Li

5113“
.

v .n . . .
Waning.

 

a... g m
«www.
32%.

AK}. r;
v1.5.

 

 

. ME"??? .

 

(£2
02 m

 

_LIBRARY
Michigan State

 

I

L

University

This is to certify that the
thesis entitled

DIALECT REGISTER SHIFT IN RELATION TO COMMUNITY

STRATIFICATION: APPLYING LINGUISTIC CAPITAL
THEORY

presented by

JOHANNA ROSE WEDDLE BOULT

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Communicative Sciences and
Disorders

 

 

 

M QflM/zm/

Major‘Proféssor’s Signature

9/92/10 7
/ /

Date

MSU is an affirmative-action, equal-opportunity employer

----.-.------------n---.--.-.—._._._.-.—A--.-.-----.--.-.—.-.-.-.------:--.-.----.-.-.-.-.-u-.--o-

 

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.

 

DAIEDUE

DAIEDUE

DATEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/C|RC/DateDue.indd-p.1

 

 

DIALECT REGISTER SHIFT IN RELATION TO COMMUNITY STRATIFICATION:
APPLYING LINGUISTIC CAPITAL THEORY

By

Johanna Rose Weddle BouIt

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Communicative Sciences and Disorders/Urban Studies

2007

ABSTRACT

DIALECT REGISTER SHIFT IN RELATION TO COMMUNITY STRATIFICATION:
APPLYING LINGUISTIC CAPITAL THEORY

By
Johanna Rose Weddle Boult

Learning to be a competent communicator entails the knowledge and skills to use
effective and appropriate language across many contexts. This includes the knowledge of
the various meanings attached to different social contexts as well as the knowledge of
specific language structures used to navigate each of these contexts. These combined
skills can be termed “register shifting”. Schooling is one of the register-shifting contexts
in which all US. children are legally obligated to participate. In this context, all children
learn that there is one register in which they speak to their peers and another in which
they speak to their teachers. Children’s ability to shift effectively from one register to
another is dependent on their access to the language of instruction (i.e. linguistic capital).
Typically developing children in all communities acquire those types of linguistic capital
that allow them to become competent communicators within their local communities. But
it is unknown whether children who speak non-mainstream English varieties have equal
access to the Standard Mainstream American English (SMAE) used for school instruction
(i.e. the academic register). Answering this question is important because of its
implications for explaining the academic underachievement of African American
children compared to other groups.

Although Socioeconomic status (SES) is often viewed as a central sociological
variable in determining patterns of academic achievement and the use of nonmainstream

English features, the present study brings the theory of linguistic capital to the discussion

by investigating the additional sociological variable of racial-residential segregation
(SEG) and the combined effects of SES and SEG on the register-shifting skills of African
American children. Register shifting is brought to the fore as one particular element of
communicative competence that reflects a child’s knowledge of appropriate language
use. Specifically, African AmeriCan children’s use of selected lexical semantic items and
request forms were elicited in a role playing (or “controlled improvisation”) task that
simulated teacher-centered and peer-centered contexts. Furthermore, participants were
asked to demonstrate both receptive knowledge (through pointing tasks) and expressive
knowledge (through completion of cloze tasks) of register-shifting skill.

The participants were 36 third-graders selected through stratified, random
sampling from four different communities within suburban Detroit: (1) High
Segregation/High SES, (2) High Segregation/Low SES, (3) Low Segregation/High SES,
and (4) Low Segregation/Low SES. Results indicated that African American children
possess peer-centered to teacher-centered register-shifting skill. Furthermore, it was
found that this ability varies depending on task type and modality. Specifically,
participants displayed more shifting behavior on lexical semantics task items than on
politeness of request forms items. Community differences were noted in terms of the
mean length of utterance and the use of address forms as alternative means of indicating
shifts in expressive politeness. The results of this study have implications for research
methodology, educational policy, and evaluation methods used by educators and Speech-
Language Pathologists (SLPs).

Keywords: register, academic register, register—shifting, code-switching, linguistic

capital, dialect density, AAE, SMAE.

DEDICATION
This dissertation is dedicated to:

my mother,
who taught-me the value of the spoken word
as a tool to communicate
the deepest sentiments of the heart;

my husband,
whose words continue to touch me
in the most unexpected and impressive ways;

and my children,
who, even before they could speak,
shared with me a new
language of love.

ACKNOWLEDGMENTS

The following people and organizations were instrumental to completion of this
dissertation: Dr. Ida Stockman, Dr. David Dwyer, Dr. Joe Darden, Dr. M. Lorraine
Wynn-Dancy and Dr. Peter Lapine; thank you for your guidance throughout the life of
this project. Special thanks are owed to Dr. Stockman whose unflagging belief in my
ideas and encouragement to persevere were invaluable; the Michigan State University
(MSU) Urban Affairs Program for funding for the first three years of my doctoral
program; the MSU Department of Communicative Sciences and Disorders, for multiple
scholarships and internship opportunities; the MSU Graduate School, for funding this
project’s completion; Dr. Mary Jo Hidecker, Dr. Diane Ogiela, Dr. Gregory Robinson,
and Robin Roots who provided extensive feedback on the writings, which eventually led
to this completed document; Tamara Corbin, Christine F oshee, Dior McDaniel, Kristine
Akers, Matt Kelly, Ashley Webb, Maggie Cooper, Rachel Ort, Annie O’Brien, Raechel
Szymanski, Kellin McGinn, Janalee Keegstra, Laurel Conrad, Meghan Hubert, Megan
Bigalke, Samantha Bissell, Elaina Swartzlander, Molly McPartland and Summer
Chapman who spent hours collecting, organizing and scoring data. Thanks especially to
Tamara Corbin who sacrificed food, sleep and even gasoline to accomplish the goals of
this study; Leah, Jayla and Jamaya Humes, who aided in collection of pilot data and
remain our dear and life-long friends; Barb Houck and Nate Higley for their wonderful
artwork. Thank you for the grueling hours it took to draw representations of the precise
linguistic constructs used in this study; and finally to all of the faculty and staff of MSU
and the University of Louisiana at Monroe (ULM) who contributed their blood, sweat

and tears both directly and indirectly so that this project might be completed.

TABLE OF CONTENTS

LIST OF FIGURES .............................................................................. x
CHAPTER 1 ...................................................................................... l
Pragmatics in Historical Context ........................................................... I
Register: An Aspect ofPragmatics............................................................5
Register and Style ................................................................... 6

Style as a Continuum of F ormality ....................................... 6

Style as Manner of Speaking .............................................. 6

Register and Style Reconciled ............................................ 7

Systemic Functional Theory of Register ......................................... 9

Register .................................................................... 10

Field ................................................................ IO

Tenor ............................................................... l3

Mode ................................................................ 14

Summary of Registral Components ........................... 16

Genre ......................................................................... l6

Ideology .................................................................... 17

Register as Linguistic Capital ..................................................... 22

Statement of the Problem for the Current Study ......................................... 25

Linguistic Context: Language as a proxy for school underachievement..27
Extralinguistic Context: Social class as a proxy for school

underachievement ................................................................. 29
Lower Class ............................................................... 29
Middle Class ............................................................... 31
The Need to Revisit Language and Social Context as Proxies for School
Achievement ....................................................................... 32
An Expanded View of Language ....................................... 32
An Expanded View of Social Context ................................. 34
Purpose of Study ............................................................................. 36
CHAPTER 2
METHOD ........................................................................................ 37
Participants ...................................................................................... 37
Selection Criteria 37
Selection Procedures .............................................................. 38
Global Screening Procedures... ...38
Language and Hearing Screening Procedures ........................ 42
Language screening procedures ............................... 43
Hearing screening procedures ................................. 43
Operationalization of the Dependent Variables ......................................... 44

