize to
content constant— the community, the suppose we find a everybody. An
diflerent subjects, population, and the community that is example of
but of the same sort, subjects: similar to (B) in such a
demographically; North American every respect generalization
different timeframe; agricultural except we now would be
same community community have difi'erent something like
(within reason: we Community with languages the following
don ’t expect things migrant farm involved—say the idea: Every
to be the same 100 workers from México immigrant normal person
years hence—as Spanish-dominant: language is raised in a
long as the First language for different. So the discourse
community is the parents; home obvious question community
same politically and language would be: Would will have at
geographically). Etc. the finding be the least one
same, that is., language.
would the
generalizations
hold?

 

 

 

 

The reader can imagine for herself the myriad ways that the conditions could be

different:

0 Different languages

0 Different communities

0 Different social classes

0 Different dominance relationships within the community

In such research, the only common factor would be that there are children under study,

and that second language acquisition is occurring.

Because such research entails multiple sets of social and linguistic factors,

definitive answers to these fantastically complex questions are not available. However,

we can hope to gain insight by careful study and analysis of specific Situations, combined

with theoretical frameworks that have been developed by other research. There is, after
all, a symbiotic relationship between these theoretical frameworks and each new
empirical study, such that the insights from each study provide ideas and insight for
theory.

I have approached my work from this position. The study that follows was
designed and implemented from this perspective. Generative linguistic theory and much
work in sociolinguistics gives us plenty of reason to believe that, indeed, there are many
aspects of linguistic behavior that are common across humans. But on the other hand, I
also recognize that the enormous complexity of social and linguistic factors involved in
the kind of research proposed here makes it necessary to be extremely cautious in
generalizations. On the one hand, there is the hOpe of being able to generalize from the
findings of a study; on the other, there is respect for the enormous complexity of the task.
My approach, therefore, was to do careful, focused study in one community, where I did
both empirical and qualitative research, in the hope of contributing to the solving of this
puzzle.

My interest in this community, in these people, is not just academic. Before
returning to graduate school, I was a bilingual teacher in this community. During that
time, I became ever increasingly interested in the patterns of acquisition of English that
my students—themselves very Similar to the participants 1 studied here—were displaying
in my classroom. In addition, I got to know their families, their parents, their
neighborhoods. I looked forward each February to the return of my migrant students, and
to the tales of Me'xico they would inevitably bring with them. These students, this

community, were actually one of the reasons I returned to school to study linguistics.

Clearly, there is a measure of personal curiosity at stake for me, in this study, as a result
of my own personal involvement with the community. This involvement benefited me in
many ways; through it, I gained access to the participants, the school site, and the school
records. But in addition, this involvement also serves to highlight the tension remarked
upon above—that between a stark scientific perspective, where all factors can be
abstracted and generalized, and the appearance on the ground, where everything is rich
and complex, and everything is idiosyncratic—where every second is like no other.

The general plan for this study was to operationalize grammatical competence in
English, and identity, and then to perform an empirical study to investigate the
correlation between these two variables. Theoretically, both of these concepts are very
broad—linguistic competence encompassing mastery of pragmatics, grammar, lexiS,
phonology, etc., and identity crossing the bounds of several fields, such as
sociolinguistics, psychology, sociology, and anthropology. Practically, however,
research is forced to operationalize such concepts by choosing a set of indicators. Of
course, there are limitations to this necessary reduction; those relevant to this research are
discussed later in this study.

From my understanding of identity theory, most of the literature supports the
claim that there is a strong positive correlation between identity and language. This also
seems commonsensical; language is emblematic of social factors. Importantly for this
research, many studies indicate that increased competence in a second language promotes
identification with that second language community. In Watsonville, which, as noted, is
predominantly Hispanic, several factors come together to encourage the acquisition of

English. While Spanish is a language that is spoken in many of the homes, and

throughout the community, English is the language of the official business. As the
language of the state, and of the country, English carries prestige. In addition, the
acquisition of English is the goal of the education system in the community. Given all of
these facts, it would not be unreasonable to expect that there would be a strong positive
correlation between the (second language) English proficiency and the identity of the
participants studied, such that the higher the English proficiency, the more Anglo the
identity. In a perfect theoretical world, that is, one in which there were no confounding
factors, this “expected” model would predict a perfect linear correlation between these

variables.

Research Questions:

At one level, I was curious about the relationship between language and identity,
but, as noted above, I am not sure that we can even ask such general questions. However,
at the very least, we can ask Specific questions, such as how these particular subjects’
identities were related to their second language acquisition, at this time and in this
situation. But then, the question remains: to what extent can these specific questions be
generalized? As a result of this, the following research questions emerged for me:

o What is the nature of the relationship between identity and English proficiency for
the participants in this study?
This entails a sub-question:

0 To what extent can such questions and answers be generalized?

Hypothesis

Given the theoretical frameworks and the research that informed me regarding the
strong correlation between language and identity, I began with a hypothesis that there
would be a strong correlation between identity and language proficiency, with my
particular subjects. Working from the “expected model,” then, I hypothesized that there
would be a strong correlation between Anglo identity and English proficiency, as
operationalized in the Methodology section of this study, for these particular subjects.

Indeed, I imagined that my study would contribute Specifics relating to particular
linguistic features and particular social factors, all within the expected model. I also
hoped that as a result of doing this particular empirical study, I would gain some insights
into the question about generalizability. However, as we will see, things proved to be not

so straight forward.

Organization of the Dissertation:

0 Chapter 1: Literature Review

0 Bilingualism

o Immigration/Migrancy

o The Community

0 Identity

0 Language Features Studied
- Chapter 2: Methodology

0 The Subjects

0 The School Site

0 Preliminary Preparation for the Study

0

Methods
. Elicitation Methods

I Scoring

Chapter 3: Results and Discussion

0

O

Frequency Data: Total Identity and Total English Proficiency
ANOVAs: Total Identity Score and Five Factors

ANOVAs: Total English Proficiency Score and Five Factors
Correlation: English Proficiency and School/Teacher Scores
Correlation: Total English Proficiency and Total Identity Scores
Summary Tables

Ethnic Identity: Can We Assess This?

Impressions: People, Food, and Vampires

Return to Main Street

Concluding Remarks on the Generalizability of the Findings

LITERATURE REVIEW

This research was hybrid in nature, in that its main focus brought together two
areas—second language English proficiency, and identity; the key concern in this work
was the relationship between English proficiency and identity. However, questions
pertaining to these issues are necessarily embedded within the larger context of
bilingualism, and so a discussion of bilingualism is also warranted. In addition, there are
two other aspects of the lives of the children who participated in this study that impact the
larger discussion. These are their status as migrants, and the community within which
they live. This Literature Review, therefore, has five sections: 1) Bilingualism, 2)
Identity, and 3) Immigration/Migrancy, 4) The Community, and 5) The Language

Features Studied.

Bilingualism

 

In the world today, it is inevitable that different languages will come into contact,
quite simply because there are many more languages than there are discrete discourse
communities. This is not to claim that languages outnumber discourse communities.

For one thing, it is extremely difficult to get an accurate count of discourse communities
because these are extremely fluid. Moreover, it is probably almost always the case that
individuals are members of numerous discourse communities, dropping out of some and
joining others, as their circumstances change. The issue then, is not numbers of discourse
communities versus numbers of languages, but rather, whether the discourse communities
are discrete, that is, linguistically insulated from all other discourse communities.

Certainly one can imagine times and places where such discourse communities could

10

have occurred. But, as the world grows smaller and transnational contact increases, such
discourse communities are probably rare to non-existent.

Among the linguistically naive, or worse, the linguistically oppressive, languages
are expected to be rigidly correlated with nationality. France, for example, has a long
history of attempts to eradicate unrecognized languages and dialects and to keep the
language “pure,” within that particular political boundary. Phillipson (1992) claims that
the defense of French can be traced back to 1883, with the establishment of the Alliance
F rancaise, and describes its work as having “intolerance of dialects and minority
languages within national borders, [and for promoting a] xenophobic national linguistic
purity” (p.106). In fact, as a result of many factors, including such oppression as the
establishment of and strict adherence to a national, “pure,” language, as well as
imperialism, many of the world’s languages are being lost; Skutnabb-Kangas argues that
today languages are “being killed and linguistic diversity is disappearing at a much faster
pace than ever before in human history” (1999, p. 188). Many linguists see this
phenomenon as an issue of human rights, and are working hard to understand it, and to
stanch its flow (Skutnabb-Kangas, 1999; Phillipson, 1992; Kontra, 1999a, 1999b;
Tollefson, 1991; Crystal, 1997). However, while this is a very urgent topic for
consideration, it is for another dissertation. What is important to the discussion here is
the fact that, linguistic imperialism and oppression aside, languages do not recognize
political boundaries. Rather, languages occur and vary according to the demand of the
speakers in the discourse communities in any particular region. The distinction between
discourse community and national affiliation becomes clear, when we compare estimates

for the number of languages in the world to the number of countries. Linguists (for

11

example, Romaine, 1995; Skutnabb-Kangas, 1999; Wei, 2000), estimate the number of
languages between 6,000 to 10,000, while the number of countries can be cited at fewer
than 200. Hamers and Blanc (2000) argue that the phenomenon of languages in contact
is actually an integral part of human behavior, citing such pressures for it as globalization
and increasing population movements because of immigration, as well as greater social
and geographical mobility.

One of the outcomes of language contact is bilingualism/multilingualism. While
there are no precise figures on the number and distribution of bilingual or multilingual
speakers in the world, in part the number estimated depends on one’s definition of
bilingualism. Grosjean (1982) estimates that approximately half of the world’s population
is bilingual; however, Wei (2000) claims that if the entire range of possible Situations in
which more than one language is used—for example, those in which two or more
languages are used regularly for work, family life, or leisure; those in which a
language(s) other than the L1 is occasionally used; those in which the foreign languages
learned in school are used for specific purposes, etc.—is to be considered to characterize
instances of bilingualism, then monolingual speakers actually make up a tiny minority in
the world today.

According to Mackey (2000), since the beginning of the 20th century the
definition of bilingualism has become continually broader. For a long time, bilingualism
was considered to entail equal mastery of two languages. As an example of a proponent
of this View, Mackey cites Bloomfield (1933, p. 56, cited on p. 26), who claimed that
bilingualism was “the native-like control of two languages.” Later, Haugen claimed that

bilingualism includes the ability to produce “complete meaningful utterances in the other

12

language” (1953: vol. 1, p. 7, cited in Mackey, 2000, p. 26). The current View of
bilingualism includes even passive knowledge of the written language, or any “contact
with possible models in a second language and the ability to use these in the environment
of the native language” (2000, p. 26, citing Diebold, 1961, p. 111). Mackey defines
bilingualism as the alternate use of two or more languages by the same individual, and
argues that this broadening of the concept is a result of the realization that the point
where a person becomes bilingual is either impossible to determine, or arbitrary.
Another factor that Mackey (2000) stresses is the interdisciplinary nature of the
study of bilingualism, claiming that it cannot be wholly described within the science of
linguistics. Romaine (1995), further notes that while there are related disciplines, such as
psychology, sociology, or education, which have been interested in aspects of
bilingualism—from its effects on cognitive processing, to bilingualism as an element of
culture conflict, or its relationship to intelligence, or to public policy—all too often
within these disciplines, following the dominant linguistic paradigm, bilingualism has
been seen as a deviation from the norm of monolingualism. For example, while many
cognitive psychology texts treat language extensively, bilingualism is rarely mentioned,
except in Special cases (1995). According to Romaine, taken separately, the various
disciplines interested in aspects of bilingualism can add little to our understanding of the
phenomenon; it is the interdisciplinary perspective advocated by Mackey—~in which the
various psychological, linguistic, and social interrelationships involved in bilingualism
complement one another—that is the most promising avenue of research. Romaine

(1995) argues that bilingualism cannot be understood except in relation to social context,

13

and therefore it falls within the field of sociolinguistics, a discipline concerned with how
language is used in society.

The relationship of bilingualism to the general field of linguistics is an interesting
one. Given even the more conservative estimate of how many bilingual speakers there
are in the world, noted above, it would not be unreasonable to claim that a comprehensive
understanding of the linguistic nature of human beings must entail a discussion of
bilingualism. In fact, J akobson claimed that “bilingualism is. . .the fundamental problem
of linguistics” (1953, cited in Romaine, 1995, p. 1). The idea, however, that bilingualism
may not in fact be a deviation from a norm of monolingualism, and may in fact be for
many (if not most) humans the normal linguistic state, departs radically from the
Chomskyan framework, which has informed much of modern linguistic theory. Within
this framework, the characterization of competence—or the knowledge of the rules of
grammar—is the central concern, and the starting point is the assumption of the
monolingual ideal speaker-listener. This research paradigm adamantly holds to the idea
that issues such as pragmatic competence, or the knowledge of the rules of social
interaction, for example, must be kept separate from the endeavor to characterize
linguistic competence, if progress in linguistic theory is to be made (e.g., Chomsky, 1986,
1995).

While this theoretical model has proven to be a very productive one—a lot has
been learned about core grammar—it does not, however, begin to address the issues
raised within the study of bilingualism. As noted, within the field, bilingualism is not
seen to be a simple result of the acquisition of parameters but instead, a very complex

phenomenon that is a result of the interaction of many factors—certainly including

14

competence in the Chomskyan sense, but also many other, social factors (Romaine, 1995;
Hamers and Blanc, 2000).

As would be expected, there are varying approaches to the study of bilingualism,
which only serves to underscore the complicated nature of the endeavor. For example,
Hakuta (1986) argues that the field Should not only study the individual, but the
circumstances which lead to the creation of bilingualism, its maintenance, and/or its
attrition. Hamers and Blane (2000) claim that when two or more languages are in
contact, of utmost importance is the relative functional use of these languages; in such
language contact situations, most often the different languages are used in different
domains and for different functions. Interestingly, Mackey (2000) argues that an
individual’s use of more than one language supposes the existence of two language
communities, but not the existence of a bilingual community. In this sense, bilingualism
is the property of the individual; “the bilingual community can only be regarded as a
dependent collection of individuals who have reasons for being bilingual” (2000, p. 26).

Mackey (2000) argues that the study of the phenomenon of bilingualism must
consider it as something entirely relative, and any description of bilingualism should be
done in terms of four criteria. The first, degree, refers to how well the individual knows
the two (or more) languages involved; the second, function, refers to what the languages
are used for, and in which domains; alternation, the third criterion, speaks to the issue of
code-switching between the languages; and the fourth, interference, focuses on the issue
of one language influencing the other. Mackey’s point is that bilingualism is a behavioral
pattern of linguistic practices which mutually modify each other, and which vary in terms

of the four characteristics mentioned above (2000).

15

In a general sense, this research falls within the domain of the study of
bilingualism. The participants were all bilingual, according to the current definition of
the term, as described above—the alternate use of two languages by the same individual:
These children were all Spanish-dominant, and while at varying points of proficiency,
were all able to use English as their second language. One of the strands of this study
centered on the participants’ acquisition of English as their second language. In assessing
this L2 acquisition, this study utilized two of the four criteria, mentioned above, that
Mackey (2000) suggests are important to any study concerned with bilingualism—degree
and interference; this will become clear in the Methods section of this dissertation. More
specifically, this research focused on the relationship of proficiency in a second
language, and identity. As a study involving the notion of identity, this work is
necessarily located within a framework of interacting social factors; the attempt to
characterize the relationship between the variable, identity—itself an assembly of social
facts—and targeted linguistic facts, brings this study into the domain of sociolinguistics.

One of the social factors that has much bearing on the lives, and the identities, of
the participants in this study is their status as migrants. We now turn to a discussion of

relevant issues surrounding immigration and migrancy.

Immigration/migrancy

There can be a tendency to think of national status dichotomously, that is, a
person is either a native of country X or an immigrant. However, a third very important
status is that of the migrant. In this case, we are talking about people who live part time
in their native country and part-time in the “host” country, usually because of seasonal

work in the host country. Politically, these people may be considered aliens, or

16

occasionally they may be nationalized in the host country. However, political
designation is only a small part of their social and ethnic status; arguably, their own sense
of identity and others’ perception of them is even more important, and this is particularly
true of children because children generally are little aware of bureaucratic designations,
though they are acutely sensitive to their local social milieu. The majority of the
participants in this study were officially classified as migrants by the school district they
were enrolled in, the Pajaro Valley Unified School District; it is therefore important to
understand a little bit about migrants in the United States and how they fit into the overall
patterns of immigration to this country.

The early part of the 20th century, right before World War I, marked a peak period
of immigration to the United States. Rurnbaut (1996) claims that the experiences of these
“white ethnies” (Italians, Poles, Greeks, Russian Jews) were shaped by class, not color,
and resulted in widespread social mobility and intermarriage. Citing Gans (1979); Alba
(1990); and Waters (1990), Rurnbaut claims that this has lead to these groups’ sense of
ethnicity as being symbolic, and exercised through Optional, familial, leisure-time forms
of activity. In contrast, the more recent and rapidly accelerating immigration to the
United States is unprecedented in its diversity of national origin, color, and class. The
1990 census reported an all-time high number of immigrants, 19.8 million, most of whom
reported themselves to be non-white (1996). The trend toward diversity is clear: “the
proportion of white immigrants declined from 88 percent of those arriving before 1960,
to 64 percent in the 19605, to 41 percent in the 19703, and to 38 percent in the 19805”
(1996, p. 121). By 1990, for the first time in US. history, Europeans were no longer the

largest immigrant population in the country, having been replaced by people from Latin

17

America and the Caribbean, with over half of these non-European immigrants arriving
during the 19808 alone (1996). Immigrants from one of these Latin American countries,
Me’xieo, have accounted for 22% of the total number of immigrants since 1970, and as a
result, make up the largest Hispanic immigrant group in the US. today (Rumbaut, 1996;
Durand & Massey, 1999).

Interestingly, Weber (2000), citing Hoerder (1994), points out that all immigrants
to the US. have historically tended to participate in “intense localism” (p. 930), which
has had the effect of these groups maintaining part of the “home-community social
structure and relations in migration” (p. 930). With the earlier immigrant groups, while
thousands of them maintained meaningful ties to their home communities——contributing
financially, participating in family decisions—many were cutoff because of distant and
slow communication (2000).

However, this does not describe the experience of the Mexican immigrant. For
generations, “Mexican immigrants have traveled northward on migratory arteries formed
long before either Mexico or the United States became nations” (Vasquez et al., 1994,
p.15), and have created communities that are some of the oldest and most widespread in
the history of the United States, with numbers that have steadily grown (1994). Because
these immigrants’ place of origin is more local than those of the immigrants from the
earlier wave (in fact, México is “right next door”), their patterns of integration into the
United States have been different, at least in part because many of them come to the US.
as migrant workers.

The majority of these immigrants come to the southwestern part of the United

States, from the states of Guanajuato, Jalisco, Michoacan, Nayarit, San Luis Potosi, and

18

Zacatecas, contiguous states in the central and western part of Mexico. Combined, these
states compose the most important migrant-sending region in México (Durand & Massey,
1999), with many of them making their way to California. Weber describes how
“Mexican communities in California of the 19205 were formed out of and based on social
ties which often had their beginnings in Mexico” (2000, p. 935). Mexicans migrated to
communities where they had family and fiiends, with newer immigrants first living with
family, and then moving close by, “lacing whole areas with families and social groups
from towns in Mexico” (2000, p. 93 5).

Santiago-Rivera & Santiago (1999) draw a distinction between place-to-place,
return, and circular migration:

Place-to-place migration occurs when an individual moves from

point A to point B. Return migration occurs when an individual

moves from point A to point B and then back to A. Circular

migration is illustrated by a move from point A to B to A and then

back to B (p. 229-10).

Circular migration, which reinforces the kind of community described above, is of
particular interest in this research. As noted, the majority of the participants in this study
were classified as migrants (at least 75.6%, and perhaps as much as 85.4% of them—the
school was in the process of determining whether 9.8% of them qualified); that is, they
were the children of migrant workers. In order to qualify for this status, and therefore
participate in the Migrant Education program, there are several criteria that must be met
by the families. The first pertains to the parents’ employment—they must work in some
officially recognized migrant occupation, for example, in the fields, the canneries, or the

nurseries in the area. Once qualified, the children are designated as migrants for two

years, after which they need to requalify, every year. To remain qualified, parents need

19

to demonstrate that they are still doing migrant work, and that they go to Mexico every
year, and remain there for at least 20 days (I. Vizcaino; personal communication,
December, 2000‘).

The Migrant Education program is federally funded, and is designed to help
students and their families overcome hardships they may encounter because of poverty or
disrupted schooling. This program has been fimded since 1966 through amendments to
Title I of the 1965 Elementary and Secondary Education Act. In California, the Migrant
Education program provides services to students in thirty-six districts in San Benito,
Santa Cruz (the county this study took place in), Santa Clara, San Mateo, San Francisco,
and Alameda counties. The services provided through this program are designed to

help migratory children overcome educational disruption, cultural

and language barriers, social isolation, various health-related

problems, and other factors that inhibit the ability of the children to

do well in school (California Department of Education [CDE],

2003, Migrant Education Program, para. 6)

The common pattern of migration for the migrant workers in Watsonville is
circular, whereby they spend most of the year in Watsonville, and then return to their
places of origin annually, generally in the middle of December when the crops have been
harvested. They then return again to Watsonville, at the end of January of each year, to
begin the cycle again (I.Vizcaino, personal communication, December, 2000). An

additional note of interest is that most (at least 75%) of the students at the school site

under study who were designated as migrant go and return from the same state in

 

' Isabel Vizealno was the school site’s Migrant teacher; I interviewed her about the programs that the
migrant students were involved in, as well about relevant aspects of their history, travel patterns,
requirements for participation in the migrant program, etc.