vi

Register Shift between Peer-centered and Teacher—centered Contexts. . ..44

Self-reflective Report of Register Use .......................................... 46
Operationalization of the Independent Variables ...................................... 47
Language Task Variables ........................................................ 47
Lexical Semantics Task .................................................. 47
Request Task ............................................................... 49
Performance Modality Type ...................................................... 49
Community Type .................................................................. 50
Socioeconomic Status (SE8) ............................................ 52
Segregation (SEG) ........................................................ 54
Stimulus Construction ...................................................................... 55
Lexical Semantics Task Stimuli .................................................. 56
Request Task Stimuli ............................................................... 56
Self-Reflective Task Stimuli ....................................................... 63
Data Collection Procedures ................................................................ 64
Task Description .................................................................. 64
Task Administration ............................................................... 65
Registral Shift Tasks ..................................................... 65
Self-reflective Task ........................................................ 68
On-Site Recording of Responses ................................................. 68
Preparation and Scoring of Data for Analysis of Register Shifting .................. 69
Data Scoring for Request Task69
Receptive Modality ...................................................... 69
Expressive Modality ...................................................... 70
Data Scoring for Lexical Semantics Task ...................................... 70
Receptive Modality ...................................................... 7l
Expressive Modality ...................................................... 71
Data Scoring for Self-reflective Task ........................................... 72
Reliability of Data Scoring ................................................................ 72
Data Analysis ................................................................................ 73
Analysis of Registral Shift ....................................................... 73
Analysis of Self-Reflective Report ............................................. 73
CHAPTER 3
RESULTS ........................................................................................ 75
Reliability of Scoring Register-Shifting Behavior ...................................... 76
Presence of Shifting Behavior ............................................................ 76
Categorical Distribution of Shifting Behavior ........................................... 79
Variation in Shifting Behavior in Relation to Community, Task and
Modality Types ............................................................................. 80
Post Hoc Analysis of Shifting Behavior across Community Type .................. 86
Categorical Distribution of Register Shifting by Community ............... 86
Alternative Linguistic Means for Indicating Register-Shifts by
Community Type ................................................................... 86
Mean Length of Utterance by Community Type ..................... 88
Method of analysis .............................................. 88

vii

Results by community type ..................................... 89

Specific Politeness Forms by Community Type ...................... 91
Method of analysis .............................................. 91
Results by community type ..................................... 91
Conclusions from Post hoc Analyses of Community Type. . . . .....93
Self-Reflective Report of Register-Shift Use ........................................... 93
Community Differences in Participants’ Reports of Whether They Speak
Differently with Teachers than Peers ......................................................... 93
Participant Reports of How and Why They Speak Differently with
Teachers and Peers ................................................................ 94
Summary and Conclusions Regarding General Findings ............................. 96
CHAPTER 4
DISCUSSION ................................................................................... 98
Research Problem and Theoretical Framework Revisited ............................ 98
Register Shift .............................................................................. 100
Register Shifting: A Modality- and Task-Dependent Skill .......................... 100
Register Shifting: A Modality Dependent Skill ............................. 101
Register Shifting: A Task-Dependent Skill ................................. 102
Relationship of Community Type to Register Shifting ............................... 104
Community Differences in Number of Strong Shifters ..................... 105
Community Differences in MLU as a Signal of Register Shift ............ 105
Definition of Community Type Revisited .................................... 105
Implications ................................................................................ l 10
Theoretical Issues ................................................................. 110
Systemic Functional Theory .......................................... 110
Controlled Improvisation .............................................. l 12
Relationship of Language and Community ......................... 112
Applied Issues ..................................................................... 113
Implications for Research Methods .................................. 113
Natural vs. experimentally controlled contexts of
observation ....................................................................... l 13
Minority participant recruitment for research studies. . ...1 l6
Participant recruitment using top-down definition of
community type .................................................. 1 18
Definition of community type in relation to language
use ............................................................... 118
Framing future studies of pragmatics ....................... 120
Implications for Clinical and Educational Assessment ................ 120
Implications for Clinical and Educational Instruction .................. 121
Educational Policy ............................................................ 122
Future Outlook ............................................................................. 123
APPENDICES .................................................................................. 126
Appendix A: Informed Consent Procedures .......................................... 127

Appendix B: Schools/Neighborhoods Designated by SES/SEG Groups. . . . . ....129

viii

Appendix C: Direct and Indirect Requests According to Modal Type ............ 132

Appendix D: Data Collection Scripts ................................................. 133
Appendix E: ABILITIES Index .......................................................... 135
Appendix F: Data Collection Forms ................................................... 138
Appendix G: Lexical Semantics Screener ............................................ 155
Appendix H: Randomized Number Sets Used For Construction of Data
Collection Lists ........................................................................... 156
Appendix 1: Reduced Price Lunch Screener .......................................... 157
Appendix J: Hearing Screening ......................................................... 158
Appendix K: Additional Statistical Analyses ........................................ 159
Appendix L: Self-Reflective Report of Register Use Questionnaire .............. 160
REFERENCES CITED ........................................................................ 161

ix

LIST OF TABLES

Table 1: Participant Characteristics .......................................................... 39
Table 2: Inclusionary and Exclusionary Selection Criteria ............................... 41
Table 3: Lexical Semantics Task Items ..................................................... 57
Table 4: Expressive Request Forms Task Stimuli ......................................... 59
Table 5: Receptive Request Forms Task Items ............................................... 61
Table 6: Task Conditions by Task Type and Performance Modality .................... 64
Table 7: Reliability of Scoring of Tasks ..................................................... 77
Table 8: Number of Children by Number of Expected Shifts ............................ 78
Table 9: Comparison of Undistributed and Distributed Shifters. . . . . . . . . . . ...79
Table 10: Extent of Distribution of Shifting Behavior by Number and Percentage...80
Table l 1: Analysis of Variance in Percent Register Shift Scores ........................ 81
Table 12: Nature of Shift Distribution by Community Type ............................. 87
Table 13: Community Differences between Teacher and Peer-Centered Mean

Lengths of Utterance (MLUs) ................................................... 89
Table 14: Mean Frequency of Rote-Polite “Please” and Address Terms Used in

Expressive Request Task by Commumty92
Table I5: Self-Report of Register Shifting: Participant Numbers and Means by

Community Type ................................................................... 95

Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:

Figure 9:

LIST OF FIGURES

The Relationship of Registral Components to Language Components. ....11

The Relationship of Register to Genre ............................................ 18
The Relationship of Register to Ideology ....................................... 20
The lntemal Location of Ideology ............................................... 21

Community Type as Defined by SES and SEG ................................ 51

3x2x2 ANOVA ...................................................................... 74
Percent Shift by Community by Task ............................................ 83
Main Effect of PRS by Modality by Community ............................... 84
Main Effect of PRS by Task Type by Community ............................ 85

xi

CHAPTER I

Communicative competence (Hymes, 1974) requires that competent language
users learn to apply their linguistic knowledge within the real world. People use this
“pragmatic” skill as a basic, human adaptation to their various social environments.

Pragmatics in Historical Context

Pragmatic skill extends beyond the empiricist notion of “linguistic competence”
(Chomsky, 1965), which is limited to the internal, static knowledge of linguistic
structure. Communicative competence, in contrast, entails the dynamic process of
language use. This distinction between linguistic and communicative competence is a
distinction brought to the fore in the early 20th century by Ferdinand de Saussure’s (1910-
I911) conceptualization of the Iangue/parole dichotomy. His thought was that the
objectified and systematic forms of language (Iangue) were the only viable candidates for
scientific inquiry. Parole can be equated with language use or pragmatics. Its dynamic
and transitory nature was thought to preclude it from being an object of scientific study.
Chomsky’s (1965) linguistic theory utilized the basic Saussurian notion that viewed the
primacy of language structure as opposed to its use as the focus of linguistic inquiry.

Subsequently, the pragmatics revolution (Duchan, 1984) of the 19705 offered
new insights into the study of language use. For example, the fathers of speech act theory
(Austin, 1962; Grice, 1968, 1969; Searle, 1975) differentiated between form and use in
arguing for the centrality of language use. Austin (1962) differentiated locutions, on the
one hand, from illocutionary and perlocutionary acts, on the other. Locutions are those
decontextua‘lized acts that convey information in and of themselves, which according to

Austin, can be done only in the abstract sense. Furthermore, Austin suggested that

locutions can be thought of as purely structural and in parallel to Saussure’s langue.
Beyond locutions, Austin proposed the levels of illocutionary force (the speaker’s intent)
and perlocutionary effect (the result on the hearer), both of which conceptualized
language as more than just form. That is, language was conceptualized as being important
in terms of the act that it was used to perform. Searle (1975) developed the notion of the
speech act by distinguishing between literal and indirect speech acts. The most famous
example of this distinction is “Can you pass the salt?” The literal act might be interpreted
as a request for information regarding one’s ability to pass the salt, i.e., “Are you
physically able to pass the salt?” But, in fact, the non-literal “indirect” act (requesting the
action of passing the salt) is the much more likely interpretation. Finally, Grice (1968;
1969) provided his conversational maxims that hold that conversational partners obey
certain rules in order to cooperate within the conversational endeavor. He proposed that
speakers use nonliteral meanings in order to communicate most effectively, and that
conversational partners use certain assumptions (e. g., that the partner is telling the truth,
making the statement as strong as possible, but not overstating the truth) in order to reach
a common understanding.

Speech act theory also led to the development of the major theories of
conversational politeness (Brown & Levinson, 1987; Goffman, 1967). Goffman wrote
about politeness in terms of “face”. Traditionally a folk notion, Goffman defined face
specifically as “the positive social value a person claims for himself by the line [i.e.,
stance] others assume he has taken during a particular contact” (p. 213). Brown and
Levinson argued further that it is in our best interest to avoid conflict by maintaining the

social face of others and thereby, being polite.

Given these basic theoretical underpinnings, researchers over the past two and a
half decades have engaged in research related to normal, adult use of pragmatic devices
as well as how children learn to use these devices. The latter of these is commonly
referred to as developmental pragmatics (N inio & Snow, 1999). Developmental
pragmatics entails the notion that the social environment contributes to and is affected by
a child’s language development. In Bloom and Lahey’s (1978) classic distinction
between content, form, and use, the process by which children internalize language based
upon situations of use was described as follows:

The mental plan that the child acquires is a system of rules-cognitive-linguistic

rules for pairing sound (or movements) with meaning in messages, and social

rules for pairing sound/meaning or movement/meaning connections with different
situations. . .At the same time that the behaviors of saying and understanding
depend on a mental plan for action, the plan itself is influenced and shaped
developmentally by the same behaviors. There is a mutual effect between
behaviors and competence---children learn about language in the process of

saying and understanding” (Bloom & Lahey, page 22).

Developmental pragmatics (according to Ninio & Snow) has been conceptualized in four
distinct domains: (1) the acquisition of intents and their linguistic counterparts, (2) the
development of conversational skills and devices (such as tum-taking, topic initiation,
maintenance, and shift), (3) the development of culturally determined norms for
language use, such as politeness rules and (4) the development of organization and
cohesion (through use of linguistic devices) in discourse.

The suggestion (in Ninio and Snow’s domains of pragmatics listed above) that
elements of language use are determined by cultural norms is a foundational principle of

the present research. Language socialization is the study of the ways that people

socialize their children about language (Goldfield & Snow, 1992; Ochs, I988; Ochs &

Schieffelin, I979; Schieffelin 1990; Schieffelin & Ochs, 1986). Language socialization
research has concerned itself with the study of how children learn to use language within
social contexts. These social contexts encompass various constellations of values, beliefs,
and norms that underlie behaviors, including linguistic behaviors. ). Interestingly, parents
tend to socialize their children’s language using primarily linguistic means, this is, by
talking to them about language use. Various language behaviors are used to teach
children the social norms of their particular community. Schieffelin and Ochs (1986), in
their landmark compilation of studies on child language socialization, highlight the
language behaviors of calling-out, repeating, prompting, teasing, etc. Each of these
behaviors can be considered a method of socialization that conveys those
aforementioned, underlying, culture-specific values. As a result of language socialization,
children are taught to behave appropriately within their own communities by means of
the instructional mode of language.

The relevance of socialization to the relationship between language use and form
was made explicit by child discourse scholars (Ervin-Tripp, 1970; 1977; Ervin-Tripp &
Mitchell-Keman, 1977; Heath, 1983; 1986). They suggested that although the functions
or uses of language are diverse, some of them are universal. But the social and linguistic
conventions for expressing universal functions vary according to one’s culture. For
example, request functions may exist universally, and yet the linguistic conventions that
indicate politeness of requests may differ from community to community. In addition,
variation exists along a developmental continuum with acquisition of functions and their

corresponding conventions varying with age.

From a developmental perspective, then, pragmatic skills can be described in
terms of one’s flexibility with application of language within and across various contexts.
That is, as we grow older, we are introduced to more and different social institutions.
Such institutions include school, organized sports, social clubs, marriage, business,
parenthood, and so on. For each institution, “rules of play” are involved (Bourdieu,
1993). In other words, there are general, behavioral norms associated with each
institutional context. For example, each institution demands the use of a different
linguistic “register”. The ability to change the way we talk when these various situations
require it, is an important aspect of normal, human development.

Register: An Aspect of Pragmatics

Register refers to the pragmatic skill that involves speakers changing the way they
talk in a given context as a reflection of specific, underlying, socially determined
meanings (Andersen, 1990; Halliday, 1970). The term, “register shift,” is used here to
describe the manner in which speakers use variation in linguistic form to serve different
functions within culturally imbedded situations of use. Different situations require one
language variety as opposed to another (e. g., Spanish as compared to English) (Reid,
1956) or simply one variety (dialect) of the same language as opposed to another
(Ferguson, 1959). In either case, a register shift entails a change in the form (morphology,
syntax, and/or phonology), content (semantics/lexical choice), and/or use (pragmatics) of
language. Register shift is reflected in the type of pronunciation, word order and
inflectional morphology use, the type of words chosen, as well as the type of discourse
structure, speech act types, politeness markers and so on. It can be argued though that the

formal, structural components of language reflect underlying and socially determined

meaning. One commonly cited example of a register is Child Directed Speech (CDS) or
“motherese”, a situation defined by the interlocutors: the adult and child). Another is
school register, in which the situation is defined by the school context as well as the
interlocutors, i.e. teacher(s) and student(s). The present study expects to validate the
bases of both types of register by investigating the differences between school register in
teacher-centered as compared to peer-centered contexts.
Register and Style

Because style and register often have been used interchangeably, it is necessary to
clarify how they are viewed in the present study. I will describe the common meanings of
style and, in doing so, explain the manner in which both register and style can be
encompassed by a well developed theory of register.
Style as a Continuum of F ormality

There are two major, traditional lenses through which linguistic style has been
viewed. The first relates to the distinction between formality and casualness. This
“continuum of formality” was first proposed by William Labov (1966) following his
famous research on /r/ use in the casual and formal speech of New York City department
store workers. In this study, Labov viewed the casual style as the most representative of
the speakers’ true vernacular and the one in which /r/-lessness was most often observed.
In contrast, the more formal style contained more standard speech features including /r/-
fullness.
Style as Manner of Speaking

A second conceptualization of style is more recent. It deals with the way of

speaking that characterizes the idiolect ((“the language or speech pattern of one

individual [. . .]” (Merriam Webster Online Dictionary, Pennebaker & King, 1999)) or the
pattern of a group of people (e. g., the African American speech style (Morgan, 2002;
Rickford & Rickford, 2000)). In this sense, style can be characterized by variation in
word choice, intonation, rhythm, loudness, and so on. As a result, such linguistic and
suprasegmental characteristics are markers of social and/or ethnic identity. From a
psychological perspective, recent research has focused on the underlying values, beliefs
and attitudes that are realized in individuals’ stylistic variation (Pennebaker & King,
1999)

Register and Style Reconciled

The present study considers register to be broader than style. Whereas, style has
been characterized either as a continuum of formality or as manner of speaking, the
theory of register used in the present research incorporates both elements of style within
its scope. Style, in this sense, is viewed as a component of register.

The casual to formal continuum of speaking is a linguistic result of one aspect of
register, namely the role and status relationship between participants as defined by the
institution. In addition, register is broader than style in that register is a reflection of
social meaning. In the case of values related to interlocutor role and status, register relates
at a deeper level to variation in the underlying values system that is reflected in language.
Thus, register goes beyond simply labeling a text as being either casual or formal as a
simple definition of style would do. It relates the level of formality to the participant role-
related and social status-related meaning(s) reflected.

In addition, “style as manner of speaking” is entailed within the theory of register

used in the present study. Speaking style used by an individual or a group has not been

connected traditionally to the value systems of those individuals or groups however,
registral theory makes this connection. Specifically, registral theory holds that an
individual’s manner of speaking results from the fact that society has agreed upon certain
“stylistic” norms that individual speakers may choose to follow. Thus, register
encompasses style as a group’s manner of speaking in that it is the linguistic reflection of
the group’s culturally embedded (generic) meaning system.

Individual speech style (or idiolect) can also be viewed as one aspect of a broader
theory of register. Drawing on the work of Hymes (1974), cultural values and beliefs of
groups are not merely external and universally accessible entities. They also are stored
mentally and used by individual human beings. In this sense, an individual’s speech style
is determined by what socially available linguistic items he or she chooses to use.