20

Mexico, Michoacan (I. Vizcaino, personal communication, December 2000), one of the
states in the most important migrant-sending regions, noted above.

Because of the generational layering of this circular migration, when these
immigrants initially travel to the host community, where their relatives and friends have
already established roots, they settle into a place that is rich with networks of social
relations that are similar in culture and language to the communities they left behind in
Mexico. These new communities become “ports of entry” (Vasquez et al., 1994, p.15)
for these immigrants, providing them “with opportunities for work and integration into
the mainstream of American culture” (Vasquez et al., 1994, p.15).

But what about the children of these immigrants? How are their lives impacted
by their parents’ lives being lived between two worlds? Almost all of the children in this
study were born in the United States, but travel regularly to Méxieo with their parents
(which is back to México for the adults, because almost all of them were born there),
spending several weeks at a time there. The home language, and therefore the native
language, for these children is Spanish, but all of them are bilingual, speaking English as
their second language (to varying degrees of proficiency). What is the relationship
between their proficiency in this second language and their identities? How does their
acquisition of English impact their identities—do they identify more with Anglo culture,
or with Mexican?

Rumbaut (1996) argues that we don’t know much about the children of the more
recent immigrants to the United States, claiming that research on their adaptation
processes is fragmentary. He further claims that

even less is known about the subjective aspects of the children’s
experience...including their modes of ethnic or national self-

21

identification, perceptions of discrimination, aspirations for their

adult futures, cultural preferences, forms of intergenerational

cohesion or conflict within their families, self-esteem and

psychological well-being, and how all these may be related to

more objective indices of their experience, such as their school and

work performance and language Shifts from the mother tongue to

English, in given social contexts (p. 122)

It is at this intersection that this research comes in. As noted, the primary focus
here is the interaction of (i.e., the nature of the relationship between) the participants’
identity and the acquisition of English as their second language. Before discussing the
two major areas of focus for this work—Identity and English Proficiency—I would like
to present a snapshot of the community, Watsonville, since that is the matrix within
which much of the action happens. I will return to a discussion of Watsonville, and how

the ineffable and rich social factors found within it interact with the language, the

bilingualism, of the community, in the Results and Discussion section of this dissertation.

The Community

Watsonville is in the Pajaro Valley, an agricultural region of central California,
approximately 90 miles south of San Francisco, inland from the Monterey Bay
approximately 5 miles. Over its history, this area has seen a range of settlement and
resettlement patterns. The first Europeans to explore this land were led by the Spaniard
Gaspar de Portola, in 1769. Included in his party were native Mexicans. The party’s
charge was to discover sites for the expansion of the Baja California chain of Missions
into the region. It is reported that when Portola and his expedition crossed into what is

now known as the Pajaro Valley, they saw a large straw-stuffed bird at the river crossing.

22

and named the river accordingly: Rio del dearo (River of the Bird) (Watsonville Public
Library, 2003, para. 2).

After the Mexican Independence from Spain, in 1821, former Mission-held lands
were given over to Mexican citizens, and the lands became ranchos. “Seven ranchos
were in the immediate vicinity of present Watsonville—San Andres, Los Corralitos,
Bolsa de Pajaro, Bolsa de San Cayetano, Salsipuedes, Laguna de Calabasas, and Vega del
Rio del Pajaro” (Watsonville Public Library, 2003, para. 2). These lands remained
Mexican until the U.S.-Mexican war, from 1846-1848, when they were lost to the US,
along with nearly half of the total Mexican territorial holdings—what is now the present
American southwest, from Texas to California. California became a state in 1850, as the
population surged because of the Gold Rush. Even after the Rush subsided, many
immigrants—Northem and Southern Europeans, Chinese, Japanese, Filipinos—continued
to move to the Pajaro Valley region because the rich and fertile soil provided many
opportunities for work in agriculture—growing and harvesting increasingly diverse crops
(Watsonville Public Library, 2003. Pajaro Valley Local History, para. 5).

Watsonville incorporated as a city in 1868. There are at least two stories about
how Watsonville came to be named. Some have it that it was named for all of the
Watsons living in the area at the time; others say that the name is derived from a Judge
who filed a famous claim against Sebastian Rodriguez, the owner of one of the large
ranchos in the region, Rancho Bolsa de Pajaro. Perhaps it was a bit of both. While Judge
John Watson lost the claim, ostensibly making his not the best name-choice, perhaps the
name stuck anyway, not only because of the notoriety of the claim, but because there

were so many Watsons in the area. The city continued to thrive as an agricultural area,

23

attracting more and more immigrants. A description of the city, just one decade later,
hints at some of the changes the city was going through:

Pajaro Street became known as Main Street as shops and

businesses moved in. By the mid-seventies, there were ten grocery

stores, six doctors, a dentist, two livery stables, three

harnessmakers, four blacksmiths, three wagonmakers, three meat

markets, seven hotels, eight saloons, two telegraph offices, one

bank, the Bank of Watsonville, one foundry, one newspaper and

one bakery in town. The young ladies’ group called the Butterfly

Club raised money for a fountain to be built in the middle of the

Plaza (Santa Cruz Library Weekly, 2003, para. 7).

By the late 1880’s, the Pajaro Valley region had grown to about 7,500 residents,
with Watsonville’s population at 2,800. One of the neighboring communities,
Freedom—which had until 1877 been called Whiskey Hill, because of its many
saloons—offered “the local cowboys, Mexican and Indian vaqueros from the neighboring
ranchos, and no doubt a patron or two from Watsonville, the chance to relax or kick up
their heels” (Santa Cruz Library Weekly, 2003). The activities that these vaqueros
engaged in included “Saturday night dances, fights, knifings, horse races along Corralitos
Creek” (Santa Cruz Library Weekly, 2003), but most interestingly, the largest festivities
were held on September 16, in celebration of Mexican Independence Day (Santa Cruz
Library Weekly, 2003).

In the final pages of this dissertation, we will revisit Watsonville’s Main Street
and Plaza, for an updated perspective on the place.

The five tables that follow are all drawn from data contained in various tables in
the US. Census Bureau’s 2000 census (US. Census Bureau, 2003. Fact Finder, 2000

Census). According to the 2000 census, by then, the total population of Watsonville had

grown to 44,265. Seventy five percent of this total, or 33,254, claimed to be Hispanic or

24

Latino, with most coming from México; the following table gives the breakdown,
including numbers for the state of California, and the United States as a whole, which
Show that Watsonville is a community that is populated by many more Hispanics/Latinos

than is typical.

Table 2: 2000 Census Data on Race/Ethnicity for Watsonville, CA, and US

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HISPANIC OR LATINO Number Percent
Total Population 44,265 100.0
Hispanic or Latino (of any race) 33,254 75.1
Mexican 29,953 67.7
Puerto Rican 39 0.1
Cuban 25 0.1
Other Hispanic or Latino 3,237 7.3
Not Hispanic or Latino 11,011 24.9
California: Total population 33,871,648 100.00
California: Hispanic or Latino 10,966,558 32.4
California: Mexican 8,455,926 25.0
US: Total population 281,421,906 100.0
US: Hispanic or Latino 35,305,818 12.5
US: Mexican 20,640,711 7.3

 

The table below, also compiled from the 2000 census, provides an overview of
several economic characteristics of this population. Of those over 16 years of age
(30,909), 63.9% are in the labor force, with a 7.9% rate of unemployment. When
focusing solely on females 16 years and over, we see that 56.1% of this population is in
the labor force, with the same rate of unemployment. The census further shows that in
61% of households, both parents in the family are in the work force. The unemployment
rate in Watsonville, as compared to that for California and the US. overall, is markedly

higher—more than twice the national rate:

25

Table 3: 2000 Census Data on Employment Status

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EMPLOYMENT STATUS Number Percent
Population 16 years and over 30,909 100.0
In labor force 19,739 63.9
Civilian labor force 19,739 63.9
Employed 17,285 55.9
Unemployed 2,454 7.9
Not in the labor force 11,170 36.1
Females 16 years and over 15,434 100.0
In labor force 8,658 56.1
Civilian labor force 8,658 56.1
Employed 7,432 48.2
Unemployed 2,454 7.9
Not in the labor force 6,776 43.9
Families, in which both parents are in
labor force 2,488 61.0
California rate of unemployment
(population 16 years or over) 1,110,274 4.3
US. rate of unemployment (population
16 years or over) 7,947,286 3.7

 

While not specifically indicating migrant status, occupational information derived
from the 2000 census shows that of the employed population 16 years and over, those
claiming the occupation category of “farming, fishing, and forestry” were 17.0% and
those claiming “agriculture, forestry, fishing and hunting, and mining” were 17.6%.
Although it is not possible to know how much overlap there is between these two
categories, they are indicative of how agricultural work fits into the overall occupational
picture for Watsonville. For a full breakdown of the other categories reported on this
census, please see Appendix A. Another source of information, the Pajaro Valley
Unified School District Data Source Book, 2000-2001, provides data that helps to flesh

out the category of “migrancy.” According to this source book, the total number of

26

students enrolled in the district was 18,445 (p.91). The total number of students in the
district participating in the migrant education program was 9,184 (p. 93), which is
approximately 50% of the total number enrolled. Again, while this information is based
on a targeted slice of the population of Watsonville, it is nonetheless indicative of the
importance of agriculture to this area; as already noted, in order to qualify for the migrant
program within the district, one of the criteria that parents need to meet is employment in
Officially recognized migrant occupations, such as the agricultural fields, the canneries, or
the nurseries. In fact, according to the City of Watsonville, in Santa Cruz county (again,
the county that the Pajaro Valley is located in), there are 815 local farms, each averaging
68 acres, and “agriculture and food processing remain the mainstay of the Paj aro Valley
economic structure” (City of Watsonville, 2003. para. 6)

Those in Watsonville claiming to be private wage and salary workers were 83.4%
of the working population, with 10.9% working for the government, and 5.4% self-

employed (table below compiled from 2000 census):

Table 4: 2000 Census Data on Class of Worker

 

 

 

 

 

 

 

 

 

CLASS OF WORKER Number Percent
Private wage and salary workers 14,423 83.4
Government workers 1,881 10.9
Self-employed workers in own not 928 5.4
incorporated business
Unpaid family workers 53 0.3

 

The household income for Watsonville, compiled from the 2000 census, is listed

on the following table, including information for California and the US. When

27

comparing median family income and per capita income, we see that Watsonville is well

below the state and national averages:

Table 5: 2000 Census Data on Income for Watsonville, CA, US

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INCOME IN 1999 Number Percent

Households 1 1,478 100.0
Less than $10,000 1,051 9.2
$10,000 to $14,999 772 6.7
$15,000 to $24,999 1,636 14.3
$25,000 to $34,999 1,816 15.8
$35,000 to $49,999 1,934 16.8
$50,000 to $74,999 2,245 19.6
$75,000 to $99,999 1,112 9.7
$100,000 to $149,999 586 5.1
$150,000 to $199,999 129 1.1
$200,000 or more 197 1.7

Median household income (dollars) 37,617

Median family income (dollars) 40,293

Per capita income (dollars) 13,205

California median family income (dollars) 53,025

California per capita income (dollars) 22,71 1

US. median family income (dollars) 50,046

US. per capita income (dollars) 21,587

 

 

 

 

 

The number of families in Watsonville below poverty level was 1,381, or 15.4%
(compared to: California, 10,6%; US, 9.2%); those families with children under 18 years
of age that were below the poverty level were numbered at 1,233, or 19.7% (compared to:
California, 15.3%; US. 13.6%); the number of families below poverty level with children
under 5 years of age was 693, or 23.5% (compared to: California, 19%, US, 17%). In
all three categories, Watsonville had higher rates of poverty than were found in either

California or the United States as a whole.

28

The educational level of this community is shown in the table below, with
comparative figures for California and the US, again drawn from the 2000 census. The
numbers Show that the educational attainment for the population in Watsonville is far

below the state and national averages:

Table 6: 2000 Census Data on Educational Attainment for Watsonville, CA, US

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EDUCATIONAL ATTAINMENT Number Percent
Population 25 years or over 24,045 100.0
Less than 9th grade 8,746 36.4
9tfi to 12th grade, no diploma 3,500 14.6
High school graduate (includes equivalency) 4,819 20.0
Some college, no degree 3,951 16.4
Associate degree 939 3.9
Bachelor’s degree 1,429 5.9
Graduate or professional degree 661 2.7
Percent high school graduate or higher 49.1
Percent bachelor’s degree or higher 8.7
California population 25 years or over 21,298,900 100.0
Less than 9‘“ grade (California) 2,446,324 11.5
Percent high school graduate or higher (Calif) 76.8
Percent bachelor’s degree or higher (Calif) 26.6
US. pgoulation 25 years or over 182,211,639 100.0
Less than 9Wgrade Q5.) 13,755,477 7.5
Percent high school graduate or higher (U .S.) 80.4
Percent bachelor’s degree or higher (US) 24.4

 

The language of the “official” business of Watsonville (government, general
business, infrastructure) is English—as it is in the wider context of California. However,
within the local Pajaro Valley Unified School District (PVUSD), there has been a
tradition of institutionalized support for the non-native English Speaking school children

in this community, including bilingual education programs, and migrant services. The

29

bilingual education programs have been for the most part transitional programs, which
have the goal of transitioning the students into English instruction by third, fourth, and
fifth grades—approximately to a 50% level by third, a 75% level by fourth, and fully
transitioned by fifih. While the emphasis is on English acquisition, in especially the
earlier grades, such a program provides native-language support for the acquisition of
literacy, and other academic skills, concurrent with English language development.
However, in 1998, California voters passed a ballot initiative, Proposition 227,
which set in motion the dismantling of the bilingual education programs, statewide. The
goal of this initiative has been to track the state’s language minority students into English
immersion programs within one year of their beginning school, regardless of their level
of proficiency in English (with limited structured support in native language, depending
on proficiency in English). The putative goal of this initiative was to improve the
acquisition of English by the state’s non-native population. Across the state, the reaction
to the requirements of Proposition 227 has been varied; while there are districts that have
moved to be in complete compliance with the proposition, others have worked to
maintain what they deem to be successful bilingual programs (Krashen, 2001). Some
schools in the Pajaro Valley Unified School District have had a strong parent response
against Proposition 227; these parents have worked with the schools their children attend
to initiate waivers to the requirements of the Proposition, stating that they desire that their
children remain in bilingual programs. According to the district’s Bilingual Coordinator,

12 of the 16 elementary schools in the district have some level of the waiver process in

30

effect (L. Verde-Rivas, personal communication, December, 20002), including the school

site under study.

Identity

Already discussed has been the inevitability of language contact in the world, and
the resultant bi- and multilingualism—given such pressures as globalization and
immigration. As immigrants move from one geographical location to another, many
complex processes are involved, at many levels, including psychological, cultural, and
social. Immigrants can “experience displacement, relocation, and redefinition of self and
community as they confront the necessity of classifying the world around them (Goldin
1999, p. 1-2). In these “transnational” contexts, identities are formed and re-formed.

While discrete or distinct ethnic or racial groups are as elusive as discrete
discourse communities (noted above), nevertheless, people believe that they live in a
world made up of such groups, and ground this belief in the terms of identity; Le Page &
Tabouret—Keller argue that these “terms function as symbols ready at hand for identities
to hang on, providing the links between individuals and groups, the instruments therefore
of identification” (1985, p. 208). Ethnicity, much like gender and class, has long been,
and is commonly taken to be, a construct within which social identities are created.
Tajfel’s (1982) work in social identity theory has informed many studies. According to
this theory, people tend to categorize their social world into groups, which consist of two
kinds—those that are made up of others who are similar to oneself, and those that are
made up of others who are different. Ethnic identity is a combination of this knowledge

of membership of social groups, and the value and emotional significance attached to the

 

2 I interviewed Dr. Lupe Verde-Rivas regarding the district’s bilingual programs, and the effects of
Proposition 227 in the district.

31

that membership (Tajfel & Turner, 1979). Phinney (1990) claims that perhaps the best
way to view ethnic identity is as a subjective feeling—one of belonging to a certain
ethnic group. Fishman (1997) defines ethnicity as macro-group “belongingness” Q).
329). Fishman warns, however, that as a term, “ethnicity” should not be taken to mean
“culture.” Ethnicity is narrower than culture; there are many aspects of culture, for
example, artifacts such as books, that are not identificational—that is, they are not
indicative of ethnicity. Fishman also claims that ethnicity is more perspectival than
culture, in that it is “fundamentally subjective, variable and very possibly non-
consensual” (1997, p. 329). Clearly, ethnicity and identity are tightly bound.

Yet, in attempting to understand identity, it is not sufficient to focus solely on
ethnicity, despite the close relationship between these constructs. It is through language.
for the most part, that one’s social identity, and ethnicity, are established and maintained
(Gumperz & Cook-Gumperz, 1982; Tabouret-Keller, 1997); further, because people can
be members of more than one group simultaneously, language becomes a means for
social comparison. It is through language that one or another of the various group
memberships a person belongs to becomes salient (Giles & Johnson, 1987). AS one of
the most noticeable characteristics of ethnic groups, spoken language is perhaps the
preeminent point of reference for understanding issues of identity and how they affect
and are affected by social processes. According to Dorian,

although many behaviors can mark identity, language is the only

one that actually carries extensive cultural content. The distinctive

sounds uttered in speaking a particular language encode meaning,

and the link between ethnic group and ethnic language becomes
much more important at this level (1999, p. 31).

32

Moreover, the link between language and identity is experienced as vital, and as
primal; in fact, “almost every aspect of language is replete with ethnic significance and
associations” (Fishman, 1997, p. 340). Inclusion in, or exclusion from, particular social
groups is regularly done on the basis of linguistic features, even a single phonemic
feature (e.g., the well known third person singular verb form without overt —s
morphology, in AAVE). According to Tabouret-Keller,

language features are the link which binds individuals and social

identities together. Language offers both the means of creating this

link and that of expressing it. Such features imply the whole range

of language use, from phonetic features to lexical units, syntactic

structures, and personal names” (1997, p. 317).

However, just as ethnicity is variable, so is the link between language and
identity. It is constructed by circumstances—contextual, historical, and social. Because
of this, many linguists interested in sociolinguistic variation have investigated this link.
ofien referred to as ethnolinguistic identity. An outcome of such research has been the
confirmation that etlmolinguistic identity is not unidimensional, and that it “is not
invariant but changes from one occasion to another” (Fishman, 1999). According to
Tabouret-Keller (1997), these changes occur through the process of identification,
whereby a person takes on, either unconsciously or consciously, aspects of another
person’s or another group’s behavior. Tabouret-Keller claims that identification is one of
the reasons for the close link between language and identity. Language in use provides
the largest set of easily adoptable features for this identification process, making it a
prime target for variation, and for studying.

Le Page & Tabouret-Keller (1985) is an important work that focuses on language

and identity, and which draws on earlier work done in Accommodation Theory (Giles,

33

1979). According to them, their work is similar to Giles’ work, in that it is also
concerned with issues of ethnicity and people’s perceptions of one another, but it differs
in important ways. Accommodation Theory centers on interactive behavioral events, and
how these reflect the way people accommodate linguistically to one another, or more
specifically, their perceptions of one another. In contrast, Le Page & Tabouret-Keller’s
work focuses on how people ascribe linguistic attributes to their perceptions of other
people, and groups (1985). Pushing further, they assume that linguistic items are not just
attributes of communities or groups, but that “they are themselves the means by which
individuals both identify themselves and identify with others” (p. 5), and that linguistic
behavior is “a series of acts of identity in which people reveal both their personal identity
and their search for social roles” (p5). In these acts, identities are born and disintegrate,
through processes Le Page & Tabouret-Keller characterize as focusing and diffusion.
Focused groups are those comprising closely interactive, and tightly-knit communities,
within which greater regularity in speech is manifested; with diffuse groups, it is the
opposite—more loosely-knit communities, and more linguistic variation. Le Page &
Tabouret-Keller borrow the concept of networks from Milroy’s (1980) work, claiming
that there is a high correlation between the degree of focusing around a particular set of
linguistic features or norms, and dense and multiplex social networks. Interestingly,
however, when Le Page & Tabouret-Keller did the research for their study, they were not
aware of Milroy’s work with networks, and admit that had they been familiar with it, they
“might well have used it [themselves]” (p. 116).

According to Le Page & Tabouret-Keller, “focusing” can be promoted by several

factors; the first is daily and close interaction in the community; the second is a sense of

34

common cause—brought on by, for example, an external threat or danger; the third is a
powerful model of some sort, such as religious teachings or symbols, a leader, or a
prestige group; and finally is the education system of the community (1985).

These researchers argue that our ability to get into “focus” with the people or the
groups we want to identify with is constrained, and that these “constraints upon our acts
of identity” (p. 182) can be categorized in four ways; we can get into focus with those we

want to identify with “to the extent that

1)

a) we can identify the groups

b) we have both adequate access to the groups and ability to analyse their behavioural
patterns

c) the motivation to join the groups is sufficiently powerful, and is either reinforced or
reversed by feedback from the groups

(1) we have the ability to modify our behaviour” (p. 182).

Of these categories, Le Page & Tabouret-Keller found positive and negative
motivation to identify with groups to be the most predictive of linguistic behavior, which
is not surprising, considering the work of linguists such as Labov, Trudgill, and Milroy,
who show similar results regarding language variation and group solidarity and/or the
desire to be distinguished from others. Le Page & Tabouret-Keller further argue that
multilingual communities provide rich, and obvious environments, in which such
motivation becomes even more apparent; in these environments, the linguistic rule-
system (or aspects of it) of the group that is perceived to be socially desirable is adopted
(1985). Tabouret-Keller (1997) further claims that environments in which there is
language contact provide good cases for studying language and “identity fusion or

disjunction” (p. 323).