However, if an idiolectal variant is adopted by others who are associated with that
individual, then the variant begins to have a shared meaning among the group members
that use it. That is, idiolectal variation can lead to registral variation. The actual element
of idiolectal variation would not be considered truly registral though until it obtained the
status of general social use and significance. Registral variation exists only on the scale
of an entire language community.

The conceptual overlap between style and register may have existed in the past
because of a narrow conceptualization of register. But in the theoretical framework used
in this study, style is acknowledged as a subcomponent or as subcomponents of the larger
concept of “register”. The description of registral theory in the next section of this

chapter will make clear where notions of style might fit.

In sum, within functional as opposed to structural linguistic theory, register is
viewed neither as a simple correlation between form and context nor as a simple means
of distinguishing casual and formal styles. Rather, register analysis, a component of
functional linguistic theory, allows the researcher to analyze language form for the very
purpose of making inferences about underlying societal values. F ormal rules of language
(phonology, syntax, morphology) are realized in this view as subordinate, yet
meaningfully reflective, components of a culturally specific system of production rules
(Bates and MacWhinney, 1979; Halliday, 1978; Prutting, 1982). Systemic Functional
theory provides a useful framework within which to view the registral changes that one
makes as meaningful and culturally embedded.

Systemic Functional Theory of Register

Systemic Functional linguistic theory (Halliday, 1978) is one of the most, well-
developed functional linguistic theories. It views language not as merely structural, but
also as a “social semiotic” or a symbolic system that allows conversational partners the
potential to co-construct meaning. For example, in a conversation involving two, young,
American, men, a reference made to one of their “cribs” would likely be understood to
refer to one of the young men’s “house” and not his “infant bed”. In this context, the
meaning of “crib” is clarified only by means of the word being spoken in a particular
social context. In this sense, the Systemic Functional school of thought deviates sharply
from the structural transforrnationists, like Chomsky and others, who have focused
traditionally on syntactic form as the most worthy subject of scientific inquiry. Systemic
Functionalists view social embeddedness as an essential context within which meaning is

imbued. Fundamentally, language cannot be understood apart from the context of its use

(Malinowski, 1935). Meaning is thus not static, overt, nor finite (Stockman, 1999). Its
potential is realized only upon use in context. Systemic Functional theory focuses not on
the linguistic structure per se, but on variations in contextual structure and their
relationship to corresponding variations in language meaning.

There are three major components of Systemic Functional Linguistic theory
(Halliday, 1978): register, genre, and ideology. Each one will be discussed in turn.
Register

Halliday (1978) defined register as the context of situation. Furthermore, there are
three and only three key aspects of register that affect language variation. These are the
type of social action (field), the interlocutor role relationships (tenor) and the
symbolic/linguistic organization of the text (mode). These situational or “registral” (i.e.
those related to the register) components are essential in that they affect the potential
interpretation of meaning.l Figure 1 identifies the three registral components and their
corresponding linguistic expressions.

Field. In general terms, field can be understood as the topic or the type and nature
of the social action in which the participants are engaged. One might better understand
the practical application of this concept by asking questions such as, “What type of task
are the participants engaged in?” or “What is language being used to talk about?” A
general example of field would be something like a third grade lesson. A specific
example of field might include a more detailed account of the aforementioned example,

such as, a third grade reading lesson using the children’s book, The Velveteen Rabbit.

 

’ There is conceptual overlap with what Halliday (1978) refers to as “register” and what Bourdieu refers to
as the “institution”. Bourdieu’s social institutions were the building blocks of larger society, made up of
different role-relationships (similar to Halliday’s “tenor”) and played-out on different social fields (similar
to Halliday’s “field”).

10

 

 

 

Mode

 

 

 

 

 

 

 

Textual

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

   

 

 

 

organization
Language
. Lexical
Politeness semantics
//
Register

Tenor

 

 

 

 

 

 

\

 

Field

 

 

 

 

Figure 1. The Relationship of Registral Components to Language Components

 

Field can be generally understood as the topic at hand. In addition, it involves the
depth (whether deep or shallow, technical or lay) with which the topic is being discussed.
The field conveys ideational meaning meant to reflect on an experience or a logical
construct. To some extent, one can get at this information by asking the five types of wh-
questions: who, what, where, when, and why? Field is reflected in the basic elements of
language: the nouns and noun phrases, the verbs, adverbs and verb phrases, and so on.
The linguistic representation of field can be found in vocabulary items that reflect the
content or substantive meaning of a message. The most characteristic lexical realization
of field is word choice in the form of technical jargon as opposed everyday language.
Particular technical lexical items are used to describe the available topics within specific
technical fields. In addition, technical grammar is often abbreviated and nonstandard in
order for those familiar with the jargon to communicate efficiently about a technical
situation.

Field also can be used to explain the use of homonyms, words that sounds the
same but mean something different, across different language communities. In one
community the word “poor” may mean “having little money” whereas in another
community the word “poor” may mean “very skinny” or “emaciated”. These similar
sounding words can be shared by two separate language communities due to borrowings
from one another or through a common borrowing from a third language. Although the
lexical items used may sound the same, they reflect very different meanings due to the
differences in cultural experiences and a difference in the range of possible meanings in

one community compared to the other.

12

Tenor. The second aspect of register presented by Halliday (1978) is tenor,
defined as the role relationship that exists between conversational participants, namely,
who is involved in the aforementioned field, and what are their respective roles and
statuses? This level of meaning is involved in the creation of interpersonal relationships.
Continuing with the examples provided above, the relevant (and basic) roles would
include the teacher and the students. The students would attribute certain status to their
peers and yet another status to their teacher. Differences in status attribution between in a
relationship between student and student as compared to between student and teacher
relationship can be attributed to power differential, frequency of contact and level of
affective involvement (Poynton, 1989). Higher status relationships are generally
characterized by greater power, less frequent contact, and less affective involvement. The
opposite is true of lower status relationships.

F ormality of language use is directly related to the role status of conversational
partners. For this reason, tenor can be equated with the type of “style” that distinguishes
casual from formal language. Informal language is used when people’s roles are equal in
power, they have frequent contact, and high affective involvement. Formal language is
used when people have a hierarchical power relationship, have infrequent contact, and
low affective involvement.

Among those linguistic items listed by Eggins (1994, p67), differences in the use
of modal verbs have been found to influence and be influenced by the tenor of a register.
In casual registers, modalization is likely to be used to express probability and/or opinion.
In contrast, modal use in formal registers is likely to express deference or suggestion.

Furthermore, modals are used to express an indirectness that signals politeness. In

l3

contrast, casual register is characterized by use of fewer modals signaling less polite
language.

In request formations, the use of the bare imperative occurs without modals, as in
the following sentence, the verb “to hand” occurs in a bare imperative form, Hand me a
turkey pot pie. Interrogatives can be modified using different modal verbs in order to
convey differences in politeness, for example Can/could would you hand me a turkey pot
pie. Use of modal verbs signal politeness, which in turn, signals a difference in the
underlying meaning system involved in a particular register. Much of this difference in
politeness is conveyed by varying levels of indirectness. Modals are devices that
contribute to the indirectness and resulting politeness of a clause containing such a
modified verb phrase.

Mode. The third and final aspect of Halliday’s (1978) register is mode. It refers, in
general, to the symbolic organization of a text which tends to vary depending on the
closeness of participants to one another (face to face conversation as opposed to reading a
manuscript that was written hundreds of years ago) and the closeness of the topic at hand
to the present conversation (for example, solving a present problem as opposed to telling
a story about a past event). The present study does not investigate mode however, I
introduce it here to provide the full scope of possible analyses within registral theory. I
have chosen not to investigate mode because the oral and face-to-face mode will remain
constant across all research conditions in the present study.

Mode is the meaning type that expresses a particular form of organization of
information. The analysis of mode answers the question, “What role does language play

in the interaction?” or “In what way is that information organized?” Mode includes both

14

the “channel” via which the communication takes place (e. g., spoken, written, face-to-
face, telephone, and so on) and the type of rhetorical device involved (e. g., persuasion,
exposition, and so on).