35

Noels et a1. (1996) is another study that is relevant to this research. These
researchers claim that increased competence in a second language promotes identification
with that second language community. They further argue that “for minority group
members, the development of second language skills and of a second ethnic identity may
undermine the original ethnic identity” (1996, p. 248). This notion is obviously key to
the current study, which at the outset, predicted that as the participants’ English
proficiency increased, so would their Anglo identity. Further supporting these
expectations is a general frame of language use in Watsonville. While the community is
predominantly Hispanic, English is nonetheless a language of prestige—it is the language
of “official” business, shored up certainly by the state-wide adoption of Proposition 227;
and while there is a tradition of institutional support for bilingual education programs in
the community, which support native-language instruction in the earlier grades, the goal
of such programs is clear—the acquisition of English.

Nonetheless, acculturation does not necessarily follow one straight path. Indeed,
there is a range of possible patterns for this process. Noels et a1. (1996) cite Berry (1980.
1990) as providing an explanatory model of possible types of acculturative processes
involved with re-identification. These include:

2)
a) separation (i.e., relatively exclusive involvement in the native ethnic community)

b) assimilation (i.e., relatively exclusive involvement in the target ethnic community)

c) deculturation or marginalization (i.e., rejection of both cultures as ethnic reference
groum)

(1) integration (i.e., high levels of implication in the native ethnic group as well as high
levels of involvement in the other ethnic community).

36

According to Noels et al., Berry and his colleagues, in their research on attitudes
toward the process of acculturation, have generally found that individuals prefer to
experience integration [as opposed to separation, assimilation, or deculturation] (citing
Berry, Kim, Power, Young & Bujaki,1989).

Le Page & Tabouret-Keller (1985) expand on their notion of a person’s ability to
modify their behavior (one of the constraints on a person’s ability to get into “focus”
with, or identify with, others, noted in 2.d. above), claiming that as second language
speakers receive new linguistic data, it is perceived in terms of models that they have
already constructed. They assert that in multilingual communities “the individual creates
for himself the patterns of his linguistic behaviour so as to resemble those of the group or
groups with which from time to time he wishes to be identified, or so as to be unlike
those from whom he wishes to be distinguished” (p. 181). Importantly, for this study, Le
Page & Tabouret-Keller claim that for children, these processes of accommodation, of
creating linguistic patterns resembling the groups they identify with, are far less difficult
than they are for adults. This is not to imply that joining a group is not complicated.
Tabouret-Keller (1997) argues that this is actually itself a very complex process that
involves much of a person’s history—subjective and intimate—a5 well as multiple social
factors, such as status in society. Accordingly,

identity is rather a network of identities, reflecting the many

commitments, allegiances, loyalties, passions, and hatreds

everyone tries to handle in ever-varying compromise strategies.

These imply language use to mark group affiliation, to reveal

permitted or forbidden boundaries, to exclude or include, etc.

(1997, p. 321). (italics mine)

Given the protean nature of identity, captured across the discussion above, any

attempt to formally study the processes of identification and acculturation is bound to fall

37

short in numerous ways. Because a lot of the processes concerned are internal and
hidden, it is practically a certainty that much will be missed. However, as stated at the
outset of this project, perhaps an accumulation of well-crafied studies which attempt to
isolate and operationalize the variables involved can bring us closer to a true picture of
the enormously complex process of identity formation.

Identity Operationalized

Identity, as operationalized in this study, is a bundling of social factors—a
constellation of symbolic information, and preferences. Again, it is frankly admitted in
this study that any such classification is a violation of sorts, because it will inevitably
leave out many important nuanced factors that are operating and interacting in highly
complex ways. Nonetheless, because studies involving the social aspects of being human
are important, many researchers have grappled with this problem, resulting in fairly well-
established techniques to discover and classify identity.

Researchers who have studied ethnic identification in children often use a
technique that requires the children to identify their preferences between dolls or pictures
that represent members of their own ethnic group, and those representing other groups
(Hamers & Blanc, 2000, citing Genessee, Tucker & Lambert, 1978 as an example). This
study borrows from this idea, but extends it to include concepts beyond people, as
indicators of group membership. That is, in questioning the participants in this study
about their preferences, I concentrated on aspects of cultural practices, instead of on
particular members of the respective groups, focusing on the following 5 general

categories:

38

Favorite food(s)

Favorite music

Sports

Holiday foods eaten by the family

Preference for place—Mexico or Watsonville

This research investigated the nature of the relationship between two variables,
identity and the acquisition of a second language; I now turn to a discussion of the

language features that were studied.

Language Features Studied3

The range of grammatical features that were examined in this study were chosen
because they reflect deep structural differences between Spanish and English, and
therefore would provide a good snapshot of the developing bilingual grammars of the
participants involved; as such, these features cover syntactic configuration (N oun-noun
compounds), morphological attachment (tense doubling), as well as several different
functions of a core grammatical element (articles). Moreover, because these are “core”
features, arguably, they can be used as a reasonable measure of proficiency.

Certainly a concern when investigating the grammars of children is whether or not
developmental factors are interacting with the data. A seminal child acquisition study
(Chomsky, C., 1969), which focused on children between the ages of five and ten,
questioned the common assumption that by age five children have mastered native
language syntax; her study claims that in fact, at age five, syntax is not fully formed, and
that between the ages of five and ten syntactic development continues. Chomsky’s study

centered on the acquisition of structures which were termed linguistically complex, that

 

3 All Spanish used in this research was reviewed by two native Speakers of Spanish, one (a linguist) from
Spain, and the other from Mexico.

39

is, ones in which the grammatical relations holding among words are not necessarily
reflected in surface structure, of the type (1969, p. 26, 33):

3) John is easy to see

4) Bozo tells Donald to hop up and down

5) Donald promises Bozo to hop up and down

Using dolls in her experiment, for items such as in 1) above, Chomsky told the children
to make the dolls “easy to see,” and for items such as 2) and 3), told them to make them
“do it.” She found, with the first type of sentence, that almost all of the five year olds
answered incorrectly, and all the nine year olds answered correctly, while the six, seven
and eight year olds showed mixed results. With the tell/promise sentences, only half of
the children knew the test structures.

Another important child language acquisition study (Goodluck, 1991),
acknowledges that Chomsky’s notion that syntax is late in developing became popular,
due to her influential study. However, Goodluck claims that since its publication, much
evidence for very early syntactically based grammars in children has been compiled
(citing, for example, Gordon, 1985; (Gleitman, 1989). Goodluck argues that Chomsky’s
‘syntax is late’ hypothesis really can’t be maintained in any strong form because all but
one of Chomsky’s tests involved lexically restricted phenomena—such as rules “for
interpreting the missing (PRO) subject of a complement clause (‘promise vs. ‘tell’) and
the distribution of missing object constructions (‘easy’/’eager’)” (1991, p. 106).

Goodluck concludes that children from as early an age as two to three years have
syntactic systems that are tuned to the syntax of the language around them, and which

reflect the core underlying principles which govern adult grammars; she further adds that

40

by age three or four, children have “acquired the basics of complex syntax (phrase
structure configurations appropriate to their language) and interpret sentences using
structurally-based principles” (1991, p. 106). Goodluck concedes that it may well be that
a child’s grammatical system is not exactly the same as an adult’s until a few years into
their schooling, but claims that the notion that a child’s syntactic system is
“fundamentally impoverished or non-existent for any extended period” is untenable,
given the evidence (1991, p. 107).

Given the research discussed above—which indicates that at an early age children
have robust core syntactic systems, with some variation for lexical restriction—combined
with the fact that the youngest group of children I tested had a median age of 7, with the
youngest at age 6.1, I did not expect to encounter developmental issues.

Before describing the grammatical features studied in this research, I would like
to briefly mention the elicitation methods used to collect data on these features. For N/N
Compounding and Article usage, two methods were used: 1) Repetition, and 2) Picture
Task, with the goal of eliciting spontaneous, unmonitored speech containing the targeted
linguistic structures. Wolfram & Fasold (1974) and Lust, et a1. (1996) argue that with
children, imitation is not merely a passive copy, but in fact reflects cognitive competence;
they claim that the oral production elicited through repetition tasks appears to quite
accurately represent children’s grammatical competence. Radford supports this idea,
claiming that there is a conventional assumption in the field “that children can
consistently imitate only those items whose morphosyntax they have mastered” (1990, p.
86). The other technique, the use of pictures in elicitation, is one of the methods most

commonly used to assess children’s linguistic capabilities (McDaniel et al., 1996). The

41

strategies I used to collect the Tense Doubling data were a bit more complicated; these
are discussed in more detail in the Methodology section of this research.
Noun/Noun Compom

Noun/Noun (N/N) compounding is a very productive process in English; in fact,
according to Ryder, this type of compounding “has been on the increase at least since
1750” (1994, p. 4, citing Leonard, 1984). In addition, and important to this study, Ryder
claims that N/N compounding in English is also one of the earliest derivational processes
used by children (1994, p. 4, citing Derwing, 1979). Supporting this, Clark (1995) states
that children as young as age two form novel nouns, the earliest typically being
compounds formed from two nouns, as in crow-bird and taxi-car (1995, p. 401 ). She
cites additional N/N compounds produced by two and three year olds, for example, car-
smoke and fire-dog (1995, p. 405-6).

In contrast, in Spanish, N/N compounding is not a productive process. Related to
this difference between Spanish and English is the fact that in English there are two ways
of expressing possession, the Saxon genitive (‘s), and the Norman genitive (also called
periphrastic genitive), which uses the NP of NP structure (Carreiras & Clifton, 1993);
Spanish has only the Norman genitive. For example, in Spanish, the English N/N
compound, dog house is realized with an NP of NP structure: casa del perro (house of the
dog)

With N/N compounding in English, it is easy to assume that a term, such as the
orange ’s juice, using the Saxon genitive, could become the orange juice, through a
simple process of phonological reduction. There are English N/N combinations in which

the first noun ends in the genitive —s that are puzzling because their status as either

42

inflected genitive constructions or compounds is ambiguous (Warren, 1978). Examples
of these follow, taken from Warren (1978):
6) driver’s seat
7) pastor’s cap

In her study, Warren’s focus is on the semantic class of such structures and she
states that whatever the function of the -s morphology is in these elements, they belong
to either the semantic class of Purpose (6. above) or Possession (7. above). While a
detailed analysis of N/N compounding in English is well beyond the scope of this paper,
it is worth briefly noting that we can also find current examples of type 7) above, in
which the -s suffix has been reduced, as in the following:
8) bachelor pad

While not completely relevant to this discussion on N/N compounds, it is
interesting to note that Spanish does have a very productive word formation process that
also results in compounds that are always nominal, but these are composed of verbal

elements and their complements (Lloyd, 1968). A couple of examples follow, taken from

Lloyd (1968):
Spanish English
9) picaflor hummingbird
prick/puncture flower
10) correfaldas woman chaser
run skirts

Through the process of lexicalization, N/N compounded elements make their way
into Spanish; the difference between these structures and the ones found in English,

though, is the direction of the head to the modifying element. In the English N/N

43

compounds, the modifying element precedes the head; in Spanish, it is just the opposite,

as can be seen in the following:

English Spanish
11) swordfish pez espada
fish sword
12) woodpecker pajaro carpintero

bird carpenter

13) wallpaper papel tapiz
paper tapestry

This difference in head direction reflects deep syntactic configuration, and was
therefore an ideal area of focus for this study. I expected that the students who were less
proficient in English—those whose grammars would be more Spanish-like—would show
more evidence of the structure of head+modifying element in their production of English
N/N compounds (i.e., they will be more likely to say bed flower for flower bed) than
those more proficient in English.

Article;

While the distribution of articles in Spanish and English is similar in certain
respects, there are points of difference that make article usage another good area to
investigate in this study. In general form, both Spanish and English have two sets of these
determiners——definite and indefinite articles, with generally similar distributions: the
definite article refers to someone or something already contextualized (by previous
mention; by being unique to the culture; etc.), the indefinite article refers to an item or
person not contextualized. Beyond this, however, there are various ways in which article

usage in Spanish and English differs.

44

The analysis that follows, and the majority of the data used in the examples on
article usage, are drawn from a volume in a series of studies undertaken in the early
19605 by the Center for Applied Linguistics (CAL), which focused a careful contrastive
structural analysis between English and the five foreign languages most commonly taught
in the United States at the time: French, German, Italian, Russian, and Spanish
(Stockwell et al., 1965).

One of the ways that article usage in Spanish and English differs is with regard to
several aspects of quantification. First of all, there can be discrepancies in class
assignment—mass or count—for semantically equivalent lexical items. What in English

is a mass noun, can be a count noun in Spanish.

 

 

English Spanish
Mass noun Mass man w/couLuzble element Plural count noun Singlular count noun

l4) thunder a clap of thunder ' truenos un trueno
thunders a thunder

15) advice a piece of advice consejos un consejo
advices an advice

16) furniture a piece of furniture muebles un mueble
furnitures a furniture

Interestingly, English accommodates speakers’ need to refer to particular
instances of these nouns by resorting to periphrastic constructions (a piece of... ). The fact
that this regularly occurs shows two things: 1) English deeply resists conversion of mass
nouns to count, and 2) speakers do not need to refer to particular instances of these nouns,
but will respect their grammatical properties when they do so. Again, these facts suggest

deep structural properties in English that prevent conversion of mass to count nouns.

45

In contrast, converting mass nouns to count constructions is quite productive in

 

Spanish:
English mass Spanish mass or count nouns
17) two bars of soap dos pastillas de jabon dos jabones
two bars of soap two soaps
18) two pieces of wood dos trozos de madera dos maderas
two pieces of wood two woods
19) two bits of information dos puntos de informacién dos informaciones

two bits of information two informations
A particular ramification of this difference pertains to mass nouns that can be
equated in both languages; Spanish, unlike English, requires the definite article with the

following mass nouns:

 

 

English mass noun §p_anish mass noun Spanish count noun
20) (the) applause e1 aplauso dos aplausos
the applause two applauses
21) (the) laughter la risa dos risas
the laughter two laughters
22) (the) support e1 apoyo dos apoyos
the support two supports

Another area within article usage that Shows variation between Spanish and
English is that of specificity—regarding definiteness and indefiniteness. According to
the Contrastive Structure Series, mentioned above (Stockwell et al., 1965), this aspect of
article usage is what causes learners of English and Spanish the most difficulty; as shown

below, this phenomenon ranges over a) nouns showing mere identification:

English Spani h
23) He is a doctor Es médico *Es un médico
He is doctor (a)

46

24) He is a friend of mine Es amigo mio *Es un amigo mio
He is fiiend mine

25) She is a teacher Es maestra *Es una maestra
She is teacher

(Note: The Spanish sentences in the third column are starred as sentences with predicates that
indicate identification; if these same sentences were meant to indicate individualization, or
emphasis, the article would be used: Es un médico=He is a doctor; Es un amigo mio=He is my
fiiend. Also, in situations of additional modification (Es un buen médico/He is a good doctor), or
in figurative speech (Es un lobo/He is a wolf), an article would be used [1965]).

and b) nouns referring to all of something:

English S anish
26) Man is mortal fl hombre es mortal
The man is mortal
27) Children are innocent _L_(§ nifios son inocentes

The children are innocent

28) Idon’t like rice No me gusta e1 arroz
No(t) me pleases the rice

Another fact related to the issue of specificity is that in Spanish, when personal
titles are used, the definite article is obligatory; this use of a double determiner—which

contrasts sharply with English—makes this an important feature of article usage to

examine:

m §E§Lni_h
29) Mr. Brown is here El sefior Brown esta aqui
30) Professor Phillips said so Q profesora Phillips 10 dijo

The last aspect of variation of article usage between Spanish and English focused
on in this study has to do with inaliable possession. In Spanish, when possession is
obvious, as with body parts especially, the definite article is used instead of a possessive
pronoun, which is the form used in English; number 33) below indicates that this

tendency to mark obvious possession with a definite article is not limited to body parts:

47

English Spani h

31) Your feet are very dirty Tienes los pies muy sucios
You have the feet very dirty

32) My whole body aches me duele todo el cuerpo
Me hurts all the body

33) Take your shirt off Quitate la camisa
You take of the shirt

While there are environments in Spanish in which this restriction is loosened, for
example with some verbs that don’t require clitic pronouns, it is nonetheless quite robust
(A. Martinez, personal communication, September, 2000’):

34) *Me duele mi mano
Me hurts my hand

35) Me duele la mano
Me hurts the hand

36) Levanto la mano
He raised the hand

37) LevantO su mano
He raised his hand

In closing this section, I would like to briefly return to the question, raised above,
of potential developmental issues involved in the acquisition of the grammatical elements
under study. A theoretical perspective that has bearing on the acquisition of articles is
one held by Radford (1990, 1995). Radford argues that in the earliest stages of child
language acquisition, the structures children acquire are lexical-thematic, and lack the
functional/nonthematic I, C, and D constituents. He further asserts that from around two
- years of age, these functional category systems are acquired, and claims that they are

implemented concurrently. While not all linguists working on these questions agree with

 

" Asuncién Martinez, a linguist and a native Spanish Speaker, was a former committee member for this
dissertation. l conferred with her on all of the Spanish grammatical elements studied in this research.

48

this latter claim (e.g. Clahsen, & Penke; Roeper; Verrips & Weissenbom, as cited in
Meisel, 1992), Radford (1995) does provide convincing evidence for his position.

Radford acknowledges that while these functional category systems appear to be
acquired from around two years of age, children this age have not mastered all of the
morphosyntactic complexities of these systems, and seem to go through a transitional
stage towards mastery. The important point here, for purposes of this study, is that
articles, which are part of the Determiner (D) System, a functional category, seem to be
acquired early. Radford cites data (1990, p. 284-288), drawn from 13 participants
ranging in age from 26 to 27 months, all of whom display quite productive use of articles.
Supporting this is data contained in another study concerned with maturational issues in
the acquisition of syntax (Borer & Wexler, 1987, p. 147). While these data are drawn
from a section of Borer & Wexler’s study focusing on the acquisition of verbal passives,
they nonetheless provide more evidence for early acquisition of articles. These come
from subjects aged 24-48 months:
38) the tree’s smashed
39) the window’s breaked again
Even at this very early age, these children are using articles, which is interesting because,
arguably, articles are semantically empty, which shows that these children have grammar
at work at this age, and not just semantics.

Again, my expectation with the article data was that my subjects whose
grammars were more Spanish-like would evidence more Spanish-like structures in their

article use.

49

Tense Doubling

The final aspect of grammatical difference between English and Spanish
investigated here I generally refer to as tense doubling, although these data include other
grammatical features as well. In Spanish Wh-questions, the main verbs are
morphologically marked for number and tense, while in English the auxiliary do bears the

tense and the main verb is in its bare form:

English Spani h
40) What did Joe buy Qué compré J osé

What bought José

41) Where did Maria go AdOnde fue Maria

To where went Maria

42) How did Elena travel COmo Viajaba Elena
How traveled Elena

Tense doubling shows up regularly in English Wh-question formation by less proficient
L2 English speakers whose native language is Spanish, who often produce strings such as
43) What did Joe bought

instead of

44) What did Joe buy

In 43), both the main verb and the auxiliary are morphologically marked for tense.

There are multiple linguistic factors that I could have focused on in these tense
doubling data; the two variations shown in 43) and 44) are not the only possibilities with
the construction of this type of string. In addition to the variation in tense, there can also
be variation in the use of the auxiliary; that is, it can be present in the data or not, or it can
have been moved, or not. The table below shows the range of responses that are possible.

The first example response listed shows no auxiliary present, and no past tense marked

50

on the main verb, the second shows no auxiliary present, but with the main verb marked

with past tense, and so on:

Table 7: Possible Range of Responses for Tense Doubling Data

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Auxiliary
- - = not present Main Verb
M = movement --- = no past tense Example
T = past tense + = past tense
-0- --- what Joe buy
-0- + what Joe bought
+M +T --- what did Joe buy
+M +T + what did Joe bought
+M —T --- what do Joe buy
+M —T + what do Joe bought
-M +T --- what Joe did buy
-M +T + what Joe did bought
-M —T --- what Joe do buy
-M —T + what Joe do bought

 

In this study, however, my focus was strictly on whether or not there was
evidence of tense doubling in the strings produced by the participants in this study. That
is, whether they produced strings such as in
46) What did Joe buy
or whether they produced strings with the tense doubled, as in
47) What did Joe bought

Again, the question of potential developmental factors interacting with the tense
doubling data is of possible concern. As noted, the production of sentences like 46)
necessarily entails more than just knowledge of the distribution of tense in such questions
in English—Wh-movement is also involved, as is the use of the auxiliary do, which could

potentially confound the tense doubling results, if there were developmental factors

51

involved. Radford’s (1990) argument (mentioned above) that functional categories are
acquired from about two years of age has direct bearing on this issue, because if he is
right, then the functional category systems involved in Wh-movement and the use of the
auxiliary do would be acquired by the age of my youngest subjects. Relating to the
issues involved in the production of these sentences, and providing evidence for
Radford’s claim, is work by De Villiers (1978), who claims that the full auxiliary system
of English is in play around age four, and that simple past tense is one of the earliest verb
expansions learned by children. Radford (1995) adds additional support, providing data
from children aged 26 months, who evidenced acquisition of auxiliaries such as do, tense
and agreement affixes, and auxiliary inversion. Data from another De Villiers (1995, p.
519) study, which is drawn from children ranging in age from 2;10 to 5;5 years, Shows
the production of sentences like:
48) What do you wanna eat
49) Who do you want to take a walk
which speaks to the question of the acquisition of Wh-movement.