Additionally, mode can be understood to exist along two separate distance
continua, spatial/interpersonal distance and experiential distance. Spatial/interpersonal
distance refers to the actual, physical distance between interlocutors.2 Within face-to-
face interaction interlocutors have the closest spatial/interpersonal distance whereas
reading a written text involves the furthest spatial/interpersonal distance between
interlocutors. Experiential distance, on the other hand, refers to the distance between
language and the social process. For example, language can be used to talk about an
immediate situation or one that has occurred long ago, in the past. In the former the
experiential distance is, fundamentally, zero whereas in the latter, the experiential
distance is great.

Mode can be characterized as a continuum of spoken to written language. Spoken
(face-to-face) language has the closest interpersonal distance whereas written language
has the least. Eggins (1994, p57) provided the following meaningful distinctions between
the spoken and written language modes. Spoken language can be characterized by: tum-
taking, context dependence, dynamic (open-ended, interactive) structure, spontaneity
phenomena, everyday lexis, non-standard features, grammatical complexity, and lexical
sparseness. In comparison, written language can be characterized linguistically as:

monologically organized, context independent, structurally synoptic with closed & finite

 

2 The “distance” referred to here is purely physical. It can be contrasted with a different type of distance
related to the aforementioned “tenor”. “Tenor” involves not physical distance, but rather a social and/or
psychological distance related to the power differences in role-relationships.

l5

 

 

 

C(‘ll‘

Elm

Cllii

CLIII

 

rhetorical staging, lexically prestigious and dense, standard and grammatically simple,
standard, formal and so on.

Language can also be affected by the aspect of mode called experiential distance.
Along this continuum, language can be used in close temporal proximity to the social
process or in reference to a social process, which is distant in that it has already occurred
in time. Additionally, mode can be characterized by the text’s thematic structure or
organization. These patterns may describe textual elements such as the use of
foregrounding, reference, presupposition, and continuity.

Summary of registral components. Given this more developed
conceptualization of register within Halliday’s (I978) Systemic Functional theory,
contexts can be better understood as having the three, aforementioned registral
characteristics of field, tenor and mode, all of which are essential to the expression and
comprehension of meaning. Context, via the elements of register, determines (l) the
manner in which meaning is expressed and interpreted and (2) the language structures
and devices used to express such meanings. This theory aids the current research in that
situational context can be analyzed in terms of one or more of its registral components.
That is, one can use a registral analysis to analyze linguistic structures for the purpose of
deducing the nature of their underlying, socially-embedded meanings.

Genre

The notion of context must be expanded beyond the immediate situation of
language use (i.e., beyond register) to encompass the general cultural context in which
communication registers are used (Eggins, 1994). In this sense, culture can be defined as

the “generic” context. The larger cultural (or generic) context gives purpose and meaning

l6

 

 

lc’l

.'
.J
J

 

 

to all of the elements that occur within a given register. We can call this larger context the
genre. See Figure 2.

Culture defines the characteristics of activities that registers serve. Examples of
everyday genres include: participating in a class lessons, asking for directions, telling
stories, engaging in business transactions, and so on. Each of these genres is a
recognizable cultural activity. It is precisely the culture (genre) and its embedded
situation (register) that give language its social meaning. Different cultures may structure
the same genres differently so that varying configurations of registral variables are
allowed to exist within comparable genres.

Ideologz

The most abstract context, which contributes to this model, is that of “ideology”
(Eggins, 1994). Ideology can be defined as the context of values and beliefs. See Figure
3. Genre and ideology are reciprocally influenced by one another in that ideology is
genre’s source, and genre helps to perpetuate and/or change ideology. Ideology exists,
from above, as an overarching, institutionalized force, which is imposed on people by
those in power. From this perspective people in powerful roles use ideologies to structure
human behavior by imposing views about how a given individual should mean. These
institutionally sanctioned meanings impose themselves upon the relation of experience
(field), the creation of interpersonal relationships (tenor) and the use of language to
organize information (mode). In this sense, our ideological predispositions program our
verbal behavior and determine how we use language in a given social institution (e. g.,

school, home, marketplace and so on) by restricting the set

17

a...“ h

      

 

Language

 

 

 

 

 

Register

 

 

 

 

 

 

 

 

Figure 2. The Relationship of Register to Genre

 

 

 

 

VI;

ll.

V. -

(Ii?

([1?

9L

ll

 

of possible linguistic acts.

Figure 3 illustrates the relationship between ideology and its embedded elements
that contribute to the components of register. It shows that language successively reflects
the register, genre and ideological framework within which it functions.

One limitation of Systemic Functional theory is that it does not explicitly offer a
personal and/or psychological manifestation of ideology. In addition to the larger, social
locus of ideologies, they can be seen to exist within the psyche of each individual person
as opposed to an abstracted group psyche.

Figure 4 illustrates this contrasting dynamic; namely, that ideology also exists at
the level of the individual, and that language is the main tool used either to perpetuate
and/or challenge such ideology. In order to illustrate this idea the internal triangle,
labeled, “Language” in Figure 3 above, should possess an “Ideology” component with an
arrow pointing outward, as illustrated in Figure 4 below. In this sense, ideology can be
viewed not merely as a set of social values and beliefs imbued from the top-down, but
also as a set of values and beliefs held by individuals who, from the bottom-up, carry out
the action of imbuing.

Figure 4 shows that, during social interaction, an individual has the choice of
whether or not to accept the status-quo to which he/she has been socialized. Ideally, a
given individual may choose to reject a given ideology, such as the power inherent to a
given role assignment, by negotiating a change in the ideologically imposed power
differential. However, ideologies, tend to reinforce the existing power structure, and

thereby favor (or value) social constructions associated with the powerful. Graphically

l9

 

n Adapted from Eggins

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/ ‘t (1994), p. 34, Genre and
,I' \‘ register in relation to
X ‘\ language
I \
’1’ \‘\
,’ \ Register:
I \
I’ ‘, Context of
,I’ ’~\ Situation
I \‘
I, ‘\
,’ \ Genre:
I \
I’ , Context
/' ‘x‘ of Culture
I, MOCIC ‘\
I \
I” \‘\
,I' ‘x Ideology:
’1’ Language “\ CODICXI
,’ \\ Of
1’ Values
,’ Field Tenor \ and
’ \
I’ ‘, Beliefs

 

 

 

 

 

 

Figure 3. The Relationship of Register to Ideology

 

20

   
 
   
 

Language

 

 

Ideology

 

 

 

 

Figure 4. The lntemal Location of Ideology

 

the overarching ideologies of a culture at large (as shown in Figure 3) tend to perpetuate
themselves. Ideological change is possible; however people tend to use ideology to
maintain given norms and values.

The perpetuation of cultural capital begins with the fact that we give value to
social symbols. Pierre Bourdieu (I983, 1991) used the example of the marketplace to
describe what he calls “cultural capital”. For example, in the marketplace, a Mercedes is
worth more than a Chevy, and a pair of Armani jeans is worth more than a pair of Levi’s.
This evaluation extends beyond the marketplace to the realm of culture in general and to
language in particular. It is not merely the objects in the marketplace to which we assign
value. All of the cultural products to which we assign value are part of a system of

cultural capital. Language is one of these cultural products.

2]

Language, as a symbol system, has social value in and of itself and furthermore,
symbolizes the social value of the speaker. Language is a tool that reflects the ideological
predisposition and other social assets of the speaker. This notion is termed “linguistic
capital” (Bourdieu, 1977).

Register as Linguistic Capital

Bourdieu (1977) argued that linguistic exchanges are a symbolic marketplace in
which claims to power are negotiated. The speaker, in the act of speaking, tries to sell a
bid for social authority and, in turn, the listener decides to what degree he/she buys that
claim. Bourdieu introduced the specific concept of linguistic capital to describe the
authority associated with the power possessed by a speaker. As with other forms of
cultural capital, linguistic capital designates access to resources. These resources afford
the speaker certain privileges. For example, those with high linguistic capital speak with
"command," i.e., the power to influence a listener toward the desired interpretation.
Furthermore, utterances are always ventured in particular situations (or markets), in
which certain external, social expectations for speech and interaction are imposed upon
conversational participants.

In the case of the present research, the particular situation or market, is envisioned
as the public school context. The linguistic market is a "system of relations of force
which impose themselves as a system of specific sanctions and censorship, and thereby
help fashion linguistic production by determining the 'price' [or value] of linguistic
products" (Bourdieu & Wacquant, 1992, p.145). Within the school system, the valued
language variety is Standard Mainstream American English (SMAE). Those with access

to this variety benefit from the capital or potential privileges it possesses. Those without

22

access or with less access to the variety are excluded from the benefits of this linguistic
form of capital.