With the tense doubling data, I expected to find evidence of morphological
attachment on the main verbs in the English Wh-questions of my subjects, again
progressing from more of it (i.e., more Spanish-like morphological attachment) to less of

it, depending on proficiency in English.

52

METHODOLOGY

Participants

The participants in this study were Spanish/English bilingual children who are
Spanish dominant, drawn from three different grade levels—first grade, third grade, and a
fourth/filth grade combination class, in an elementary school in Watsonville, which is
located in an agricultural region of California. Forty one subjects participated in this

study, with the following breakdown, by total number at each grade level, age, and sex:

Table 8: Participant Age and Sex, by Grade

 

 

 

 

 

 

 

 

 

 

 

 

Grade level # of Median Age Sex

participants Female Male
First 13 7 5 (38.5%) 8 (61.5%)
Third 14 9 6 (43%) 8 (57%)
Fourth/Fifth combo 14 10 6 (43%) 8 (57%)
TOTAL/AVERAGE 41 8.7 17 (fl.5%) 24 (58.5%)

 

Taken together, this pool of participants represents a fairly homogeneous group,
with regard to several factors, for instance, location of birth. The majority (90%) of these
children are first generation American-born, with the majority of their parents being
immigrants from México. And of these, 31 of the 41 children, or 75.6%, were born in
Watsonville. For all of the participants, and their parents, the information on birth
location was drawn from school records; while the data on the participants is complete,
two of the 41 participants’ parents’ information is missing. The table below breaks down

the detail of these data.

53

Table 9: Birthplace of Participants and Parents

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PARTICIPANTS

Where born Total number % Cumulative %
Watsonville 3 1 75.6% 75 .6%
Other location in California 6 14.6 90.2%
Mexico 4 9.8 100%
TOTAL 41 100%

PARENTS

Where born (Parent I/Parent 2) Total number % Cumulative %
México/Me’xico 37 87.8% 87.8%
Mexico/California 1 2.4% 90.2%
Mexico/Unknown 1 4.9% 95.1%
Unknown/Unknown 2 4.9% 1 00%
TOTAL 41 100%

 

All of the 41 participants in this study, across all grade levels, began their

schooling in kindergarten, in the same school district as their current school, indicating

that most of them have lived only in Watsonville, given the data in Table 9 above.

Further, 27 of the 41 participants, or 65.9%, have attended only this school, indicating

that they have lived in the same neighborhoods since they started school.

Another factor complicates this picture, however. While a majority of these

participants were born and raised and have attended school in Watsonville, a majority of

them are also classified as migrants. The following table illustrates four classifications of

these participants within the category of migrant status, as determined by the school.

54

 

Table 10: Migrant Status

 

 

 

 

 

 

Catpgory Number % Cumulative Total
Migrant 31 75.6% 75.6%
Not yet decided 4 9.8% 85.4%
No longer Migrant 2 4.9% 90.3%
Non-migrant 4 9.8% 1 00%
TOTAL 41 100%

 

 

 

 

 

 

Those classified as migrants (75.6%) are shown in the first row of this table. The second
row in the table represents those participants the school was still in the process of
deciding about—the migrant program personnel were investigating whether or not the
family met the criteria, mentioned above in the Literature Review, for migrant status.
The third row in the table indicates those participants who once were classified as
migrants, but who have been reclassified because their families no longer meet the
requirements to qualify for this status; finally, the fourth row shows the participants who
were classified as non-migrant—9.8%.

Another classification that the school uses with its student population is whether
or not they qualify for free or reduced lunches. This determination is based on family
income and follows the criteria set forth in the California Department of Education’s
“Income Eligibility Guidelines for Free and Reduced Price Meals or Free Milk,” in this
case, guidelines that are effective from July 1, 2000 through June 30, 2001 (see Appendix
B). The table below shows the breakdown for the students in this study, with a majority
of them, 73.2%, qualifying for free lunch, and another 14.6% qualifying for reduced

lunch, for a combined total of nearly 90% of the students:

55

Table 11: Free/Reduced Lunch Status

 

 

 

 

 

 

Cagggry Number % Cumulative Total
Free Lunch 30 73.2% 73.2%
Reduced Lunch 6 14.6% 87.8%
Denied 3 7.3% 95.1%
No Data 2 4.9% 100%
TOTAL 41 100%

 

 

 

 

 

 

My assumption, given the information detailed in Table 9 above, was that the
home language for at least the majority of these participants would be Spanish. To
confirm this, I embedded questions about home-life into my casual conversations with
these participants. Essentially, I asked them whether they spoke Spanish or English at
home, and with whom. I coded their responses as either 1=Spanish; 2==Spanish and
English (with Spanish mentioned first, and/or more often); and 3=English and Spanish
(with English mentioned first, and/or more often). The table below indicates, according
to each participant’s reporting, the breakdown on home language; all participants
reported Spanish to be a language of the home, with 66% claiming solely Spanish, and

another almost 30% claiming Spanish before English:

Table 12: Home Language

 

 

 

 

 

Category reported Number % Cumulative Total
Spanish 27 65.9% 65.9%
Spanish/English 12 29.3% 95. 1%
English/ Spanish 2 4.9% 100%
TOTAL 41 100%

 

 

 

 

 

 

56

School Site

The Pajaro Valley Unified School District has 24 schools in total, 16 of which are
elementary schools. The school where all data for this study were collected is one of
these elementary schools. It is an older school, built in the middle of last century. Over
the years, because the original buildings could no longer accommodate the growing
student population, additional classrooms have been added to the school site. In addition,
there is a large multi-purpose room that is used as a lunchroom daily, and when needed,
as an auditorium for school assemblies, parent meetings, etc. The schoolyard is also
large, with a separate play-area designated for the youngest students, the kindergarteners.
In the larger yard, there are swings and a jungle-gym located in a sandy area. There is
also a black-top area, containing several basketball courts. Another prominent feature of
the yard is a large grassy field, which is used for various sports—soccer being the game 1
most often saw played there. Recesses are staggered by grade-level: the first, second, and
third graders have recess at one time, and the fourth and fifth graders at another.

At the time of data collection, there were just over 600 students enrolled at this
school. These students were enrolled in 28 classes in total, 11 of which were bilingual
classes, 8 structured English immersion classes, and 9 English-only classes. The class
levels ranged from kindergarten through fifth grade. The following table, drawn from
data contained in the Pajaro Valley Unified School District Data Source Book, 2000-

2001, reflects several demographic characteristics of the students enrolled at this school:

57

Table 13: School Site Demographics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

% of Total
DEMOGRAPHICS Enrollment
Race
% Hispanic 89.6
% White 9
% Other 1.4
Socio-Economic Level
% Free and reduced lunch 80
Enflsb Proficiency
% Limited English 62
% Fluent English (designation for NNS of English) 12
% English Only 26
Migrant Status
Migrant students 39

 

Preliminary Preparation for the StudsI

Because of my previous affiliation with The Paj aro Valley Unified School
District, arranging for data collection was somewhat easier than if I hadn’t had any
contacts in the district. Having taught there for four years, at two different schools, I
knew that the district was the kind of place that would provide the opportunity for me to
collect the type of data that l was targeting. That is, I knew that I would find a population
and a community within the district that met my basic criteria—that the participants be
Spanish dominant, and that the community be at least somewhat agricultural, so that data
on migrant children could be collected.

I contacted an administrator at one of the elementary schools in the district.
Although I had never worked with this person, I had worked with several teachers at that
site. After one of these teachers briefly told the administrator about me, I sent her an

email, introducing myself and the project. Although she at first expressed some interest

58

in having me collect data at her school, she ultimately did not allow this to happen,
claiming that she really wanted the research done at her school to be done by the school’s
own teachers.

I subsequently learned that one of the colleagues I had taught with, at both of my
previous school Sites, had been recently appointed as a school site administrator within
the district. I contacted this person and briefly explained my project to her—including
logistical criteria I had, such as which grade levels I wanted to target, etc. I was very
pleased with her response to my request; she was not only willing to have me collect my
data at her school site, but was happy for me to do so, saying that she very much
supported such work. She did, however, leave the final decision up to the three
classroom teachers that would be involved. She herself framed the project for these
teachers, and luckily obtained permission from them for me to spend one week in each of
their classrooms collecting data from their students. Each of these teachers turned out to
be as cooperative as the school principal had been—they opened their classrooms and

their files to me.

Methods

The data collected for this study were of three general types: 1) linguistic
information, obtained from experimentally controlled data; 2) information on
participants’ identities, gained through casual conversation; and 3) demographic
information, obtained from school records.

Starting with the older students, I worked my way through the designated class
levels, spending one full week in each. In each class, the teachers had chosen 15 students

to participate in my study. I spent recesses and lunch periods with all the students from

59

each class, and in addition, pulled the students who were targeted for my study out of the
classroom, one-by-one, for audio taped data collection sessions. For this, the school
principal set me up in an office that was not in use. There I was given a desk, where I set
up my instruments and my tape recorder. Each student visited me in that office several
times, for the experimental sessions, and for the casual conversations, all of which were
audio recorded. In the experimentally controlled sessions, participants responded, in
English, to a series of experimental tasks that targeted the morpho-syntactic units
outlined above. In addition, I also recorded, in a field notebook, data obtained through
conversations with the children on the schoolyard or at lunch.

Elicitation Methods

In the collection of the N/N Compounding and the Article data, this study used
two elicitation techniques: 1) Repetition, and 2) Picture Task. These were designed to
elicit spontaneous, unmonitored speech containing the targeted linguistic structures (c.f.
Wolfram & Fasold, 1974; McDaniel et al., 1996; Radford, 1990).

In the tasks that used repetition, after being instructed on how to proceed, the
participants repeated the target sentences, after I had stated them. With the picture tasks,
the participants responded to questions such as, “What is this?” or “What happened
here?”, while looking at particular pictures. For a list of the actual target structures,
elicitation sentences, elicitation techniques, questions, and pictures used for the N/N
Compounding and the Article Usage data, please see these Appendices:

0 Appendix C: Noun-noun compounding data

0 Appendix D: Article usage data

60

The elicitation of the Tense Doubling data was less straightforward. In order to

elicit the targeted structure, for example,

50) What did Joe buy

(repeated from above, #40 in the Literature Review), I had to create a context for the
participants to ask such a question in. I began with a practice session or two with each
participant, being very careful to not produce the targeted structure in my set-up. For
example, I explained to the participants that we were going to play a game in which I was
going to tell them about some event from last night, and they were going to ask me a
question about that event. I then said, for example, “The dog wanted to chew something
last night, so he did,” and told the participants to ask me a question, based on the
information I’d given them, starting with what.

The older participants, the 5‘“, 4th and most of the 3rd graders, were able to
proceed right through this task, after one or two examples of how it was done, though
some of the 3rd graders required several examples up front, to be able to complete the
task. However, many of the 1St graders were not able to understand the set-up, or what it
was that they were to do in our sessions; with these participants, I changed strategies. I
set the targeted context up in English, and then asked, in Spanish, the question I wanted
them to ask in English; these participants were to translate the question I had produced,
into English. While this elicitation strategy, used with the 1St graders, is different from
the one I used with the older participants, I was nonetheless able to assess their ability to
construct the targeted questions in English, and more importantly, to assess whether or
not there was tense doubling in their questions. An example of the procedure used with

the younger participants follows:

61

a. I set up the context: “John went to the store last night and wanted to buy
something, so he did.”

b. I ask thquestion. in Spanish: “gQue' compro Juan anoche?”
What did Juan buy last night?

c. The participant translates this question into English.
The structures used with all participants in the collection of these Tense Doubling data,
regardless of elicitation method used, varied along several dimensions:

0 The amount of linguistic material between the head noun in the matrix clause of
the given context, for example, the difference between 1) and 2) below:

1) a. John went to the store last night and wanted to buy something, so he did
b. What did John buy last night?

2) a. The boy in the yellow hat went to McDonald’s last night and wanted to
buy something, so he did

b. What did John buy last night?
0 The type of Wh-question used: what, where, how

0 The type of verb used as the targeted main verb in the question response; these
verbs were of three general types:

1) completely irregular: buy/bought
2) alveolar: play/played
3) bisyllabic: want/wanted
For a list of the actual questions used in the collection of these data, as well as what type
of verb was used, and in what environment, please see Appendices E and E. 1.
In collecting the identity data, I engaged in several casual conversations with each
participant, across the period of data collection, and in both English and Spanish.

Questions targeting the identity information were embedded in these conversations. The

62

variable I operationalized as Identity reflects the participants’ preferences across cultural

practices that break down into the following 5 general categories:

Favorite food(s)

Favorite music

Sports

Holiday foods eaten by the family

Preference for place—Mexico or Watsonville

In addition to these general categories, I also asked the participants whether or not they

had any friends who spoke only English. The nine sub-categories for the identity

questions I asked are as follows:

51)

a. Favorite food

b. Favorite singer

c. Favorite soccer team

d. Favorite American football team

e. Preference for soccer or American football

f. What the family ate for Thanksgiving

g. Preference for living in México or Watsonville
h. Which is better—Mexico or Watsonville

i. How many English-only (E/O) fiiends

The demographic information collected on each participant was provided by the

school and included the following categories:

52)

a. Sex
b. Grade level
c. Age
d. Where born

c. Where parents born

63

f. Migrant status

g. Grade first entered US school

h. Where first entered US school

i. Free lunch

j. Home language, as reported by subject
k. Whether go to Mexico (regularly)

I. When/how often go to México

In addition to these demographic data, I also obtained information from the school on the
school’s language scores for each participant. These scores were of two types:

1) Assessment of the student’s oral English language ability, and oral Spanish
language ability, as tested by the school’s Bilingual Resource Teacher

2) Assessment of each participant’s English, by his/her classroom teacher, along the
following measures: listening, speaking, reading, writing

Earns

As noted, in the experimental portion of this study, I elicited data from the 41
participants in English, targeting the language features described above, which instantiate
deep structural differences in English and Spanish: N/N compounds, article usage, and
morphological attachment (tense doubling). After scoring each response, 1 derived a
“total grammar” score, which I operationalized as a measure of each participant’s
proficiency in English. Although the grammatical data in this study are rich with detail,
for this research, I scored the responses as either “target-like,” or “non-target-like,” for

each set of grammatical structures as follows:

64

Table 14: Scoring Rubric for Language Features Studied

 

 

 

 

 

 

 

 

Scored as: % of total

Language Target- Non-target- # of score, by

Feature like like items structure
N/N Compounding l -0- 16 24%
Articles 1 -0- 38 58%
Tense Doubling 1 -0- 12 18%
Total 1 -0- 66 100%

 

 

 

 

 

In scoring the identity data, I was primarily interested in whether my subjects’

identities were more Mexican or

more Anglo—according to the categories I had used to

operationalize this variable. To this end, I scored the 9 sub-categories (listed above in

#51, repeated here) in the following way:

Table 15: Scoring Rubric for Identity Data

 

Identity sub-category

Range of scores possible

 

 

Mexican ‘1 47 Anglo

 

1.1. Favorite food

-2=Mex; -1=Bicult/Mex; 0=Bicult/Mixed;
1=bicult/Anglo;2=Anglo

 

1.2. Favorite singer

-1=Mex; 0=Mixed; 1=Anglo

 

1.3. Favorite soccer team

-1=Mex; 0= Other; 1=US

 

1.4. Favorite American football

team -1 don’t have/don’t like fb; l=have fav

 

1.5. Preference for soccer or

- 1=socc; 0=same; 1 =football

 

American football
1.6. What the family ate for -2=Mex; -1=Bicult/Mex; 0=Bicult/Mixed;
Thanksgiving 1=bicult/Anglo;2=Anglo

 

Watsonville

[7 Preference for living in México or -1=Mex; 0=same; 1=Wats

 

1.8. Which is better—Mexico o
Watsonville

r -1=Mex; 0=same; 1=Wats

 

 

 

 

1.9. How many English-only (E/O) -1=none; 1=1 or more

friends

TOTAL SCORE POSSIBLE -ll ‘4 ; +11
Mexican Anglo

 

 

65

 

All of the grammatical and identity data were transcribed from the audio tapes
and compiled from my field notebook, and scored. These data, along with a score for
each of the categories of the demographic data, and the school and teacher language
scores, were entered into an Excel spreadsheet.

In addition to this formal analysis of the grammatical and the identity data, I was
concerned throughout the data collection to gather the children’s subjective impressions
about the various communities that they interacted in. Toward this end, after I had asked
each participant the identity questions concerned with “place,” that is, the items 51 . g) and
51.h) above, I asked each one this set of questions:

1) What is the same about Mexico and Watsonville

2) What is different about México and Watsonville

These questions were designed to elicit open-ended responses; that is, I made no attempt
to classify them according to any pre-determined categories. I will discuss these data in

the Results/Discussion chapter of this dissertation.

66

RESULTS AND DISCUSSION

The central concern of this study has been the relationship between the acquisition
of English as a second language and identity. The prediction, at the outset, was that there
would be a strong correlation between English proficiency and Anglo identity. Before
addressing the main question of this research, which involves the two variables, Total

Identity Score and English Proficiency Score, let’s look at each of these separately.

FrequencLData: Total Identity and Total English Proficiency

First, considering identity, we can see from Table 16 below, that these data have a
fairly normal distribution. The total range possible for this score was -11 to +11, with the
negative numbers indicating Mexican identity and the positive numbers Anglo. The
range of scores for the participants in this study were from —10, on the Mexican identity
side, to +6, on the Anglo. The table below Shows that 63% of the participants
demonstrated varying levels of Mexican identity, almost 10% were neutral (scoring —O-),
and 27% demonstrated varying levels of Anglo identity.

Table 16: Total I Score—F

Score F Percent Cumulative Percent
-l

     

2.
4.
12.

22.
31.
43.
63.
73.
78.
90

p—I

97

2.
2.
7.
4.
4.
9.
12.
19.
9.
4.
2
4.
2.
2.
0.

.—
O

67

Drawing from similar frequency data, we see from Table 17 that the English
proficiency data also show a fairly normal curve. The total range possible for this score
was from 0-66. The range of scores found for the participants in this study was from 23
to 61. It is interesting to note that all of the participants’ scores were in the upper 2/3 of
the range possible, with no one falling into the lower 1/3, indicating that all of them were
at least somewhat proficient in English.

Table 17: English Proficiency Score—Frequency

Score F Percent Cumulative Percent

2.
2.
2.
2.
2.
2.
2.
4.
4.
4.
2.
7.
2.
9.
2.
2.
4.
2.
2.
2.
4.
4.
2.
4.
2.
2.
2.
2.
2.
0.

10

 

68

In attempting to get as broad a sense of these data as possible, I conducted various
analyses of variance (ANOVA) among different groupings of the data, looking for
significant and non-significant differences among the groups. What follows is, 1)
analyses of the identity variable, compared across several other categories of the data,

and 2) analyses of the English proficiency variable, across the same set of categories.

AN OVAs: Total Identity Score afli Five Factors

Considering first the total identity score as the dependent variable, I compared
this with several standard sociolinguistic categories—themselves larger group identity
factors: sex, age, and socioeconomic status. Table 18 below shows the difference
between sex and identity, which was not significant (p=0.875). As noted above, the total
range possible for the identity score was -11 to +11; male mean identity score was —1.17;

female mean identity score was —1.35.

Table 18: ANOVA, Total Identity Score and Sex

 

 

 

 

 

 

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups .345 1 .345 .025 .875
Within Groups 533.216 3 13.672
Total 533.561 4

 

 

The participants in this study were grouped into three categories, according to
age: 6-7 year olds; 8-9 year olds; and 10-11 year olds, a grouping that theoretically allows
for enough distinction to show trends, if they exist. The mean identity score for the 6-7
year olds was —1.83; for the 8-9 year olds, it was —1.36; and for the 10-11 year olds, it

was 0.67. These differences were not significant (p=0.715). Table 19 below reflects this:

69

Table 19: ANOVA, Total Identity Score and Age

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups 9.347 2 4.673 .339 .715
Within Groups 524.214 38 13.795

 

Total 533.561 4

 

 

 

 

 

 

 

 

The measure I used to gauge socioeconomic status was whether or not the
participants qualified for the free/reduced school lunch program. As noted in the
Methods section above, this qualification is based on the California Department of
Education’s “Income Eligibility Guidelines for Free and Reduced Price Meals or Free
Milk,” in this case, guidelines that are effective from July 1, 2000 through June 30, 2001
(see Appendix B). These participants were classified dichotomously, as either qualifying
for free or reduced lunches, or not. The mean identity score for those qualifying for
free/reduced lunch was —1.36, those denied had a mean identity score of —1 .33. Again,
the difference between these scores is not significant (p=0.990). The table below shows

the results of the AN OVA.

Table 20: AN OVA, Total Identity Score and Lunch Status

 

 

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups .002 1 .002 .000 .990
Within Groups 514.972 37 13.918
Total 514.974 38

 

 

 

 

 

 

In addition to these variables, I also conducted ANOVAS on home language and
identity, and migrant status and identity. Neither of these showed significant differences.

As noted above, the data for the home language variable were categorized in three ways:

70

1) respondents who claimed Spanish as the home language; 2) respondents who claimed
Spanish and English, listing Spanish first, or as most-often used; and 3) respondents who
claimed English and Spanish, with English listed first. The mean identity score for the
first group was —1.11, for the second, -1.50, and for the third, -1.50 as well. These

differences were not significant (p=0.951). Table 21 below reflects this.

Table 21: AN OVA, Total Identity Score and Home Language

 

 

 

 

 

 

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups 1.394 2 .697 .050 .951
Within Groups 532.167 38 14.004
Total 533.561 40

 

 

The data for the migrant status variable were coded dichotomously; the
participants were classified as either 1) migrant, or 2) all others. This latter category
included those who a) were no longer migrants (parents no longer working in the
officially recognized migrant occupations, or the family no longer making annual treks to
México), b) were classified by the school as non-migrant, and c) the school was still
deciding about. Those classified as migrants had a mean identity score of—1.35, and
those who were not had a score of -.90, a difference that is not significant (p=0.73 7), as

shown in the table below.

Table 22: ANOVA, Total Identity Score and Migrant Status

 

 

 

 

 

 

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Grogas 1.564 1 1.564 .1 15 .737
Within Groups 531.997 39 13.641
Total 533.561 40

 

 

71

All five of the above listed AN OVAs were one-way, in that they operated on only
one independent variable. As we see, not one of the five larger group identity factors
tested showed significant difference, when compared with the identity variable in this
study.

A Closer Look at 3 of the 9 Identity Sub-categories

Before moving on to the analyses of the English proficiency data, I would like to
briefly describe an aspect of the identity variable, discussing here three of the nine sub-
categories of this measure. These centered on preferences with regard to sports, designed
to elicit participant responses as to 1) their favorite soccer team,5 2) whether they had a
favorite American football team, and 3) whether they preferred soccer or football. When
I had originally designed the elicitation questions for the identity sub-categories, I wanted
to focus one section on sports, since there tend to be strong allegiances around this
cultural practice. I deemed that soccer would be a good measure, and asked the
participants who they would like to see win the World Cup. Naively, I assumed that their
responses would fall into the super-category of nation—either Mexico or the United
States. What I found, however, was that not one of the respondents associated soccer
with the United States. Instead, the teams that they named, that is, Atlas, Cruz Azul,
Chivas, and America, were regional teams from within Me'xico (actually, the first three
listed are afiiliated with Guadalajara and Mexico City, and the last, America, is a national
team). While I found this fact to be quite interesting, I nonetheless had to develop other

questions to get at the data I wanted. Because soccer was so clearly affiliated with

 

5 The assumption I made with regard to favorite soccer team was that their choices would be either
Mexican or American teams

72

México for these respondents, I surmised that juxtaposing it to another obvious cultural
symbol, American football, would serve my purpose. So, I asked the participants what
their favorite American football team was, and then further, which sport they preferred.
The tables below reflect the frequency data on these sub-categories.

In Table 23, we note that almost 98% of the participants have a favorite soccer
team that is Mexican. The one person whose response was coded as “other” named
Brazil as his favorite team. When interviewed on this topic, most of the respondents

named their favorite team quickly and with much animation, boys and girls alike.

Table 23: Favorite Soccer Team—Frequency

 

 

 

 

 

 

 

 

Team Frequency Percent Cumulative Percent

Mexican 40 97.6 97.6
Other 1 2 .4 1 00.0
Total 41 1 00.0

 

 

Table 24 below shows the responses on the question about American football.
Here, in contrast to the soccer data, almost 37% of the participants indicated that they
don’t have a favorite team, or that they don’t like American (which they called “tackle”)

football.

Table 24: Have Favorite American Football Team—Frequency

 

 

 

 

 

 

 

Responses Frequency Percent Cumulative Percent

Have 26 63.4 100.0

Don’t have/don’t like 15 36. 36.6
Total 41 100.

 

 

 

73

The data in Table 25 reflect the same trend of a preference for soccer over
football. Sixty eight percent of the respondents prefer soccer, 24% prefer football, and

7% like both.

Table 25: Prefer Soccer or Football—Frequency

 

 

 

 

 

 

Preference Frequengy Percent Cumulative Percent
Soccer 28 68.3 68.3
Both 3 7.3 75.6
Football 10 24. 100.0
Total 41 100.0

 

 

 

 

 

ANOVAs: Total English Proficiency Score and Five Factors

Turning now to the other main variable in this study, English proficiency,
additional one-way ANOVAS were conducted, using this measure as the dependent
variable, and the same five independent variables listed above (sex, age, socioeconomic
status, home language, and migrant status). Again, the total range possible on the English
proficiency score was from 0-66. The mean English proficiency scores for sex have the
following breakdown: males—45.21, females—43.53. These differences are not

significant (p=0.612), as shown in the table below.

Table 26: ANOVA, English Proficiency Score and Sex

 

 

 

 

 

 

 

 

 

 

Sum of df Mean Square F Sig.
ngares
Between Groups 28.050 1 28.050 .261 .612
Within Groups 4190.194 39 107.441
Total 4218.244 4

 

 

74

Age, on the other hand, does show a significant difference (p=0.000). The mean
English proficiency score for the 6-7 years olds was 34.92, the mean score for the 8-9
year olds was 45.36, and for the 10-11 year olds it was 51.40. Table 27 below shows the

results of this ANOVA test.

Table 27: ANOVA, English Proficiency Score and Age

 

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups 1826.513 2 913.256 14.510 .000
Within Groups 2391.731 38 62.940
Total 421 8.244 40

 

 

 

 

 

 

 

Rounding out this picture on the age variable is data Obtained from each of the
classroom teachers whose students were under study. These teachers had recently
assessed their students’ English language ability, on a measure that ran from 1 through 20

possible points. The table below reflects a crosstabulation of these data:

Table 28: Crosstabulation of Teacher’s Overall English Score

 

 

 

 

 

 

Teacher Current Overall English Score: Range 1-20
Grd Tot
I I 6 7 8 9 10 ll 12 13 14 15 16 17 18 19 20 Part
I

1 2 5 3 2 l 13

l 2 2 3 S l 14

4/5 1 2 6 2 l l 14

Total 1 2 5 5 4 3 6 1 1 2 6 2 l 1 1 41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The general trend with the combined teacher data is that the older students

demonstrate more proficiency in English than the younger ones. This, combined with the

75

results from the AN OVA run on age and English proficiency, discussed above, seemingly
indicate that these participants’ acquisition of their second language, English, is
progressing over time, which is to be expected. Of course these data do not confirm this,
because they don’t reflect these participants’ proficiency over time. However, the data in
the table that follows do speak to this issue. These are data from the English language
tests that were administered to these participants by the school when they first enrolled,
which for most was in kindergarten. For the first grade participants, that was only one
year before; however for the older children, these initial Enlish assessments done by the
school date back several years. Table 29 below shows that 34 of the 41 participants, or
83%, received scores in the lowest two levels of this test; thirty two, or 78% of the total,

scored in the lowest category.

Table 29: Crosstabulation of School Initial English Scores

 

 

 

 

 

 

 

School Initial English Scores: Range 1-5
GRADE TOTAL
1 1 2 3 4 I
1 7 2 l 3 13
3 13 1 14
4/5 12 2 14
Total 32 2 3 4 41

 

 

 

 

 

 

 

Adding to the developing picture of these participants’ language, Table 30 below
shows their Spanish language, as assessed by the school, again when they first enrolled.
These data show that these students were Spanish dominant, in that 98% scored in the

highest two levels.

76

Table 30: Crosstabulation of School Initial Spanish Scores

 

 

 

 

 

 

 

School Initial Spanish Scores: Rang 1-5
GRADE 3 4 5 TOTAL
1 l
l 10 3 13
3 10 4 14
4/5 1 9 4 14
Total 1 29 11 41

 

 

 

 

 

 

Returning to the set of ANOVAs that were run with English proficiency as the
dependent variable, when compared with the third factor, socioeconomic mean scores (as
measured by free/reduced lunch qualification), we find again that there is no significant
difference (p=0.734), as shown in Table 31 below. The mean English proficiency scores
for this measure were: 1) for those qualifying for free/reduced lunch—44.22, and 2) for

those denied—46.33.

Table 31: ANOVA, English Proficiency Score and Lunch Status

 

 

 

 

Sum of df Mean Square F Sig.
Squares
Between Groups 12.342 1 12.342 .117 .734
Within Groups 3904.889 37 105.538
Total 3917.231 38

 

 

 

 

 

 

 

 

The mean English proficiency scores for the home language data were as follows:
those claiming Spanish—44.41, those claiming Spanish and English—44.67, and those

claiming English and Spanish—45.00. These differences are not significant (p=0.995):

77

Table 32: AN OVA, English Proficiency Score and Home Language

 

 

 

 

Sum of Df Mean Square F Sig.
Squares
Between Groups 1.059 2 .529 .005 .995
Within Groups 4217.185 38 110.979
Total 421 8.244 40

 

 

 

 

 

 

 

 

Finally, we look at migrant status, compared with English proficiency. Here, we
do find a significant difference (p=0.004). As noted in the Methodology section of this
research, 31 of the 41 participants (75.6%) were classified as migrants. These
participants had a mean English proficiency score of41.97, out of a total possible score
of 66, while those not classified as migrant had a mean score of 52.40. The results of this

ANOVA are reflected in the table below.

Table 33: ANOVA, English Proficiency Score and Migrant Status

 

 

 

 

Sum of (If Mean Square F Sig.
Squares
Between Groups 822.876 1 822.87 9.452 .004
Within Groups 3395.368 39 87.061
Total 4218.24 40

 

 

 

 

 

 

 

 

These results are to be expected. As discussed above, the school’s criteria for
being classified as migrant include having parents who work in an officially designated
migrant occupation, and traveling to Mexico annually. This last criterion is further
specified: In order to qualify for the migrant program, these students need to stay in
México for at least 20 days, every year; as noted above, most of these migrants leave in
mid-December and return to Watsonville at the end of January (I. Vizcaino, personal

communication, December, 2000). Given this, by definition, the participants in this study

78

who were classified as migrants spend less time in an environment where English
competes with their Spanish; when in Watsonville, even though the majority of these
participants’ home language is Spanish, and they are predominantly Spanish speakers,
they also receive much English input. At school, they are involved daily in English
language development programs, they associate with classmates and teachers who are
English speakers, the language of the official business of the city they live in is English6,
much of the television they have access to is in English, etc. In contrast, during the time
that they are in Mexico, they are immersed in Spanish, whether or not they attend school
during their stay there. It would seem obvious that these participants’ English language
development and levels of proficiency would be affected by these trips to Mexico,

making the results above not surprising.

Correlation: English Proficierm am School/Teacher Scores

In addition to the analyses above, I wanted to assess the relationship between the
scores derived from the measure I used for English proficiency and those obtained
through the school’s testing of oral English proficiency (administered by the bilingual
coordinator), as well as those obtained through the teachers’ testing of the overall English
ability (which included both reading and writing scores) of their students. To this end, I
ran correlations between my measure and these two others. There was a fairly strong
correlation (.680) between the English proficiency score I obtained, and the school’s; this
was significant at the (p=0.01) level. The figure below displays a scatterplot analysis of

this correlation:

 

6 Several of the students remarked that they often had to act as translators for their parents, having to use
their English in conversation with people who would visit their homes—mail carriers and other delivery
people, for example.

79

Correlation of Scores: School English Score and English Proficiency Score

 

5.5

5.0‘ D an an:

4.5“
4.0- r: I: :1 l3 :1 c1 1:: I:
33‘
3.0“ D :1
2.5“
2.05 D

1.5

SCHOOL ENGLISH SCORE

 

 

 

W

20 3'0 40 5'0 60
ENGLISH PROFICIENCY SCORE

Figure 1: Correlation of Scores—School English and English Proficiency

As is clear from this graph, there is quite a bit of overlapping of scores on these
two measures, by participant. In addition, there is one possible outlier, which may be
affecting the results. This participant scored 23 on the English proficiency score (total
range possible 0-66), and 2 on the school English score (total range possible 1-5). I
decided to remove this participant’s information, and rerun the correlation. This run
showed a moderate correlation of .409, which was not significant.

The other correlation I ran, between my English proficiency scores and those Of
the teachers, was strong (.689), and significant (p=0.01). The scatterplot analysis for this

is shown in Figure 2:

80

Correlation of Scores: Teacher English Score and English Proficiency Score

 

 

 

 

o

U 30

m

:

VJ

S 20‘ “a

o a D a

Z I: n gt: a a

I-LI a a

E 10: a a a can 3": a" = a °

3 a a a an D a

e ° . .° . .
20 30 40 50 60 7o

ENGLISH PROFICIENCY SCORE

Figure 2: Correlation of Scores—Teacher English and English Proficiency

While the English proficiency measure used in this study, and the school’s
assessment of these participants’ English was only moderately correlated (Figure 1), the
strong correlation between the teachers’ assessment of their students’ English provides a

measure of external validity to the English proficiency variable used in this study.

Correlation: Total English Proficiencyand Total Identity

Now, with a broader sense of how the two main variables (English proficiency
and identity) interact with other classifications of the data in this study, we move to an
assessment of the relationship between these two variables, which, again, was the central
focus of this research. As noted in the Literature Review of this study, it was expected
that there would be a strong correlation between English proficiency and Anglo identity.
To test this, I ran a correlation between these two variables. The results of this analysis
Show very little correlation: (.119), which is not significant. Figure 3 below shows the

scatterplot analysis for these data:

81

 

 

 

 

:3 Correlation of Scores: Total Identity Score and English Proficiency Score
8 10.
m u
r: 0-1 :3 final: 2: an a u can: an
2 a u a a :1
L11 I: a
e a a E D
._1 ~10'l n
<
1-4
8 -20
20 30 40 50 60 70

ENGLISH PROF ICENCY SCORE

Figure 3: Correlation of Scores—Total Identity and English Proficiency

These unexpected results caused me to wonder about each of these main

variables, and to move to assess them separately.

Unexpected Results: Assess English Proficiency Vamble

First, considering the English proficiency scores, these showed a strong
correlation (0.689) with the teachers’ assessment of the participants’ overall English
(Figure 2 above), which, as noted, gives them some external validity—arguably
indicating that this variable actually measures proficiency. In addition, the three sub-
categories of the English proficiency scores (article usage, noun-noun compounding, and
tense doubling), when compared separately with the identity variable, still showed very
little correlation; these results are shown in the correlation matrix in Table 34 below, and
indicate that taking these sub-scores together as a total score is not unreasonable. It is
true that the article usage and noun-noun compounding variables do each show slight
correlations with identity, while the tense doubling data is slightly negatively correlated

with identity; however, none of these results are significant.

82

Table 34: Correlation—Total Identity Score/three sub-categories of Engl. Prof. Score

 

 

 

 

 

 

 

 

 

 

N-N Total Identity

Article Usgge Compounding Tense Doublin Score

Article Uspgc 1 .573** .403** .231

N-N .573** 1 .640** .210
Compoundin

Tense Doublin .403** .640M 1 -. 138

Total Identity .231 .210 -.138 1
Score

 

*"' Correlation is significant at the 0.01 level (2-tailed).

We can also see from the table above that the three sub-categories are all

 

modestly correlated with one another. Still, there are differences among these data that
are worth considering briefly. The following three frequency tables show the range of
scores for each grammatical element. The article data had a total score possible of 38; the
participants’ scores ranged from 22-3 8, with 12 of the 41 subjects (approximately 30%)
receiving the two highest scores possible, 37 or 38. Another 30% received the next

highest score of 36, reflecting that approximately 60% of the participants received scores

in the 95th percentile or above.

Table 35: Articles—Frequency Data

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Scores Frequency Percent Cumulative Percent
22 1 2.4 2.4
25 1 2.4 4.9
26 1 2. 7.3
28 1 2.4 9.8
29 1 2.4 12.2
30 1 2.4 14.6
31 1 2.4 17.1
32 2 4.9 22.0
33 2 4.9 26.8
34 2 4. 31.7
35 4 9.8 41.5
36 12 29.3 70.7
37 7 17.1 87.8
38 5 12.2 100.0
Total 41 100.0

 

 

83

The noun-noun compounding data had a total score possible of 16; the
participants’ range was from 0-14, which indicates that this feature was more difficult for
them than the articles were. Approximately 50% had a score of 6 or below (3 8th

percentile or below):

Table 36: Noun-noun Compounding—Frequency Data

Score F Percent Cumulative Percent
9.
17.1
26.
34.1
39.
48.
51
58.
65.
73
78.
87.
95.1
100.

11

9.
7.
9.
7.
4.
9.
2.
7.
7.
7.
4.
9.
7.
4.

1
Total 100.

 

The tense-doubling data had a total score possible of 12; the scores the
participants received ranged from 0-11, with almost 66% scoring at 5 or below (40‘h

percentile or below).

84

Table 37: Tense-doubling—Frequency Data

 

 

 

 

 

 

 

 

 

 

 

Scores Frequency Percent Cumulative Percent
0 18 43.9 43.9
2 3 7.3 51.2
3 1 2.4 53.7
2 4.9 58.5
5 3 7.3 65.9
3 7.3 73.2
8 2 4.9 78.0
9 5 12.2 90.2
1 l 9.8 100.0
Total 41 100.0

 

 

 

 

 

 

The scores listed above seem to indicate that these participants are much more
proficient in article use than in the use of the other two features studied. This could mean
that the articles were acquired early for these participants. However, there is another
possible explanation for the differences in the scores. Perhaps the elicitation task was
faulty. As can be seen in Appendix D, four of the five sub-categories of the article
variable used repetition to elicit responses; the fifth, inaliable possession used a picture-
task7; although repetition tasks are purported to assess grammatical competence (as
discussed in the Literature Review of this study), perhaps because in most of the prompts
given to the participants by the researcher the targeted element was one of the first words
spoken, it became overly salient to the participants in their repetition of the strings.
Unfortunately, the determination of which of these factors may be more at play here is
beyond the scope of this work. Moreover, given the general acceptance of repetition

tasks as a measure of grammatical competence, coupled with my desire to have as broad

 

7 The other variables did not use repetition; the noun-noun compounding data was elicited using only
picture tasks (please see Appendix C), and the tense-doubling data required that the participants produce
novel strings, not repeat ones given by the researcher (Appendix E).

85

a picture of the English grammatical competence of these participants as possible, I
retained all three grammatical features in my measure of English proficiency.
Unexpected Results: Assess Identity Variable

I continued my investigation into the unexpected results of a very weak
correlation between English proficiency and identity with a focus on the identity variable.
Based on advice from the statistical consultant in the MSU computer center, I pulled this
variable apart, running separate ANOVAS on each of the nine sub-categories of identity,
and comparing these with the English proficiency scores. The aim here was to review the
mean scores of these measures, and to look for trends. I suspected that perhaps these sub-
categories were pulling in different directions, and therefore neutralizing the total effect.
Again, these sub-categories were:
53) a. Favorite food

b. Favorite singer

c. Favorite soccer team

d. Favorite American football team

e. Preference for soccer or American football

f. What the family ate for Thanksgiving

g. Preference for living in México or Watsonville

h. Which is better—Mexico or Watsonville

i. How many English-only (E/O) friends

Not one of these tests was significant, and as expected, the scores did pull in

different directions. What follows are three sets of tables, each set including one table

86

that groups the sub-categories of identity that Show similar trends, followed by the

AN OVA source tables for each of these sub-categories.

Non-significant Trends in the Data

The first set of tables Show the mean scores for those sub—categories of the
identity variable that show a trend that is opposite what was expected in this study, that
is, for those participants showing more Mexican identity, as measured in this study, and
higher English proficiency scores; again, the highest score possible for English

proficiency was 66:

Table 38: Mean Scores for Identity Sub-scores with trends opposite expected

 

 

 

 

 

 

 

 

 

 

Sub-category: Which is better—Mexico or Watsonville

Type of response Number Mean English proficiency score
Mexico 19 45.42
Same 5 45.20
Watsonville 17 43.29

Sub-category: Favorite soccer team

Type of response Number Mean English proficiency score
Mexico 40 44.73
Other 18 36.00

 

 

 

 

The AN OVA source tables for each of these measures follow:

Table 38: ANOVA—Better/México or Watsonville and English Proficiency Score

 

 

 

 

 

 

 

 

 

 

Sum of dfl Mean Square Sig.
Squares
Between Groups 43.283 2 21.641 .197 .822
Within Groupsl 4174.961 38 109.867
Total 4218.244 4

 

 

 

8 As noted, this one person who did not name a Mexican team as his favorite, named Brazil.

87

Table 39: AN OVA—F avorite Soccer Team and English Proficiency Score

 

 

 

 

 

 

 

 

 

 

Sum of d 1 Mean Square F Sig.
Squares
Between Group 74.269 1 74.269 .699 .408
Within Groupsl 4143.975 39 106.256
Total 4218.244 40

 

 

 

The scores for three of the sub-categories of the identity measure showed non-

linear trends. Below is a table showing the mean English proficiency scores for these:

Table 40: Mean Scores for Identity Sub-scores with non-linear trends

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sub-cute ory: What the family ate for Thanksgiving”
Type of response Number Mean English proficiency score
Mexican food 10 46.30
Bicult/Mex food 1 33.00
Bicult/Mixed l 5 42.67
food
Bicult/Anglo 1 3 l .00
Anglo 14 47.00
Sub-catggo : Preference for soccer or American football
Type of response Number Mean English proficiency score
Soccer 28 44.89
Same 3 41.67
Football 10 44.30
Sub-cateflry: Preference for living in Mexico or Watsonville
Type of response Number Mean Englishgproficiency score
México 18 44.06
Same l 45.00
Watsonville 21 44.62

 

 

 

 

 

9 The responses regarding Thanksgiving food were coded as a) Mexican, b) Bicult/Mex, c) Bicult/Mixed,
d) Bicult/Anglo, and e) Anglo. Responses like tacos al pastor, pozole, were scored as Mexican; responses
naming foods from places such as McDonald’s and Burger King (which are well established in México)
were scored as Bicultural; responses such as turkey, mashed potatoes, pumpkin pie were scored as Anglo;
responses with a mixture of Mexican and Anglo foods were scored as Mixed.