Bourdieu (Bourdieu & Wacquant, 1992) argued that access to legitimate language
(i.e., the formal language of the school environment) is not equal, and that linguistic and
general cultural competence are monopolized by some. This inequality is solidified in
academic institutions, which provide advantages to those from cultural environments in
which generations of people have used the preferred language forms. This outlook
suggests that access to the discourse and discursive practices of formal schooling (i.e.,
academic discourse) is differentially accessible (Lemke, 1990). For those students who
enter the classroom with competence in its discursive practices, access to the academic
institution is facilitated. In other words, access to academia is easier for those with more
than less access to the academic register. At the same time, students with access to
SMAE are more likely to be perceived as successful students based on the teacher's
judgment of their ability. Within the overall school context, language background and/or
the ability to switch into the language of instruction is linguistic capital that can be
converted to academic reward.

Those who are less socially favored, use the less favored language variety, (e.g.,
AAE) and will continue to be judged less favorably than those who do not. That is, those
with access to the standard variety, SMAE, use their privilege to restrict those without
access to the privileged linguistic system from social activities (jobs, schooling and so
on) that require its use. These social activities entail general cultural capital, to which
access is restricted by use of the privileged language variety. The overall control of

resources afforded by those with cultural capital remains skewed. Those with control of

23

resources, including linguistic resources, tend to retain control at the expense of
excluding others with less access to those resources.

Clearly it is not merely the formal language variety that gives one linguistic
capital within the school setting. It is also how one uses the structural features of a
language system appropriately and powerfully within that situation of use. The
knowledge, ideologies, presuppositions, beliefs, and values must be considered a part of a
person’s linguistic capital. These underlying values and beliefs and their linguistic
reflections are parallel to what has been defined as “register” in this dissertation research.
Consequently, knowledge of and the ability to use registers (viz., school register,
teacher register and peer register) were considered as elements of linguistic capital,
which vary according to one ’s access to other sorts of cultural capital.

Because meaning is culturally determined and institutionally embedded, not
everyone is willing or able to uncover these patterns, which are typically hidden by
unexamined ideology. This is likely the case because of the way that people use power.
Those in power tend to benefit from the reproduction of power differences. In the case of
academic institutions, it is well known that there is “unequal distribution of social power
and knowledge of teachers as compared to students. . .throughout the world” (Homberger
& Chick, 2001, p. 34).

Even though teachers may be uncomfortable or unwilling to make their position
of higher status overt, examination of underlying meanings and practices in the social and
academic realms is necessary in order to empower students generally (Ayers, 2004) and
those from non-mainstream backgrounds, in particular. Non-mainstream children need to

learn not only the mainstream, formal, linguistic code but also the social meanings in

24

which language form is embedded. This means making overt those dynamics such as
power, status, privilege, evaluation and devaluation that occur within and around the
communication process (Delpit, I995). Delpit summarized her point of view by stating
that,

“. ..students must be [explicitly] taught the codes needed to participate fiilly in the

mainstream of American life, not by being forced to attend to hollow, inane,

decontextualized subskills, but rather within the context of meaningful

communicative endeavors. ..[T]hey must be allowed the resource of the teacher’s

expert knowledge, while being helped to acknowledge their own expertness as

well; and that even while students are assisted in learning the culture of power,

they must also be helped to learn about the arbitrariness of those codes and about

the power relationships they represent.” (p. 44)
Since language is the main tool by which these social actions occur, we must examine
language structure but not limit our inquiry to structure per se, in order to more
effectively educate non-mainstream students. We must also teach our students to
investigate the social norms of the home and school communities including the socially
determined norms of meaning that are relevant to a given community. Register-shifting
skill is a fundamental element within this suggested, pedagogical space. In sum, in order
to teach our students SMAE we must teach not only structural knowledge but also
functional skills such as the specific skill involved in shifting registers.

Statement of the Problem for the Current Study

Given that register-shifting skill is essential to the academic and life success of
children, it is amazing that we know so little about the manner in which this skill
develops (Andersen, 1990). The development of register-shifting ability is an

understudied area in general and in African American English speakers, as a case in

point.

25

Basic research is needed about how African American children shift registers.
More generally, we need a broader framework for investigating register use. A well—
developed theoretical framework is especially important within a multicultural context in
which different issues of social adaptation arise. In the case of switching from peer-
centered to a teacher-centered register in the school context, the Standard Mainstream
American English, or SMAE, of a given region is usually expected. Given that all
children in the United States (US) are legally obligated to go to school, exposure to
SMAE is unavoidable. Thus, linguistic demands will be placed on minority language
speakers who must adapt their home language to a greater extent than mainstream
language speakers. For SMAE speakers the task is to switch from the casual to formal
variety within the same dialect, whereas for Afiican American English speakers the task
is to switch from the casual to formal variety and from one variety to another (AAE to
SMAE). This dissertation research focused on AAE speakers because of the potential
impact that casual to formal register shift may have on their education.

AAE, the variety that many African American children bring to school, possesses
systematic, rule-govemed patterns that differ fiom SMAE in every domain of language,
i.e., phonology, semantics, syntax and pragmatics. Examples of AAE phonological
patterns include 1. postvocalic consonant deletion/reduction (e.g., “fire-)fi, dial-)dad”),
2. “-g dropping” (or replacement of / a / with / n /): (e.g., “waitin’, walkin’ ”) and 3. th
substitution (voiceless syllable finally and voiced syllable initially) (e. g., with-)wif,
birthday-9birfday, this-)dis) (Labov, 1972a; Pollock, 1998; Stockman, 1996c;
Stockman, 2007; Wolfram, 1969). Some AAE syntactic patterns are: 1. zero copula or

auxiliary in the present tense (e. g., “He running.” and “They cold.” but not *“They cold”

26

meaning “They were cold”) and 2. uninflected 3rd person, singular, progressive —s in
order to indicate habitual action (e. g., “He walk to the store everyday.” and “My momma
like to read,” (Dillard, 1972; Green, 2002; Rickford, 1999; Rickford & Rickford, 2000).
Some AAE lexical (semantic) items are kitchen (meaning “the back of the neck where
hair grows short and curly”), money (meaning “friend”), tote (meaning “to carry
something heavy”) (Dillard, I977; Smitherman, 2000; Turner, 1949). Pragmatic devices
characteristic of AAE include verbal strategies such as “signifying” (Morgan, 2002): “a
verbal art of insult in which a speaker humorously puts down, talks about, or

needles. . .the listener” (Smitherman, 1977, pl 18).

Because these features and meanings of AAE contrast with those of SMAE (the
language of school instruction), AAE speaking students may be less comprehensible to
teachers and/or simply less valued by them. In this sense, decreased access to the
knowledge of the standard language code and the meanings it entails, may directly impact
the academic success of AAE speaking children.

Linguistic Context: Language as a Proxy for School Underachievement

The linguistic differences between AAE and SMAE have been well documented
by linguists over the past 40-50 years. Furthermore, the difficulty inherent in register
shifting, while simultaneously shifting from one linguistic code to another, has led some
to suggest that language difference is the reason for the reduced academic achievement of
African American children (Baratz, 1969; Charity, 2004; Labov, 1995).

According to the National Assessment of Educational Progress (NAEP,
Commonly called the “Nation’s Report Card”) on the educational attainment of US.

Children from 1971-1999, African American children seem, on the surface, to have made

27

gains toward closing the performance gap that separates them from the White majority.
However, if one investigates the data more closely, most of the reduction in educational
gaps between Whites and African Americans, occurred before 1986. Since then, the gaps
have either widened or remained thesame (Craig & Washington, 2004b; Nettles, 2000).
Throughout the late-19805 and the 19905 the average scores of Afiican American
students have remained well below those of Whites. For example, at age 17 the reading
achievement of Afiican American students was lower in 1999 than it was in 1988, a
reversal of the gains made over the previous two decades.

This racial comparison must be tempered, however, by the differences between
African American and White children relative to basic as compared to advanced reading
skills. At the basic skill level, of finding facts in simple stories and drawing inferences
based on short passages, the gap between Whites and African Americans has narrowed
dramatically; 95% of African American l7-year-olds reach this basic level, which is
higher than their 82% in I971, as compared to 98% of Whites in both years. But at more
advanced levels, a tremendous gap remains. At the level of understanding and analyzing
relatively complicated information, Whites have moved up only 3 percentage points to
46% over the three decades. African Americans have increased from 8% to 17%, which
is a noticeable gain, but is inconsequential given the 29 percentage point disparity that
remains between Whites and African Americans.