88

The AN OVA source tables for each of these measures follow:

Table 41: ANOVA, What family ate for Thanksgiving and English Proficiency Score

 

 

 

 

 

 

 

 

 

Sum 03 dfl Mean Square F Sig.
Square
Between Groupsl 484.811 121.203 1.169 .341
Within Groupsl 3733.433 36 103.706
Tota 42182441 40

 

 

 

 

Table 42: AN OVA, Pref. for Soccer or Amer. Football and English Proficiency Score

 

 

 

Sum 03 (111 Mean Square F Sig.
Square
Between Groups] 28.799 2 14.399 .131 .878

 

Within Groupsl 4189.445 38 110.249
Total 4218.244 40

 

 

 

 

 

 

 

 

 

Table 43: AN OVA, Preference for living in Me'xico or Watsonville

 

 

 

Sum 03 dfl Mean Square F Sig.
Square
Between Groups] 3.478 2 1.739 .015 .985

 

Within Groupsl 4183.897 37 113.078
Total 4187.375 39

 

 

 

 

 

 

 

 

 

Finally, four of the nine sub-categories of the identity variable showed scores that
had trends in the expected direction. That is, the higher the scores were in English
proficiency, the more Anglo the identity. The table below shows the mean English

proficiency scores for these four sub-categories:

89

Table 44: Mean Scores for Identity Sub-scores with expected trends

 

Sub-category: Favorite singer

 

 

 

 

 

Type of response Number Mean English proficiency score
Mexican 5 42.33
Mixed 7 43.57
Anglo 24 45.17

 

 

Sub-category: Favorite American football team

 

 

 

 

Type of response Number Mean English proficiency score
Don’t have/like 15 43.33
Have fav team 26 ' 45.19

 

 

Sub-category: Favorite food

 

 

 

 

 

 

Type of response Number Mean English proficiency score
Mexican food 5 39.40
Bicult/Mex food 13 45.38
Bicult/Mixed 21 46.29
food
Bicult/Anglo 1 23.0010

 

 

Sub-category: How many English-only (E/O) friends

 

 

 

 

 

Dpe of response Number Mean English proficiency score
No EO friends 25 42.24
1+ EO friends 16 48.06

 

 

The source tables for each of the AN OVA tests run on these four sub-categories

of the identity variable are listed below:

Table 45: ANOVA, Favorite singer and English Proficiency Score

 

 

 

 

 

 

 

 

 

 

 

 

Sum 03 dtl Mean Square F Sig.
Square
Between Group 44.538 2 22.269 .195 .823
Within Groupsl 3876.381 34 114.011
Total 3920.91 36

 

 

'0 This one participant’s score does not show the same trend as the others in this category. I reran an

ANOVA with this person’s scores removed. The results were still not significant (0.385). I’ve included the
information from the original ANOVA test here in this table for accuracy.

90

 

Table 46: ANOVA, Favorite American football team and English Proficiency Score

 

 

 

Sum 03 dll Mean Square F Sig.
Square
Between GroupsL 32.872 1 32.872 .306 .583

 

WithinGroupsl 4185.372 39 107.317
Total 4218.244 40

 

 

 

 

 

 

 

 

 

Table 47: AN OVA, Favorite food and English Proficiency Score

 

 

 

 

 

 

Sum o (I Mean Square F Sig.
Square
Between Groupsl 669.337 3 223.112 2.265 .098
Within Grog; 3546.563 36 98.51
Total 4215.900 39

 

 

 

 

 

 

 

Table 48: ANOVA, English-only Friends and English Proficiency Score

 

 

 

 

 

 

Sum 03 df Mean Square F Sig.
Square
Between Groupsl 330.746 1 330.746 3.318 .076
Within Groupsl 3887.497 39 99.679
Total 4218.244 40

 

 

 

 

 

 

 

Before discussing the results above, I would like to further comment on the
identity variable from two other points of reference, in order to more fully frame this
measure.

Crosstabulation of Total Identity Scores by Grade

First, a crosstabulation of the identity scores for all of the 41 participants, by
grade level, shows that these scores are fairly widely distributed (Table 41 below). The
first grade identity scores range from the lowest to the highest scores given (-10——+6).
The scores for the other two levels are also fairly spread out, even though there is some

clustering around —1 or -0-:

91

Table 49: Crosstabulation of Total Identity Scores by Grade Level

 

 

 

 

 

 

 

 

Total Identity Score
Grd Tot
-10 -8 -7 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
1 1 l l 1 3 1 1 1 1 1 1 1 3
3 l 1 2 l 3 2 2 2 14
4/5 1 l l 2 4 1 2 1 l 1 4
Tot 1 1 3 2 3 4 4 8 4 2 2 3 2 1 1 41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

When these identity results are compared to those for English proficiency, their
wide distribution across grades stands out. Below are three tables, which are derived from
a crosstabulation of English proficiency scores by grade level, for comparison. Because
the range of these scores is from 23-61, I was not able to capture these data in one table,
and had to distribute them across three. Still, it is quite clear that, unlike the identity data

above, the English proficiency data generally cluster by grade/age level.

Table50-a, b, c: Crosstabulation of English Proficiency by Grade Level

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

English Proficiency Scores: Range 23-50
23 25 26 29 31 33 35 36 41 45 50 Total i
T“ 1 1 1 l 1 1 1 2 1 2 1 13
English Proficiency Scores: Range 34-60
34 35 39 40 41 45 47 48 51 52 56 60 Totall
3"‘1 1 1 2 1 2 1 l 1 1 1 1 1 14
English Proficiency Scores: Range 43-61
43 45 46 48 53 54 55 56 57 58 59 61 Total I
"Em/5‘“ 1 1 1 1 2 2 1 1 1 1 1 1 14

 

92

 

The second issue I would like to comment on is regarding the validity of the
identity variable.
Validity of the Identity Variable

Given the unexpected results, perhaps the measure itself is faulty, and is not
actually measuring identity. Research done by Cuellar et a1. (1995) speaks a bit to this
issue, in a round-about way. This work, which focuses on acculturation and the
interactions of culture and personality, has resulted in the establishment of several
successful scales for measuring acculturation. One of these is relevant to the current
study; this is the Acculturation Rating Scale for Mexican Americans-II (ARSMA-H)
(Cuellar, Arnold, & Maldonado, 1995), which is composed of two sub-scales, an Anglo
Orientation Subscale and a Mexican Orientation Subscale, with total scores in either
direction indicating orientation toward the targeted group. While Cuellar et al.’s work
with this measure focuses on different social behavior than this study does, that is,
personality traits and characteristics, and is therefore not wholly relevant to this
discussion, the scale itself is. Although the ARSMA-II scale (please see Appendix F) is
not identical to the measure used here for identity, it has some bearing on the assembly of
social facts chosen to make up this variable. The ARSMA-H scale differs from the
identity variable used here in that it targets adults versus children, and in that it overtly
asks question about identity/identification—as opposed to the more covert method of
elicitation of identity information used in this study, which centered on the children’s
preferences. Nonetheless, there is a fairly good overlapping of the type of information
sought in both measures. As noted, there are six general categories that the nine sub-

categories of the identity variable used in this study fall into, repeated again here:

93

0 Favorite food(s)

0 Favorite music

0 Sports

0 Holiday foods eaten by the family

0 Preference for place—Mexico or Watsonville

0 English speaking fiiends

While the The ARSMA-II scale does not use the categories of sports and

preference for place, it does target three of the above-mentioned categories (four, if
favorite food and holiday foods are considered to be part of one general category—food)
in its questions—food, music, and English speaking friends. Given the good overlap of
sub-categories between the measure used here for identity and the well-established
ARSMA-II scale, I feel fairly confident that the identity variable used in this study is

valid.

Summary Tables

Several points fall out from the analysis above. Below are two summary tables of
the results from the ANOVA tests listed above, which provide accessible reference points
for the discussion that follows. The first table references the results from both the

identity score and the English proficiency score, each compared with five factors:

94

Table 51: Summary—AN OVA Results for Identity/English Prof. + Five Factors

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANOVA Results: Identity Score and Five Factors
Factor Significance
Sex 0.875
Age 0.715
Socioeconomic Status 0.990
Home Language 0.951
Migrant Status 0.737
ANOVA Results: English Proficiency Score and Five Factors
Factor Significance
Sex 0.612
Age 0000*
Socioeconomic Status 0.734
Home Language 0.995
Migrant Status 0004*

 

 

 

To briefly review, not one of the five factors listed in Table 51 above, when
compared with identity, showed any significant differences. However, when these five
factors were compared with English proficiency, two out of the five, age and migrant
status, showed significant differences (these are marked in the table). These results were
to be expected, as discussed above.

Recall that, given the unexpected result of no correlation between the two main
variables in this study (English proficiency and identity) separate ANOVAs were run on
the sub-categories of the identity variable and English proficiency, with the aim of
identifying trends in the data, if they existed. The summary table for the results of these

ANOVAs follows:

95

Table 52: Summary—AN OVA results for Engl. Prof. + 9 sub-categories of Identity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANOVA Results: English Proficiency Score and Nine Sub-categiries of Identity
Category Significance
Trend opposite than expected:
Which is better—Mexico or Watsonville 0.822
Favorite soccer team 0.408
Non-linear Trend:
What the family ate for Thanksgiving 0.341
Preference for soccer or football 0.878
Preference for living in México or Watsonville 0.985
Trend as expected:
Favorite singer 0.823
Favorite American football team 0.583
Favorite food 0.098
How many English-only (EO) friends 0.076

 

 

 

Please note that while this table maintains the distinction between trends in the
data in order to remain consistent with the earlier tables, this does not imply significance;
all of the data in this table reflect non-significant differences. Nonetheless, one of the
sub-categories listed, how many English-only (E0) fi'iends, shows a significance of
(0.076), which is close to the accepted level of (0.05). With not much of a stretch, one
could argue that these results were somewhat significant”. The discussion of this
variable (which I will heretofore refer to as EO) that follows, proceeds from this
assumption.

A Closer Look at the E0 Variable
The E0 score is interesting, in that in some sense, it is very much like a network

score, as characterized by Milroy (1980). Milroy’s important research in sociolinguistics

 

” The sub-category listed above this one, favorite food, shows a significance level of (0.098). It is
tempting, because these results are also not “too far” from the accepted level, to wrap this variable into the
discussion of reasons for significance; however, because these results are twice the distance from the
accepted level than those for the number of E/O fiiends variable, this may be too much of a stretch, and I
will not include this variable in my discussion.

96

has established the notion that analysis of social networks can be used very successfully
to study individual linguistic behavior in communities. This work itself draws on earlier
research by Boissevain (1974), which is based on the idea that in their social lives,
individuals behave meaningfully not only as members of groups such as sex, class,
particular occupations, etc., but as individuals. Importantly, individuals act to create and
maintain key social relationships, in what Boissevain characterizes as exchanges. Milroy
takes these ideas further, claiming that the social network

becomes a mechanism whereby pressures, resulting from

obligations contracted within the network, are applied to influence

an individual’s behaviour. If the individual wishes to protect

social relationships, these constant obligations must be honoured

(1980,p.49)

Milroy states that when considering the influence of a social network on behavior,
the structural and content characteristics of these networks become important, the most
important structural characteristic being density. A dense network is one in which the
people an individual associates with, or has links to, also have links to one another; in
contrast, a person with a network of low-density would have contacts, none of whom
knew each other. Another important dimension of a network is its interactional or
content characteristics, which, according to Milroy can be either multiplex or uniplex. In
a network, people who are connected in a single capacity only are said to have
relationships that are uniplex; a person who is linked to another in varying capacities—
for example, as a classmate, a neighbor, a friend, or a family member, etc.—is said to
have multiplex network ties. According to Milroy, “multiplexity and density are

conditions which often co-occur, and both increase the effectiveness of the network as a

norm-enforcement mechanism” (p. 52). As noted earlier, there is much overlap between

97

Milroy’s characterizations of social networks and Le Page & Tabouret-Keller’s sense of
groups as being either focused (closely interactive and tightly knit) or diffuse (loosely-
knit, entailing more linguistic variation).

In this study, recasting the E0 variable in terms of network can only at this point
be metaphorical, as a description of social relationships, because detailed analysis of the
participants’ networks was not done. However, given the demographic distribution of the
participants in this study, we can assume that most of them have dense and relatively
multiplex social networks: they live in the same neighborhoods, have extensive family
ties in the neighborhood, have the same socioeconomic status (SES), attend the same
schools, have the same home language, etc., and therefore the linguistic norm of their
community, Spanish, is being reinforced. Those with E0 friends, by definition, have
broader relationships that are less dense, some of which may be uniplex; this factor could
be implicated in their higher English proficiency scores.

Considering EO as a measure of social networks in this way, in a sense
reclassifies it as a large group identity factor, similar to several other factors used in this
study—age, sex, and SES (see Table 51 above). Like one of these, Age, EO appears to
be a better predictor for English proficiency than the other sub-categories of the identity
variable. In this way, it is also like the migrant status variable (also listed in Table 51
above), which is another large group identity factor that appears to be a good predictor of
English proficiency. One might at this point think, then, about removing the E0 sub-
category from the identity variable, because it better fits the other grouping. However,
this move would not be as straightforward as it would seem. EO also differs from age,

sex, SES, and migrant status in an important way. While the other factors do not involve

98

personal choice, EO very much does. And in this way, E0 is like the other eight sub-
categories of the identity variable that it was originally grouped with; all of these factors
reflect individual patterns of choice.
Large Group Identity Factors vs Factors Entailing Individual Choice

In a certain sense, then, E0 is a hybrid category, spanning two domains—that
occupied by the other large group identity factors of age, sex, SES, and migrant status,
and that of the variables entailing personal choice. This second factor has already been
discussed. As noted, Le Page and Tabouret-Keller (1985) found motivation to identify
with groups (which involves personal choice) to be quite predictive of linguistic
behavior. That is, it is not just the opportunity to identify with groups, but the desire to
do so, that is crucial to predicting linguistic behavior. Given this, one might be tempted
to argue that the assembly of social facts operationalized in this study as identity, because
it includes only variables that entail personal choice, should have been a strong predictor
of English proficiency. However, motivation/personal choice can be a double-edged
sword. It can exert its pressure either positively or negatively. That is, it can drive
movement toward a group norm, or away from it (c.f. Milroy, 1980; Milroy & Milroy,
1992; Trudgill, 1972; 1974). In addition, while this factor has been shown to predict
linguistic behavior, it is also, by nature, highly variable. Recall the wide distribution of
identity scores by grade level in this study (see Table 49). In fact, given most of the
evidence here, it would appear that the lower level ethnic identity factors used in this
research are not very revealing at all. Thus, it may be that the E0 variable is significant

primarily because it is a large group identity factor.

99

Ethnic Identity: Can We Assess This?

In considering ethnic identity, one might continue to move away from the lower
level factors in the discussion. We’ve seen that ethnic identity can be thought of as a
construct of various features and/or categories. Yet it can also be thought of as a super-
category itself. In that sense, this study could have proceeded without an assessment of
ethnic identity at all, but rather with the assumption that the ethnic identity of all of the
participants was the same—that of California Mexican-American. I could have
categorized them as such, and focused on other aspects of the bilingual lives of these
children, and perhaps come away with a less murky picture. However, I would have been
no closer to understanding the relationship between public identity categories, such as
California Mexican-American, and the lower level ethnic identity factors that inevitably
entail personal choice, and which presumably are a truer reflection of identity. Moreover,
as Fought (2002) remarks, assigning speakers to a category like ethnicity cannot take the
place of careful study of the social practices that such a category is made up of, claiming
that such classification “is useless to us without an understanding of the construction of
ethnicity by individuals and communities” (p. 444). This leaves us in a tough position,
because while ethnic identity is extremely difficult to study, in order to gain any insights
at all, such study must not ignore the lower level identity factors. An example of this fact
was highlighted earlier (see Results and Discussion section, sub-section A closer look at 3
of the 9 identity sub-categories) in this study. As noted, my assumption that the
participants’ responses regarding their preferences for soccer team would fall into the
super-category of nation did not come close to capturing the nuanced cultural information

actually represented in their real team preferences. These participants had strong

100

allegiances (as shown in their responses) to particular regional teams in Mexico—teams
that they claimed their families supported also.

At the outset of this study, I expressed hope that some insight into the relationship
between the English proficiency and ethnic identity of these participants would be gained
through this study. While neither clear answers nor expected results were obtained on
this issue, I did obtain results, albeit perhaps obvious and to be expected, with regard to
the categories of age and migrant status, and how they relate to English proficiency for
the participants in this study; this, in addition to confirming expectations, also provides a
point of departure for future study. Additionally, it appears that as a measure of English
proficiency, the one constructed here may be worth using again, or parts of it. However,
because there was a difference in the range of scores obtained for the three sub-categories
of this measure, before using it again as is, I would want to investigate the possible
reasons for the high scores on the article usage. The variable regarding how many
English-only speaking associations the participants had was telling, especially when
conceived of as a type of network score; these data hint at perhaps a fruitful avenue for
future research. It could be very worthwhile to study these participants (or ones like
them, that is, from the same community/neighborhoods/school) again, with a focus on
social networks.

In spite of the enormous complexity involved in the study of the formation of
ethnic identity, or perhaps because of it, I am left still intrigued about the social lives of
these participants. Ethnic identity is not, as Zentella (1997) notes, automatic, nor is it
simply a compilation of group customs, practices, cultural artifacts or linguistic features;

social identities are instead, actively constructed, in particular communities. These

101

California Mexican-Americans, migrants most, children all, have a particular story to tell.
It is clear that their experience, which includes their language experience, is “a complex
phenomenon that defies a characterization based on simple comparisons” (Vasquez, O.,
et al., 1994). While Watsonville itself contains a generational, stratified community—
including recent arrivals and established immigrants from México, alongside other ethnic
groups (see Table 2 above)—the participants in this study represent a slice of this
community, not one that cuts across categories, but one that perhaps contains a type. This
makes them not a random sampling of the community at large, but a snapshot of a
particular moment in Watsonville history. These children, both at the individual level
and at the group, are second generation Americans with an active, living memory of the
land of their parents: Mexico. Studding the experience of assimilation/integration for
most of these participants is the experience of being migrants who make annual treks to
and from México. But, even more poignantly for the lives of these children, because of
the generational and circular migration into and out of this community, México has bled
into Watsonville, creating a dynamic exchange of cultures, and languages. This exchange
encompasses not only these participants’ own local social networks, but the world
outside—a world that they are exposed to through the media, their schools, etc. In this
way, these children become “increasingly fluent in an Anglo cultural Milieu at the same
time that they become more fluent in a local Mexican culture” (Rosaldo, 1985, p. 12
(italics mine). This sense of the world(s) that these participants live in is congruent with
recent research that centers on transnational identities (Goldin, 1999). This work moves
away from characterizing the processes of identity formation across borders through

simple distinctions—such as geographical location, or national designation, such as

102

Mexican and American—and toward more “unbounded concepts” (p. 2). Goldin’s sense
is that the processes of identity formation are fluid in nature, but yet are designated within
the broader concepts of homogenization and differentiation; it is then, through these
inter-connected processes that “groups and individuals draw the lines of commonality
and differences” (1999, p. 5).

And, these lines are drawn in particular places. In addition to the quantitative data
collected during this study, I also wondered about the impressions the children had of
their communities in Watsonville and in Mexico. Did they distinguish between the two
places, and if so, in what ways? During the casual conversations I had with them, in both
Spanish and English, I embedded two questions, 1) What is the same about Mexico and
Watsonville, and 2) What is different about Mexico and Watsonville? Attempting to get
uncluttered responses, I asked only these very open-ended questions, and waited for their
responses, which I did not guide in any way. I asked first what was the same, and then
moved to the question about differences. All of the respondents answered both of these
questions easily and quickly, except for the one who was born in the community, near
Watsonville, and who claimed he had never been to Mexico.

What follows is a compilation of the responses that these participants gave to
these two questions. Qualitative in nature, these data are included here as a supplement
to the larger discussion of this project. The impressions that these children have of the
two places they inhabit are quite telling, and do in fact reflect a blurring of borders, an
overlapping of worlds; their impressions describe not so much distinct communities, but

“Porous, web-like ethnoscapes” (Goldin, 1999, P- 3)-

103

Iii:

llér
chur
ems

dint
SUE“
1116111
no.1
Nhne
hdsl
phce
hide

efiher

hhgs
“hie
“EH50
C00kt.

andlh

WOple

mllpla

Impression_s: People. Food. and Vampires

The participants, across all three grade levels, reported much similarity between
México and Watsonville along physical lines. They claimed that the houses, the
churches, the stores, the plants, the trees, the ranchos, the mountains, the parks, and the
cars are the same. One fifth grader provided some detail with regard to the ranchos,
claiming that in Mexico the ranchos had mainly cows, and in Watsonville, mainly
strawberries. Another thing he mentioned, in the next breath (perhaps triggered by his
mention of the strawberry fields), was that in Watsonville they work a lot, but in Mexico,
no. The participants also claimed that the games they play in both places are the same——
Nintendo and soccer were named. Several of them did mention, however, that in México,
kids were more “free” (to play unattended; to explore at will). Many stated that the two
places had the same animals—dogs, cats, cows; an exception to this was two of the first
graders, who reported that México has bulls and burros, and Watsonville doesn’t have
either.

Food was also reported to be the same in both places by several of the children,
who specifically named enchiladas, beans, and rice. One of the third graders said that
while the food is the same, the cheese and the tortillas in México are much better than in
Watsonville. One difference noted by one of the fifih graders was that in México food is
cooked outside, but in Watsonville, people liked to eat inside, at places like McDonald’s
and Burger King.