The gaps in educational attainment between Whites and African Americans
parallel the disparities in the numbers enrolled in high school honors and college

preparatory courses as well as in scores on college entrance exams (Nettles, 2000). These

28

disparities have a remarkable impact on potential college attendance and the subsequent
employment attainable with such an education.

The possibility that the language itself is the reason for less than optimal
academic performance, has receivedenough clout to set legal precedent in the Ann Arbor
Parents vs. Martin Luther King, Jr. Elementary (i.e., the “Ann Arbor”) case in which the
school district was mandated to take AAE-speaking children’s language into account
when teaching SMAE reading skills. This same notion was at the heart of the Oakland
(California) Unified School District’s proposal to teach their students standard English by
bridging the gap between their home and school varieties. Each of the two
aforementioned cases was predicated on the underlying assumption that the
disproportionate, academic underachievement of African American children is based on
the failure of schools to take their language background into account in the educational
process.

Exrtralinguistic Context: Social Class as a Proxy for School Underachievement

Language is not the only proposed reason for reduced school success among
African American children. Scholars have pointed to the extralinguistic, social context
for which socio-economic status (SES) has served as a proxy. SES is commonly
measured in terms of parental income, education, and/or occupation. Low SES has been
associated generally with school underachievement (Hart & Risley, 1995; Lareau, 2003).
Lower Class

Head Start, a program stemming from Civil Rights legislation, was an
instrumental piece of US. president Lyndon B. Johnson’s War on Poverty. Head Start

Was developed as a solution to a broad range of social, cognitive and health disparities

29

among pre-school children in the US. The implicit, founding argument of Head Start was
that being poor deprived one of educational opportunity and something needed to be done
to address this negative consequence.

However, recent data show that whatever gains African American children make
by participating in Head Start are lost by the time they reach eight years of age,
approximately their second to third-grade year. Currie and Thomas (1995) found that
Head Start is associated with large and significant gains in preventive health services, as
well as mathematics, reading and vocabulary scores among both Whites and Afi'ican-
Americans. However, among African-Americans, most of these gains are lost quickly. In
fact, greater access to preventive health services, is the only area addressed by Head Start
in which both Whites and African Americans benefit equally. By the time Afiican
American children are eight years-old, there is no difference between those who attended
Head Start and those who did not on measures of math, reading or vocabulary test scores.
Furthermore, Head Start significantly reduces the probability that a White child will
repeat a grade but it has no effect on grade repetition for Afiican-American children.

Suggested reasons for this racial disparity are 1) heterogeneity in program content
and 2) variance in program benefits depending upon the child’s background inclusive of
home resources and the type and quality of school attended. The first of these two
explanations suggest that the large variation in methods with which Head Start program
are implemented may affect outcomes for African American children. In particular, it was
suggested that programs that are attended by African American children tend to focus
more on factors affecting health than socialization and cognition (Currie & Thomas,

1 995). The second explanation suggests that the relative poverty and low level of

30

educational resources in African American communities oppose the gains made in Head
Start. In addition to its association with school underachievement, low SES tends to be
associated linguistically with greater use of local, vernacular than standard, prestige
language features. However, the correlation between language use and school
underachievement should not be used to infer that low SES is the only factor to influence
AAE speaking children’s access to the language of instruction.

The educational disparity between African American and White American
children may also be due to teachers’ racial biases, a factor which has been given
relatively little attention in the language variation literature. In general it has been shown
that racism negatively affects education (King, 1991; Sedlacek & Brooks, 1977; Troyna
& Hatcher, 1992) and that programs designed to decrease racism have positive effects on
educational outcomes (Gregory, 2000; Lawrence & Tatum, 1997).

Middle class

African American children’s decreased academic success extends beyond lower
social class status. Middle class African American children also tend to underachieve
academically. According to a special tabulation of the long-term, trends data from the
The Nation's Report Card (NAEP), the achievement gaps are just as wide or even slightly
wider between African American and White students with college-educated parents as
they are for students whose parents have much less education (Nettles, 2000). A
summary of current research reveals that even with adequate socioeconomic resources,
African American children continue to show underachievement tendencies (Jencks &

Phillips, 1998; Spencer, 2000; Steele, 1992). “Something else [besides SES] has to be

3]

involved...” (Steele, 1992, p.70). If so, then SES cannot be the sole answer to the
educational underachievement of African American children.

The Need to Revisit Language and Social Context as Proxies for School Achievement

The rationale for the present Study is guided by the assumption that both
language and social context have been viewed far too narrowly to account for school
achievement. Before considering decreased academic achievement directly, we need
more basic research that expands how we study both language and social class as
mitigating causes of school underachievement. Two main issues warrant consideration.
The first one relates to the way that language is studied. In particular, language as a
proxy for school underachievement is often studied structurally (i.e., with a focus on
grammar and/or phonology) rather than functionally (i.e., with a focus on pragmatic
factors). The second one has to do with isolating the study of social class from racial
segregation.
An Expanded View of Language
Past research has traditionally focused on language simplistically as grammatical

and/or phonological structure. In fact, well-known educational policies such as those of
the Oakland Unified School District and the Ann Arbor Unified School District have
been informed by structural differences between home and school languages. In order to
remediate the problem of the difference between home and school language, policies
were enacted to teach children the structure of the SMAE language code. In contrast, the
present study will investigate language not simply as decontextualized, structural,
grammatical or phonological form but instead as culturally constructed meaning that

is used within a specific context.

32

The present research hypothesizes that the interaction between language and the
academic achievement of African American children is much more complicated than has
been previously assumed. While it is fundamentally acknowledged that access to the
language of instruction (SMAE) is acontributing factor, the present research holds that
difficulty with access extends far beyond the SMAE code itself. There is also a problem
with Afiican American children’s access to the culturally-embedded meanings of the
school environment that are conveyed by SMAE. In other words, there is a specific
linguistic mismatch between home and school registers. Only a registral view of
language, which embeds language form with social meaning, can account for the full
extent of the linguistic mismatch between African American children and their schools.
Furthermore, the nature of the social contexts in which registers are embedded requires
systematic investigation, to the end that the community variables relevant to the
acquisition of school register can be identified.

We need a way to effectively investigate children’s understanding and use of
register shift as evidence of their knowledge of the socially embedded meanings that such
shifts convey. Registral features cannot be studied in the absolute. That is, they are by
definition context-dependent. In general, the researcher can only attach knowledge to an
observable state. This is commonly done in studies of language development and
disorders by observing receptive or expressive language behaviors. To study registral
features, we must embed feature use within a meaningful context because knowledge is
always context-dependent. For this reason, the present research assumes that its chosen,
peer-centered and teacher centered school contexts give distinct meaning to the linguistic

Structures used therein.

33

There are two particular linguistic areas that lend themselves to the investigation
of register shifting: 1) lexical semantics, which may reveal what a child knows about
culturally appropriate word choice differences across interlocutors, and 2) request forms,
which may reveal what a child knows about the relative deference between interlocutors.
These language structures can be investigated by observing how African American
children shift registers across peer and teacher-centered contexts, the two culturally
embedded situations of interest in the present research. Changes in use of linguistic
structures across meaningful contexts will reveal children’s knowledge of how to
appropriately modify meanings through register shifting in the school setting.

An Expanded View of Social Context

In addition to the modified definition of language, viewed as register, in the above
discussion, the current research was also driven by the assumption that the view of social
context should be modified. In particular, SES must be abandoned as a lone proxy for
language in the context of school performance. Instead, we should embrace the view that
SES interacts with other variables to influence children’s access to SMAE. I propose that
the extralinguistic variables of racial-residential segregation (SE G) in combination with
SES interact to influence SAME access. Both variables in combination may impact such
access in a way that neither variable alone can do.

This assumption is supported by a recent study (Orfield & Lee, 2005). It revealed
that the combination of poverty and racial segregation negatively affected various
measures of school success such as teacher quality, test scores and drop-out rate. What
remains unclear is whether African American students from communities with different

constellations of SES and segregation possess similar levels of access to SMAE, the

34

language of school instruction. When combined with SES, the additional variable, SEG,
creates different community constellations. In this manner, specifically defined social
subgroups are expected to differ in their levels of linguistic capital. It is within these
carefully defined social contexts that register-shifting knowledge might be differentially
used. This combination of the variables of SES and SEG allowed the present research to
focus on children from schools within two high segregation neighborhoods, one that was
upper-middle class and one that was impoverished, and schools within two low
segregated neighborhoods, one that was upper-middle class and one that was
impoverished.