The children’s responses got particularly interesting when they spoke of the
people and language(s) in both places. Not one said that the people were different in the

two places, and several of them claimed this as one of the similarities about México and

104

Watsonville. In fact, one third grader reported that “las personas son iguales,” (“the
people are exactly the same”) in both places. While none of the first graders mentioned
language, quite a few of the third and fifth grade participants reported this as either a
similarity or a difference, with varying responses; several spoke in terms of “more”
Spanish in Mexico than in Watsonville, and another mentioned, referring to Watsonville,
that “they speak Spanish here, too.” One of the fifth graders claimed that “a lot talk in
Spanish in Mexico, and in Watsonville, some of them talk in English and some in
Spanish.” Another fiflh grader reported that “more people speak Spanish in Michoacan,
and more people Speak English here.” Another mentioned school with regard to language,
claiming that in México only Spanish is used in the schools, but here both English and
Spanish are.

At a more abstract level, more differences emerged, interestingly, with the older
students—the third and fifth graders—and not the younger ones. Two of the third
graders mentioned that the flag and the country for each place were different, and one
fifth grader mentioned that money was different in each place. Another fifth grader
mentioned the pledge of allegiance as being different in both places, saying that he didn’t
know México’s, but his Dad did; he was able, however, to tell me that “it’s about going
to war and fighting people who want to take your land.” The fact that these differences
were mentioned by the older students most likely reflects an emergent awareness of
difference at a more abstracted level of nationhood, but overall these participants’
strongest impressions seem to be that there is a lot of overlap between the two places,

especially on issues that most directly affect their lives.

105

While my descriptions above make mention of the place names, most often in
their responses the children made reference to the two places through the use of the terms
“here” and “there,” (acti and alld) instead of to “México” and “Watsonville.” While this
might be a stretch, this could reflect a sense of the two places as being quite proximate,
and not as being separated by thousands of miles and an international border. Given their
impressions—that these are two places that in general look very much alike, are inhabited
by the same people, who eat the same kinds of foods, play the same kinds of games, and
where their native language, Spanish, is spoken by many, if not all—perhaps for these
children the more intimate terms of acci and allci best capture their sense of the span of
these inter-connected communities; perhaps reflected in their impressions is a sense that
the edges of their transnational worlds are fuzzy, so that that which runs between them is
in many ways a “borderless border” (Bonilla, et. al., 1998). For these young people, the
lines of commonality and difference seem to be measured differently; several remarked
that the main difference between the places was that acd everything is covered in cement,
and allci there is much more dirt—on the roads, and in the fields; another young girl said
that the only difference was how the two places smelled. One first grade boy’s
impression was most revealing—perhaps of a secret truth. This one, when asked about
what was different between the two places, looked at me rather scomfully, as if I had just
asked him a question that anyone should know the answer to. He exclaimed, “pues, allci

hay vampiros, y acct; no!” (“well, over there, there are vampires, and here, 1101”).

Return to Main Street

In closing, I’d like to return focus to the community in which these participants

have drawn many of their lines of commonality and differences; what follows is a series

106

of descriptions and still-shots of parts of Watsonville. As promised earlier—in the
section of this research that gave a historical overview of Watsonville—we revisit Main
Street, and the center of downtown Watsonville for an updated View. For convenience,
I’ve copied again below the passage from the section above on Watsonville history,
which contains a description of the city in the 1870’s:

Pajaro Street became known as Main Street as shops and

businesses moved in. By the mid-seventies, there were ten grocery

stores, six doctors, a dentist, two livery stables, three

harnessmakers, four blacksmiths, three wagonmakers, three meat

markets, seven hotels, eight saloons, two telegraph offices, one

bank, the Bank of Watsonville, one foundry, one newspaper and

one bakery in town. The young ladies’ group called the Butterfly

Club raised money for a fountain to be built in the middle of the

Plaza (Santa Cruz Library Weekly, 2003, para. 7).

Today, the center of town signals a strong Mexican influence, which mixes with
the also apparent Anglo presence. The Plaza is still a focal point, but the fountain
designated by the Butterfly Club is not at its center. Instead, there is a gazebo, ringed by
walk-ways lined with benches. The messages in this public space, marked on trash-cans
and recycling bins, for example, are all bilingual—Spanish and English. Below is a

picture of this Plaza, which I took on a weekend morning; this picture shows the gazebo,

a few folks on a bench, and the bilingual trash receptacles.

107

llllllllllllli or
hit:
. PROIIIIIIID

 

Figure 4-b: Trash Bin—Spanish & Figure 4-c: Trash Bin—English

108

As I moved away fiom the Plaza and onto Main Street, I noticed another bilingual
public announcement from the City. In its campaign to keep the streets safe and drivers
within the speed limit, the City hung bilingual banners encouraging people to “Share the
road, with caution, considerately, and with respect.” Below is a shot of one side of this

banner—the Spanish side.

 

Figure 5: Bilingual Watsonville City Banner

Continuing down Main Street with my camera, I ran into one of the locals, who,
seeing that I was taking pictures of the area, boldly asked to have his picture taken (in
Spanish). Posing there, in his Stetson and leather boots, this man seemed to me to be the
quintessential vaquero. As I was taking his picture, a small boy came out of one of the
stores along the street, and smiled at me until I took his picture as well. In preparation,
he climbed up onto the back of a pick-up truck parked just there and posed for the shot.

Below are these pictures.

109

 

Figures 6-a & b: E1 Vaquero, and Child on Pick-up Truck

110

Across the street I noticed two adjacent storefronts, each with signs that were not
bilingual, but which were in English and Spanish, respectively. These caught my eye;
side-by-side, each reflected official business. One was the Social Security
Administration office, and the other was a fairly fancy jewelry store. Each, with its
business being declared through a sign that was written in only one of the two main
languages of the community seemed to send a particular message; however, taken
together, as two buildings close enough to have a common wall, on a street among other
building that carried bilingual signs, that message of one language-only appeared to be

over-ruled.

 

Figure 7: Social Security Administration Office and Joyeria

111

I noticed two other business signs along Main Street that seemed to convey this
same spirit—of marking one’s own message in a particular language, but within a domain

that recognizes the other. The first of these two signs is below:

£11205 . posors

‘ a BURRITOS o MILANESA’ a,“
a sopas oMARlseostI

filly?” ‘ll. ISUL"

 

Figure 8: Super Grullense Taqueria Atotonilco

This sign, for SUPER GRULLENSE TAQUERIA A TOTONILCO, was on the window of a
taqueria (taco stand), and is very interesting in that the detailed cultural content conveyed
is in only one language. That is, the name of the place, and the list of the various foods
served in this establishment were all in Spanish, and not translated into English. One of

the words in the name, Grullense, means person from El Grullo, which is a town in

112

J alisco, Mexico. Another word, Atotonilco, a Nahautl (indigenous language of Mexico)
name, is also a place in Jalisco. Jalisco, as noted above, is one of the states that compose
the most important migrant-sending region in Mexico, and many of the people in this
community come from there. None of this detail, nor that of the different foods listed, is
captured in the catch-all phrase at the bottom of the sign, which is the English term,
Mexican Food. What I find interesting here is that while it is obvious that a particular
population is being targeted by the advertisement for this restaurant in Spanish, there is
also an acknowledgement (a tipping of the hat, if you will) of another segment of the
community—the English Speakers.

The second picture, below, demonstrates the same phenomenon, but in reverse.
In it, we see cultural content that is marked in English only, but again, a translation of
sorts is found at the bottom of this sign—this time, as a single Spanish term, carniceria,
which means butcher shop; interestingly, this term in Spanish does not necessarily imply
custom cutting and smoking of meat. The English message, like the Spanish one in the
picture above this one, seems to be targeting a particular population—perhaps the
ranchers or hunters in the area; however in this sign there is also a nod toward another
segment of the population—people who, whether or not they are interested in custom-

smoked meat, are speakers of Spanish.

113

 

Figure 9: Stevie G’s Meats

In addition to the goal of walking down Main Street to document evidence of the
bilingual/bicultural world the participants in this study lived in, I also had another
objective. At one end of this street there was a taquerl'a that I had come to know years
before, when I worked in the community. I wondered if it was still there, and if the tacos
they served were still just like the ones I had eaten so many times in México. I made my
way toward where I remembered this taqueria to be, and was happy to find it there,
bustling with business. In addition to the various tacos served there, because it was a
weekend, both pozole and menudo were on the menu. These stews, made with hominy
and either pork (pozole) or tripe (menudo), which are very commonly eaten on the

weekends in México, are famous for their ability to cure a hangover. As I lined up to

114

place my order for tacos de carnitas (roast pork), I noticed a woman in front of me, with
her pot to be filled with either pozole or menudo, to take home with her. I also noticed
that I was the only Anglo around. All around me Spanish was being spoken, and
Mexican music was playing; this, mixed with the smells of the meats being grilled for the
tacos and of the pozole simmering away, made me momentarily forget where I was.

I attempted to capture that moment in the picture below, but wasn’t very

successful:

E“ i O.“ i:
‘» i“). 2 n3 ’ A”
' ‘ “”2 ""53 .odlllns :1
. mTu-Iob 6:” ”an!“ “
l ’ n .V
' “it. r233 ,é—ttu 2‘ uh: ,‘ .
' - 1.. var ,
J. on. '1'. .
. do 0"“: dc
C

0:“ uchiladIzos d: '

. bQue b ‘L I Wheosigiuimifi
‘ V S'

 

 

Figure 10: Taqueria on Main Street

Still, in this picture, we see more evidence of the dynamic cultural and linguistic mix that
is Watsonville: the weekend advertisement, in Spanish, for pozole or menudo and the
Various other foods offered; the breakfast burritos, served Monday through Friday

mornings (apparently targeting those rushing off to work), advertised in English. And off

115

to the lefi of the menu, taped to a comer of the plate-glass window, a tell-tale sign that
provides a back-drop for this particular context—one displaying the American flag and

the words, “united we stand.”

Concludingrempflgs. on the generalizabilig of the findings

Because of the seemingly idiosyncratic nature of identity as operationalized here,
it is possible that this study falls under the category of “worst case” on the continuum of
generalizability reviewed on the introductory pages of this research (Table 1, page 4).
That is, given that the identity variable was composed, for the most part, of lower level
identity factors that are ultimately reflections of individual choice, and therefore by
nature, idiosyncratic, perhaps the results from this study are not generalizable. However,
it is possible to see a study such as this one as a baseline of sorts for research on the
relationship between English proficiency and identity, in this community; subsequent
research could be implemented replicating this work, either very closely, or in a model
that would hold varying factors constant. Another framework for subsequent research in
this community would be to run a longitudinal study, following these participants over
several years, and tracking potential shifts in their English proficiency, and their
identities. The results of this kind of work could provide insights into trends and patterns
that are generalizable within this community. If such work were done, that would bring
this study along on the continuum of generalizability (Table 1, p. 4), and into the second
column, that referring to a specific community.

There is work that has been done in this community that is relevant to this
discussion. Matute-Bianchi (1986) studied high school students in Watsonville, and the

interrelatedness of school performance, minority status, and ethnic identity. While

116

Matute-Bianchi’s research focused on different issues than this study did, and is therefore
not completely applicable here, certain insights from her study are, because they range
over participants from the same community as the ones studied in this research, but who
are 5-10 years older. Matute-Bianchi claims that there are five categories that most of the
Mexican-descent students in the high school can be placed into. These are:

0 Recent-Mexican immigrant

o Mexican-oriented

0 Mexican American

0 Chicano

o Cholo
Matute-Bianchi assigned the students she interviewed to these categories, and asked them
about the categories, and to identify themselves, according to them. She found that the
“recent-Mexican immigrants,” who were predominantly Spanish speaking, claimed an
identity of Mexicano, and think of México as their home; further, the majority of these
respondents claimed that it was economic opportunity that brought them to the US.
Those students categorized as “Mexican-oriented” were most often bilingual; these
students also claimed an identity of Mexicano, but had strong ties to both the US. and
Mexico. In addition, they saw themselves as distinct from the other categories. The third
category, “Mexican-American” students, were born in the US, and according to Matute-
Bianchi, were much more American-oriented than the first two groups. These
respondents claimed differences between themselves and the people in the other groups;
interestingly, Matute-Bianchi notes that these students find the terms “Chicano” and

“Cholo” offensive. The group, the “Chicanos,” made up 40-50% of the Spanish—sumame

117

students; these identified themselves as Mexican or Mexicano, and did not find the term
“Chicano” offensive. Finally, the “Cholos,” who, according to Matute-Bianchi, make up
a very small and disappearing group, are identified by others as gang-oriented or gang-
sympathetic. Interestingly, in her study, Matute-Bianchi does not report how those
categorized as “Cholos” identify themselves.

I’ve described the categories from Matute-Bianchi’s study in some detail because
this description highlights issues that are relevant to this study; while it is acknowledged
that it is difficult to compare the two studies because they assessed variables that were
different, and were operationalized differently, they nonetheless each drew participants
from the same community, and both were concerned with issues of identity. The
categories of identity, both those assigned and those claimed by the respondents in
Matute-Bianchi’s study, are of interest. In their range, these potentially corroborate the
discussion above on the variability and porousness of identity; but more importantly,
these may be indicative of possible paths of development that the participants in this
study could follow. A longitudinal series of studies that could track this, and the
development of English proficiency, could be of some interest, and potentially could
result in findings that are generalizable within this community.

Beyond the community of Watsonville, there is research that has been done in the
same agricultural region of California; Vasquez etal., (1994), drew its data from a
community that is also predominantly “Mexicano,” and describes a commmrity that is
very similar to Watsonville. As a study that is concerned with the relationships between
language, culture, and learning, Vasquez et a1. (1994) overlaps a bit with this study;

however, the orientation, which includes pedagogical implications of language learning,

118

is different. Nonetheless, this community could serve as a point of reference for future
research done in Watsonville.

Clearly, the potential generalizability of the results from this study depends on
follow-up work. However, there are frameworks, themselves generalized from other
research, that can be implicated in the results found here, and as such, can add to their
explanation. Ethnolinguistic identity theory (Giles & Johnson, 1987) claims that people
in a subordinate group will hold on to their ethnolinguistic identity, and will most often
not become fully proficient in the dominant language of their community, if the following
conditions exist:

0 They identify themselves strongly as members of a group, and language is an
important symbol for identity

0 They are aware of alternatives to their social status

0 They perceive the ethnolinguistic Vitality of their group to be high

0 They perceive the boundaries to their ethnic identity and language to be closed

0 They don’t identify with many other social groups
Of these conditions, ethnolinguistic vitality is particularly pertinent to this discussion; this
is a measure that is itself based on several components, including 1) demographic
strength (group numbers, birthrates, migration, concentration and distribution in the
community), 2) institutional support for the language (in the education system, the media,
the government), and 3) status (social, political, economic, linguistic prestige).

Given the demographic data presented in this study, as well as the descriptions of

the educational support programs available within the school district (bilingual and
migrant programs), combined with the results of this research, it would appear that the

ethnolinguistic vitality of these participants is high. Because of this, these participants, in

119

this community, appear to be holding fast to a Mexican-oriented identity, while becoming
more proficient in English. The question of how this vitality affects whether or not these
children become fully proficient in English (a claim entailed in the Ethnolinguistic

identity theory, noted above), is a matter for future research.

120

APPENDICES

121

Appendix A: Watsonville Occupation Characteristics

(From 2000 Census)

 

IDP 3 PrOfile OfSelected Economic CharaCtenStlcS 2000..“

Data Set:C ensus 2000 Summgy File 3 (SF 3) — Sample Data
Geographic Area: Watsonville city, California

 

lSubject _ lNumber _ _. Percent

Employed c1v111anvpopulatlon 16 years and
over

[OCCUPATION

Management professronal and related
occupations

 

17, 285%

l
l
l
l
l
l
l

 

2 899.

 

lSerVice occupations

 

 

Sales and office occupations e 3: 43§I

lF arming, fishing, and forestry occupation’smm

 

occupations

 

Production, transportation, and material-W
moving occupations

Mm... ._.._ -W-wufi.-v--__......_.,.....-n-.--.__....-..-__._..-.. ......... . -,_.._-._,.,._..___._.....___.-._.--__.. _..,.... .
r

lINDUsTRY

1
Construction, extraction, and maintenance I
l
l
l

 

 

 

mining

19.9;
17.0“;

Agriculture, forestry, fishing and hunting, and I , ‘

100.0

16.83
17.9

9.9%

18.6

 

lConstruction l l ,1 93 l

m- _.._.._ ”—7—..- Wm, ....._.---__......................---,-_,_._._ m.-.._...2_..a..~.._---_ _-.........___.._.n».n.ar..-_..--.e—-

 

[Manufacturing I 2,364fll .. .

lWholesaletrade A. I I. H l l H 1,082

IRéfhjlfi-ade l 1,952. .. W ., ,

 

 

lTransportation and warehousing, and utilities I 533l
llnformatlon l

 

Finance, insurance, real estate, and rental and
leasing '

r—-.-.._.._‘..- -Ww....__._ . SW.

ProfeSSional, seientific, management, .
administrative, and waste management services i

 

 

1,310?E

 

 

Arts, entertainment, recreation, accommodation;

and food services R3901;

 

I
l

lOther services (except public aMImstratron) l ‘ ”"738 l

 

 

122

7.6

lEducational, health and social services ml 2 5,703.?“ 14.9

8.0?

IPublic administration _ --. ' " ' 440* ~~ ~ ~---

2.5;

Appendix B: Income Eligibility Guidelines for Free Lunch

California Department of Education April 2000
Nutrition Services Division

INCOME ELIGIBILITY GUIDELINES
FOR FREE AND REDUCED PRICE MEALS OR FREE MILK
(Effective from July 1. 2000 through June 30. 2001)

Children from households With incomes at or below the following levels
may be eligible for free or reduced price meals.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FREE ELIGIBILITY SCALE REDUCED PRICE ELIGIBILITY
SCALE
Lunch. Breakfaét. Milk Lunch. Breakfast
HOUSEHOLD
SIZE YEAR MONTH WEEK YEAR MONTH WEEK
1 $10,855 $ 905 $209 $15,448 $1.288 $ 298
2 $14.625 $1.219 $282 $20,813 $1.735 S 401
3 $18.395 $1.533 $354 $26,178 $2.182 $ 504
4 $22,165 $1.848 $427 $31,543 $2.629 s 607
5 525.935 $2.162 $499 $36,908 $3,076 $ 710
6 529.705 $2.476 $572 $42,273 $3,523 5 813
7 $33,475 $2,790 $644 $47,638 $3.970 $ 917
8 $37,245 $3.104 $717 553.003 $4.417 $1.020
For each additional family member. add:
+$3.770 +3 315 +$ 73 I +$ 5.365 +5 448 +5104

 

 

 

 

 

 

 

 

° BOTH of the above scales appear in the Sample Media Release

' ONLY the Reduced Price Eligibility Scale may appear in the Lettetr to Households (for those
schools participating in the National School Lunch Program).

 

' ONLY those schools that participate in the Special Milk Program. and offer free milk should
publish the free eligibility scale in both the Sample Media Release and the Letter to Households

123

Appendix C: Noun/Noun Compounding

Noun/Noun Compounds: 16 Items

For each of the 16 NfN Compound items, the elicitation technique used was
“Picture Task,” in which a different picture was shown to each subject, and the

question, “What is this?” was asked.

Copies of the pictures used for this task are included in this Appendix B,
following this page. The pictures are labeled Appendix B] through Appendix
B.16. The order of these pictures follows the order found in the table below.

 

Target Noun/Noun Compound

 

Muscle man

 

Fire truck

 

Flashlight

 

Houseboat

 

Key-chain

 

Kite-tail

 

Doghouse

 

Paper-bag

 

Raincoat

 

Anteater

 

Mousetrap

 

Handcuffs

 

Banana peel

 

Chocolate Rabbit

 

Bird cage

 

 

 

Remote control

 

124

Appendix C]: Muscle man

 

125

Appendix C.2: Fire truck

 

126

Appendix C.3: Flashlight

 

127

Appendix C.4: Houseboat

 

 

128

Appendix 05: Key-chain

 

129

Appendix C.6: Kite-tail

 

130

 

 

Appendix C.7: Doghouse

 

131

Appendix 08: Paper-bag

 

132

Appendix C.9: Raincoat

 

 

133

 

Appendix C.10: Anteater

 

134

$$$¥

 

Appendix Cl 1: Mousetrap

 

135

 

Appendix C. 12: Handcuffs

 

136

Appendix C. 13: Banana peel

 

137

Appendix C.14: Chocolate Rabbit

 

138

Appendix C. 15: Bird cage

 

1..

 

.n.

 

 

 

 

 

1:: ' m TN" £1 g.' I. ? 1:11;”... H‘ _ ‘

 

 

 

 

 

139

 

Appendix C. 1 6: Remote control

 

140

Appendix D: Article Uglge

A) Mass nouns: 9 Items
0 Spanish requires the definite article; English disallows articles

 

 

 

 

 

 

 

 

 

 

 

 

 

Target Mass Elicitation Sentence Elicitation Technique
Noun
Meat my dog eats (the) meat Repetition
Music Mary loves (the) music Repetition
salt my brother cooks with (the) salt Repetition
Sand the beach has lots of (the) sand Repetition
People (the) people are funny Repetition
Air the balloon is filled with (the) air Repetition
Snow (the) snow is cold Repetition
Honey I like (the) honey in my tea Repetition
Bread (the) bread is made with flour Repetition

 

B) Specificity regarding mass nouns referring to ALL of something: 5 Items
0 Spanish requires the definite article; English disallows articles

 

 

 

 

 

 

 

 

 

Tapget Noun Elicitation Sentence Elicitation Technique
children (the) children are friendly Repetition
rice I don’t like (the) rice Repetition
dogs (the) dogs bark loudly Repetition
coca-cola (the) coca-cola is delicious Repetition
ice crean I like (the) ice cream Repetition

 

C) Specificity regarding personal titles: 6 Items
0 Spanish requires the definite article (double determiner); English disallows

 

 

 

 

 

 

 

 

 

articles
Tam title Elicitation Sentence Elicitation Technique
Mr. Martinez (the) Mr. Martinez is here Repetition
Mr. Brown (the) Mr. Brown is old Repetition
Father Sanchez (the) Father Sanchez went home Repetition
Mr. Brown This is for (the) Mr. Brown Repetition
Ms. Evans (the) Ms. Evans came to the party Repetition
Father Sanchez That is for (the) Father Sanchez Repetition

 

 

Please note: The Elicitation Technique used for items D) and E) below was “Picture
Task.” Copies of the pictures used for these tasks are included in this Appendix D,
following this page. The pictures for items in D) below are labeled Appendix D]

141

 

 

 

through Appendix D.9. The pictures for items in E) below are labeled Appendix D.lO
through Appendix D. 1 8. The order of these pictures follows the order found in the tables
in D) and E), below.

D) Specificity regarding nouns showing mere identification: 9 Items
0 Spanish disallows articles”; English requires an article

 

 

 

 

 

 

 

 

 

 

Targt Noun Elicitation Sentence Elicitation Technique
Doctor He is (a) doctor Picture Task
teacher She is (a) teacher Picture Task

baseball player He is (a) a baseball player Picture Task
Nurse She is (a) nurse Picture Task
Writer He is (a) writer Picture Task
waitress She is (a) waitress Picture Task
Musician He is (a) guitar player, etc. Picture Task
Musician He is (a) musician, etc. Picture Task
astronaut He is (an) astronaut Picture Task

 

 

 

 

 

*(Note: Regarding this type of construction, as noted in #23 in the main document,
repeated here Spanish disallows the article:

23) Es médico *Es un médico
He is doctor (0)

However, if this same sentence were meant to indicate individualization, or emphasis, the
article would be used: Es un médico=He is a doctor. Also, in situations of additional
modification (Es un buen médico/He is a good doctor), or in figurative speech (Es un
Iobo/He is a wolf), an article would be used [Stockwell etal., 1965]).

 

E) Inaliable possession: 9 Items

0 Spanish, with possession obvious, requires the definite article; English requires a
possessive pronoun

 

 

 

 

 

 

 

 

 

 

 

Elicitation Sentence Elicitation Technique
He hurt (his) head/tooth Picture Task
He hurt (his) whole body Picture Task
He is raising (his) hand Picture Task
She hurt (her) feet Picture Task
She broke (her) arm Picture Task
He has dirt on (his) feet Picture Task
He took off (his) shoes Picture Task
He hurt (his) neck Picture Task
She is combing/brushing (her) hair Picture Task

 

 

142

 

Appendix D.1: Doctor

 

143

Appendix D.2: Teacher

 

144

Appendix D.3: Baseball Player

 

145

Appendix D.4: Nurse

 

146

 

Appendix D.5: Writer

 

147

Appendix D.6: Waitress

 

 

148

Appendix D.7: Guitar Player

 

149

Appendix D.8: Musician

 

150

Astronaut

Appendix D.9

 

151

Appendix D.10: He hurt (his) head/tooth

 

152

Appendix D.12: He is raising (his) hand

 

154

Appendix D. 1 3: She hurt (her) feet

 

155

Appendix D.l4: She broke (her) arm

 

156

Appendix D.15: He has dirt on (his) feet

 

157

Appendix D. 16: He took off (his) shoes

 

158

Appendix D. 1 7: He hurt (his) neck

 

159

Appendix D.l8: She is combing/brushing (her) hair

 

160

Apmndix E: Tense Doubling

o Tense Doubling Structures: 12 Items

Listed below, in A) and B), are the two elicitation methods used—one for the older
subjects, and one for the younger—as well as the actual structures used.

A) Elicitation method used with the older subjects—with those from the 4“'/5tll
combination class, and the 3rd grade class:
0 Context given by researcher in English (items listed in a. below)

0 Subjects ask question in English, based on the context given; researcher prompts
subjects with the first word in the targeted question, e. g., what, where, how, who
(target structures listed in items b. below)

1. a. John went to the store last night and wanted to buy something, so he did
b. What did John/Juan buy last night?

2. a. The boy with the yellow hat on wanted to eat something last night, so he did
b. What did the boy in the yellow hat eat last night?
3. a. The girl with the brown hair wanted to drink something yesterday, so she did

b. What did the girl with the brown hair drink yesterday?

4. a. The boy in the yellow hat went to McDonald’s last night and wanted to buy
something, so he did

b. What did the boy in the yellow hat buy last night?
5. a. Juan wanted to eat something last night, so he did
b. What did John eat last night?
6. a. Maria wanted to drink something yesterday so she did
b. What did Maria drink yesterday?
7. a. The boy in the yellow hat wanted to go somewhere yesterday, so he did
b. Where did the boy in the yellow hat go yesterday?

8. a. The girl with the brown hair saw her boyfriend last night and wanted to dance
perfectly for him, so she did

b. How did the girl with brown hair dance last night?
9. a. John wanted to play some game last night, so he did
b. What game did John play last night?

161

10. a. The woman with the red dress on wanted to play some game last night, so she did
b. What game did the woman with the red dress on play last night?

11. a. The man wanted three tacos last night
b. What did the man want last night?

12. a. The very old man wanted three apples last night
b. What did the very old man want last night?

B) Elicitation method used with the younger subjects—with those from the 1" grade
class
0 Context given by researcher in English (items listed in a. below)
0 Researcher asks the target response question in Spanish (items listed in b. below)

0 Subjects translate this question into English, based on the context given, and the

question in Spanish; if necessary, researcher prompts subjects with the first word
in the targeted question, e.g., what, where, how (target structures listed in items c.
below)

1. a. John went to the store last night and wanted to buy something, so he did.

b. aQué compro Juan anoche?

c. What did John/Juan buy last night?

2. a. The boy with the yellow hat on wanted to eat something last night, so he did

b. aQué comio e1 niflo de la gorra amarilla anoche?

. What did the boy in the yellow hat eat last night?

0

. The girl with the brown hair wanted to drink something yesterday, so she did

0"!»

. aQué tomo la nifia del pelo café ayer?
c. What did the girl with the brown hair drink yesterday?

4. a. The boy in the yellow hat went to McDonald’s last night and wanted to buy
something, so he did

b. aQué compro e1 nifio de la gorra amarilla anoche?
c. What did the boy in the yellow hat buy last night?

5. a. Juan wanted to eat something last night, so he did
b. aQué comio Juan anoche?

c. What did John eat last night?

162

6. a. Maria wanted to drink something yesterday so she did
b. aQué tomo Maria ayer?
c. What did Maria drink yesterday?
7. a. The boy in the yellow hat wanted to go somewhere yesterday, so he did
b. gAdonde fue el niflo de la gorra amarilla?
c. Where did the boy in the yellow hat go yesterday?

8. a. The girl with the brown hair saw her boyfriend last night and wanted to dance
perfectly for him, so she did

b. aComo bailo la nifia del pelo cafe anoche?
c. How did the girl with brown hair dance last night?
9. a. John wanted to play some game last night, so he did
b. aQué juego jugo Juan anoche?
c. What game did Juan play last night?
10. a. The woman with the red dress on wanted to play some game last night, so she did
b. gQué juego jugo la mujer del vestido rojo anoche?
c. What game did the woman with the red dress on play last night?
1 1. a. The man wanted three tacos last night
b. aQue queria el hombre anoche?
c. What did the man want last night?
12. a. The very old man wanted three apples last night
b. gQué queria el hombre muy viejo anoche?

c. What did the very old man want last night?

163

Appendix E.1
As noted in the Methodology section of this dissertation, the structures used for data
collection on Tense Doubling vary, along the following dimensions:

0 The amount of linguistic material between the head noun in the matrix clause
of the given context, and the verb that becomes the targeted main verb in the
question response

0 The type of Wh-question used: what, where, how

0 The type of verb used as the targeted main verb in the question response

Below is a table that breaks down each of the structures used, listed in Appendix D,

according to the three ways these varied:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Item # type of verb used type of Wh- Yes/No: extra
question used linguistic material
1 verb type l-buy/bought What N
2 verb type: l—eat/ate What Y
3 verb typg l-drink/drank What Y
4 verb type: l-buy/bought What Y
5 verb type: l-eat/ate What N
6 verb type: l-drink/drank What N
7 verb type: l-go/went Where Y
8 verb type: 2-dance/danced How Y
9 verb type: 2-plaJ/played What N
10 verb type: 2-play/played What Y
11 verb type: 3-want/wanted What N
12 verb type: 3-want/wanted What Y

 

 

164

Appendix F: ARSMA-II (Acculturation Scal_e)

Below is a representation of the ARSMA-II, drawn from Cuellar, 1., Arnold, B., &
Maldonado, R. (1995). There is also a version of this scale in Spanish.

The responses are according to the following Likert scale:

(1) Not at all

(2) Very Little/Not very Much

(3) Moderately

(4) MuchNery Often

(5) Almost Always/Extremely Ofien

1. I speak Spanish ................................. (l) (2) (3) (4) (5)
2. I speak English ................................. (1) (2) (3) (4) (5)
3. I enjoy speaking Spanish ..................... (1) (2) (3) (4) (5)
4. 1 associate with Anglos ....................... (1) (2) (3) (4) (5)
5. I associate with Mexicans and/or (1) (2) (3 ) (4) (5)
Mexican Americans ...........................
6. I enjoy Spanish language music ............. (l ) (2) (3) (4) (5)
7. I enjoy listening to English (1) (2) (3) (4) (5)
language music .................................
8. 1 enjoy Spanish language TV ................. (1) (2) (3) (4) (5)
9. l enjoy English language TV .................. (1) (2) (3) (4) (5)
10. i enjoy English language movies ............. (1 ) (2) (3) (4) (5)
11. 1 enjoy Spanish language movies ............ (1) (2) (3) (4) (5)
12. I enjoy reading books in Spanish ............. (1) (2) (3) (4) (5)
13. I enjoy reading books in English ............. (1) (2) (3) (4) (5)
14. I write letters in Spanish ....................... (1) (2) (3) (4) (5)
15. I write letters in English ....................... (1) (2) (3) (4) (5)

165

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

My thinking is done in the
English language ...............................

My thinking is done in the
Spanish language ...............................

My contact with Mexico has been ...........
My contact with the USA has been ..........

My father identifies or identified
himself as “Mexicano” .........................

My mother identifies or identified
herself as “Mexicana” ..........................

My friends while I was growing up
were of Mexican origin ........................

My friends while I was growing up
were of Anglo origin ...........................

My family cooks Mexican foods .............

My fi'iends now are of
Anglo origin .....................................

My friends now are of
Mexican origin ..................................

I like to identify myself as an
Anglo American ................................

I like to identify myself as an
Mexican American .............................

1 like to identify myself as Mexican .........

I like to identify myself as Mexican .........

166

(1)

(1)

(1)
(1)
(1)

(1)

(1)

(1)

(1)
(1)

(1)

(1)

(1)

(1)
(1)

(2)

(2)

(2)
(2)
(2)

(2)

(2)

(2)

(2)
(2)

(2)

(2)

(2)

(2)
(2)

(3)

(3)

(3)
(3)
(3)

(3)

(3)

(3)

(3)
(3)

(3)

(3)

(3)

(3)
(3)

(4)

(4)

(4)
(4)
(4)

(4)

(4)

(4)

(4)
(4)

(4)

(4)

(4)

(4)
(4)

(5)

(5)

(5)
(5)
(5)

(5)

(5)

(5)

(5)
(5)

(5)

(5)

(5)

(5)
(5)

BIBLIOGRAPHY

167

BIBLIOGRAPHY

Boissevain, J. (1974). Friends of friends, networks, manipulators and coalitions. Oxford,:
Blackwell.

Bonilla, F. (1998). Borderless borders: U.S. Latinos, Latin Americans, and the paradox
of interdependence. Phiadelphia: Temple University Press.

Borer, H. and Wexler, K. (1987). The maturation of syntax. In T. W. Roeper, E. (Ed.),
Parameter setting. Dordrecht: D. Reidel Publishing Company.

California Department of Education, (2003). Migrant education program. Available:
http://cde.ca.gov/iasa/migregions.html [2003, May].

Carreiras, M. and C., C. Jr. (1993). Relative clause interpretation preferences in Spanish
and English. Language and Speech, 36(4), 353-3 72.

Census, U. S. (2003). FactFinder, 2000 Census. Available:
http://factfinder.census.gov/servlet/BasicFactsServlet [2003, April].

Chomsky, C. (1969). The acquisition of syntax in children fiom 5 to 10. Cambridge,
Mass.,: M. I. T. Press.

Chomsky, N. (1986). Knowledge of language: its nature, origins, and use. New York:
Praeger.

Chomsky, N. (1995). The minimalist program. Cambridge, Mass: The MIT Press.

Clark, E. (1995). Later lexical development and word formation. In P. M. Fletcher, B.
(Ed.), The handbook of child language. Oxford: Blackwell.

Crystal, D. (1997). English as a global language. New York: Cambridge University
Press.

Cuellar, 1., Arnold, B., & Maldonado, R. (1995). The acculturation rating scale for
Mexican-Americans-II (ARMSA-II): A revision of the original ARMSA scale.
The Hispanic Journal of Behavioral Sciences, 17 (3), 275-304],. The Hispanic
Journal of Behavioral Sciences, 1 7(3), 275-304.

de Villiers, J. (1995). Empty categories and complex sentences: The case of wh-
questions. In P. M. Fletcher, B. (Ed.), The handbook of child language. Oxford:
Blackwell.

168

Dorian, N. (1999). Linguistic and ethnographic fieldwork. In J. Fishman (Ed.), Language
and ethnic identity. New York: Oxford University Press.

Durand, J. and M., D. (1999). Retablos of Mexican migrants to the United States. In L.
Goldin (Ed.), Identities on the move: Transnational processes in North America
and the Caribbean Basin. Albany: Institute for Mesoamerican Studies, University
at Albany.

F ishman, J. (1997). Language and ethnicity: The view from within. In F. Coulmas (Ed.),
The handbook of sociolinguistics. Oxford: Blackwell.

Fishman, J. (1999). Sociolinguistics. In J. Fishman (Ed.), Language and ethnic identity.
New York: Oxford University Press.

Fought, C. (2002). Ethnicity. In J. K. Chambers, Trudgill, P. and Schilling-Estes, N.
(Ed.), The handbook of language variation and change. Malden, MA: Blackwell.

Giles, H. (1979). Ethnicity markers in speech. In K. R. G. Schere, H. (Ed.), Social
markers in speech. Cambridge: Cambridge University Press.

Giles, H. and J ., P. (1987). Ethnolinguistic identity theory: A social psychological
approach to language maintenance. International Journal of the Sociology of
Language, 68, 69-99.

Goldin, L. (1999). Transnational identities: The search for analytic tools. In L. Goldin
(Ed.), Identities on the move: Transnational processes in North America and the

Caribbean basin. Albany: Institute for Mesoamerican Studies, University at
Albany.

Goodluck, H. (1991). Language acquisition: a linguistic introduction. Oxford, UK ;
Cambridge, USA: Blackwell.

Grosjean, F. (1982). Life with two languages: an introduction to bilingualism.
Cambridge, Mass.: Harvard University Press.

Gumperz, J. and Cook-Gumperz, J. (1982). Introduction: Language and the
communication of social identity. In J. Gumperz (Ed.), Language and social
identity. Cambridge: Cambridge University Press.

Hakuta, K. (1986). Mirror of language: the debate on bilingualism. New York: Basic
Books.

Hamers, J. F ., & Blane, M. (2000). Bilinguality and bilingualism ( 2nd ed.). Cambridge ;
New York: Cambridge University Press.

169

Kontra, M. (1999a). "Don't speak Hungarian in publicl"--A documentation and analysis
of folk linguistic rights. In M. Kontra, Phillipson, R., Skutnabb-Kangas, T., &
Varady, T. (Ed.), Language: A right and a resource. Budapest: Central European
University Press.

Kontra, M. (1999b). Some Reflections on the Nature of Language and its Regulation.
International Journal on Minority and Group Rights, 6, 281-288.

Krashen, S. (2001). Are children ready for the mainstream after one year of structured
English immersion? TESOL Matters, 11(4).

Le Page, R. B., & Tabouret-Keller, A. (1985). Acts of identity: Creole-based approaches
to language and ethnicity. Cambridge ; New York: Cambridge University Press.

Lloyd, P. M. (1968). Verb-complement compounds in Spanish. Tiibingen,: Niemeyer.

Lust, B., Flynn, S., and Foley, C. (1996). What children know about what they say:
Elicited imitation as a research method for assessing children's syntax. In D.
McDaniel, McKee, C., & Cairns, H. (Ed.), Methods for assessing children 's
syntax. Cambridge: MIT Press.

Mackey, W. (2000). The description of bilingualism. In L. Wei (Ed.), The bilingualism
reader. London: Routledge.

Matute-Bianchi, M. (1986). Ethnic identities and patterns of school success and failure
among Mexican-descent and Japanese-Amerian students in a California high
school: An ethnographic analysis. American Journal of Education, 23 3-255.

Meisel, J. M. (1992). The acquisition of verb placement: functional categories and V2
phenomena in language acquisition. Dordrecht ; Boston: Kluwer Academic
Publishers.

Milroy, L. (1980). Language and social networks. Oxford: B. Blackwell.

Milroy, L. and. Milroy, J. (1992). Social network and social class: Toward an integrated
sociolinguistic model. Language in Society, 21(1), 1-26.

Noels, K., Pon, G. and Clement, R. (1996). Language, identity and adjustment: The role
of linguistic self-confidence in the acculturation process. Journal of language and
social psychology, 15(3), 246-264.

Pajaro Valley Unified School District. (2000). Pajaro Valley Unified School District
Data Source Book. Watsonville, CA: PVUSD.

170

Phillipson, R. (1992). Linguistic imperialism. Oxford: Oxford University Press.

Phinney, J. S. (1990). Ethnic identity in adolescents and adults: Review of research.
Psychological Bulletin, 108, 499-514.

Radford, A. (1990). Syntactic theory and the acquisition of English syntax: the nature of
early child grammars of English. Oxford, Ox., UK ; Cambridge, Mass., USA:
Blackwell.

Radford, A. (1995). Phrase structure and functional categories. In P. M. Fletcher, B.
(Ed.), The handbook of child language. Oxford: Blackwell.

Romaine, S. (1995). Bilingualism ( 2nd ed.). Oxford, UK ; Cambridge, Mass., USA:
Blackwell.

Rosaldo, R. (1985). Assimilation revisited. Palo Alto: Stanford University, Stanford
Center for Chicano Research.

Rumbaut, R. (1996). The crucible within: Ethnic identity, self-esteem, and segmented
assimilation among children of immigrants. In A. Portes (Ed.), The new second
generation. New York: Russell Sage Foundation.

Ryder, M. E. (1994). Ordered chaos: the interpretation of English noun-noun
compounds. Berkeley: University of California Press.

Santa Cruz Library Weekly, (2003). Watsonville Part I-Spuds and Apples. Available:
http://santacruz.about.com/library/weekly/aa090797c.htrn [2003, April].

Santiago-Rivera, A., and Santiago, C. (1999). Puerto Rican transnational migration and
identity: Impact of English-language acquistion on length of stay in the United
States.

Skutnabb-Kangas, T. (1999). Linguistic diversity, human rights and the "free" market. In
M. Kontra, Phillipson, R., Skutnabb-Kangas, T., & Varady, T. (Ed.), Language: A
right and a resource. Budapest: Central Europena University Press.

Stockwell, R. P., Bowen, J. D., & Martin, J. W. (1965). The grammatical structures of
English and Spanish. Chicago,: University of Chicago Press.

Tabouret-Keller, A. (1997). Language and identity. In F. Coulmas (Ed.), The handbook of
sociolinguistics (Vol. Blackwell). Oxford.

171

Taj fel, H. (1982). Social identity and intergroup relations. Cambridge [Cambridgeshire] ;
New York, Paris: Cambridge University Press;Editions de la Maison des sciences
de l'homme.

Tollefson, J. W. (1991). Planning language, planning inequality: language policy in the
community. London ; New York: Longrnan.

Trudgill, P. (1972). Sex, covert prestige and linguistic change in the urban British English
of Norwich. Language in Society, 1, 179-195.

Trudgill, P. (1974). The social difikrentiation of English in Norwich. Cambridge [Eng]:
University Press.

Vasquez, O. A., Pease-Alvarez, L., & Shannon, S. M. (1994). Pushing boundaries:
language and culture in a Mexicano community. Cambridge [England] ; New
York: Cambridge University Press.

Warren, B. (1978). Semantic patterns of noun-noun compounds. , Acta Universitatis
Gothoburgensis, Goteborg.

Watsonville, City of. (2003). History of Watsonville. Available:
http://www.ci.watsonville.ca.us/information/history.html. [2003, April].

Watsonville Public Library, (2003). Pajaro Valley local history. Available:
http://www.watsonville.lib.ca.us/lochistory.html [2003, April].

Weber, D. (2000). Historical perspectives on Mexican transnationalism: With notes from
Angumacutiro. In N. Klahn, Castillo, P., Alvarez, A., and Manchon, F. (Ed.), Las
nuevasfronteras del siglo XXI. Mexico City: La J omada Ediciones.

Wei, L. (2000). Dimensions of bilingualism. In L. Wei (Ed.), The bilingualism reader.
London: Routledge.

Wolfram, W. and Fasold, R. (1974). The study of social dialects in American English.
Englewood Cliffs, N.J.: Prentice-Hall, Inc.

Zentella, A. C. (1997). Growing up bilingual: Puerto Rican children in New York.
Malden, MA: Blackwell Publishers.

172

 

llljllgjllrjjlljll[13)le