The four resulting groups arguably have different life experiences that may
differentially influence access to and use of the SMAE school register (i.e., the one that
matches the language of instruction). Social context is likely to influence their linguistic
capital in the form of access to SMAE school register. For example, two African
American children from an upper-middle class background, who live in neighborhoods
that differ in the amount of racial segregation are likely to have two different types of life
experience due to both: (1) their resources that SES level provides and (2) their
opportunities to and/or restrictions on networking with people, which are imposed by
SEG level. Similarly, a comparable pair of children from low SES backgrounds will have
differences in resources depending upon whether their neighborhood is segregated or not.
As such examples illustrate, an analysis that incorporates both SES and SEG may provide
a more comprehensive explanation of the social context factors that are relevant to

accessing mainstream language varieties than either variable investigated in isolation.

35

In this study I investigated the specific effect of social context, as broadly defined
by the intersection of SES and SEG, on African American children’s access to,
evaluation and exploitation of the linguistic capital afforded by SMAE as reflected in
their variable use of peer-centered to teacher-centered register shift.

Purpose of Study
This study aims to determine if and how linguistic register shifting among African
American children is related to membership in a larger social community as defined
by the combined characteristics of SES and segregation. Specifically, the following
questions are posed:
1. Do African American third graders demonstrate knowledge of peer-centered as
compared to teacher-centered register shift?

2. Is register shift related to membership in a larger community of language users as
distinguished by the combined effects of segregation and SES?

3. Within each community type, is register shift related to (a) linguistic task
conditions, i.e., lexical semantics or request forms and/or (b) type of performance

modality, i.e., the receptive or expressive mode?

4. Is children’s ability to express, metalinguistic knowledge about register-shifting
skill related to their display of register shifting on experimental tasks?

36

CHAPTER 2
METHOD
Participants

Selection Criteria

Participants were 36, typically developing, African American third graders
between 7; 6 and 9; 9 years. The focus was on third-graders for two primary reasons.
First, current evidence suggests that gains made by African American preschoolers tend
to be lost by 3rd grade (Craig & Washington 2004b; Nettles, 2000). This seems to be a
critical grade level at which the academic preparedness of African American children is
crucial. By 3rd grade, all children, including African Americans, have also gained
knowledge about politeness of request forms, a skill central to the methodology
underlying this study (Ervin-Tripp, 1976, 1977; Mitchell-Keman & Keman, 1977).
Selection was based on participants’ public school attendance within specific
neighborhoods.

Neighborhoods were operationalized in terms of census tracts, the method that
best captures the geography of true neighborhoods. Census tracts are identified generally
as small and relatively permanent subdivisions of a county defined by the US. Census
Bureau. Most census tracts contain fewer than 8,000 residents. They represent

“neighborhoods” in that they often contain homogeneous population characteristics,
inclusive of economic status and living conditions.

Census data were used to characterize each neighborhood in terms of different

levels (High and Low) of racial-residential segregation (SEG) and socio-economic status

37

(SES). These characteristics resulted in four neighborhoods, which are labeled as follows:
HiSSLoSG or High Social Status/Low Segregation, LoSSLoSG or Low Social
Status/Low Segregation, LoSSHiSG or Low Social Status/High Segregation, and
l-IiSSHiSG or High Social Status/High Segregation. A cohort of three students from the
HiSS/HiSG group was not included in the main analyses but was superimposed in most
tables and figures in the results section in order to view the potential relationship of this
relatively underrepresented group to the other community groups in the study. The
addition of these three students to the participant pool accounts for the grand total number
of 36 participants used.

It was desired that each participant would be randomly selected from a gender-
stratified sample in order to include a balanced representation of boys and girls. Due to
difficulty with participant recruitment, this goal was not obtained. However, relatively
gender-balanced samples were obtained in all community types except the LoSS/LoSG
(N = 11) group in which most (11 = 8) volunteers were girls. The HiSS/LoSG sample (N =
10) was composed of four (40%) girls and six (60%) boys and the LoSS/HiSG sample
was composed of seven (58.33%) girls and five (41.66%) boys. Table 1 provides a
detailed description of participant characteristics.

Children from neighborhoods and/or schools that did not meet the selection
criteria were excluded from the study. Table 2 specifies the other inclusionary/
exclusionary criteria used to select participants.

Selection Procedures
Global Screening Procedures

An initial screening was conducted in order to identify potential participants by

38

Table 1

Participant Characteristics

 

 

Participant Grade Age Sex Community Hearing F RPLc DELV
Number Type3 Screenb Screend
1 3 8,6 M HiSS/LoSG P N P
2 3 8,4 F HiSS/LoSG P N P
3 3 9,2 M HiSS/LoSG P N P
4 3 8,11 F HiSS/LoSG P N P
5 3 8,7 M HiSS/LoSG P N P
6 3 8,1 M HiSS/LoSG P N P
7 3 8,7 F HiSS/LoSG 1:c N P
8 3 8,3 M HiSS/LoSG P N Pi
9 3 8,6 M HiSS/LoSG P N P
10 3 8,1 F HiSS/LoSG P N P
1 1 3 8,10 M LoSS/LoSG P N/Ag Pi
12 3 9,11 M LoSS/LoSG P N/Ag Pi
13 3 9,3 F LoSS/LoSG P N/Ag P
14 3 8,6 F LoSS/LoSG P N/Ag P
15 3 8,4 F LoSS/LoSG P N/Ag P
l6 3 9,5 F LoSS/LoSG P N/Ag P
l7 3 8,6 F LoSS/LoSG Pf N/Ag P
18 3 9,2 F LoSS/LoSG P N/Ag P
19 3 9,2 M LoSS/LoSG F° N/Ag P
20 3 7,7 F LoSS/LoSG P N/Ag P
M 3 9,1 F LoSS/LoSG P N/Ag P
22 3 9,0 M LoSS/HiSG P N/Ag P
23 3 8,4 F LoSS/HiSG P N/Ag P
24 3 8,9 F LoSS/HiSG F" N/Ag P
25 3 8,10 1: LoSS/HiSG P N/Ag P
26 3 9,12 M LoSS/HiSG P N/Ag P
27 3 9,9 F LoSS/HiSG P N/Ag P
28 3 8,8 F LoSS/HiSG P N/Ag P

39

Table 1 (cont)

29 3 8,8 M LoSS/HiSG P N/Ag P“
30 3 8,1 1 F LoSS/HiSG P N/Ag P
31 3 9,0 M . LoSS/HiSG P N/Ag P
32 3 8,8 F LoSS/IIiSG P N/Ag P
33 3 9,7 M LoSS/l-liSG P N/Ag P
34 3 9,2 F HiSS/HiSG P N P
35 3 9,0 F HiSS/HiSG P N P
36 3 9,1 M HiSS/HiSG P N P

 

aHiSS/LOSG = High Social Status/Low Segregation, LoSS/LoSG = Low Social
Status/Low Segregation, LoSS/HiSG = Low Social Status/High Segregation, HiSS/HiSG
= High Social Status/High Segregation. b P = Pass, F = Fail; hearing screenings were
generally passed at 20 dB for the frequencies 1000, 2000 and 4000 Hz and failed
otherwise. ° F RPL = indication of eligibility for “free or reduced price lunch”. d DELV
screening: P is indicative of “lowest” or “low to medium” disorder risk; F is indicative of
“medium to high” disorder risk. ° The participant passed hearing screening for all
frequency except one, which was passed at 25 dB. f The participant passed the screening
at 20 dB at all frequencies except 1000 Hz which was passed at 45 dB in the right ear and
55 dB in the left ear. 3 N/A = specific information not available due to school district
policy regarding privacy. h The participant was categorized as medium to high risk for
disorder; due to difficulty recruiting subjects participant was included in the main data
analyses (as the other non-optimal participants) then scores were compared to central
tendencies to investigate status as an outlier. ' The participant had reported teacher or
caregiver concerns about speech fluency.

40

Table 2

Inclusionary and Exclusionary Selection Criteria

 

Inelusionary Criteria

Exclusionary Criteria

 

Residence

School attendance

Race
Ethnicity/Nationality
Consistent Community
Membership

Language Development
Hearing Ability

A resident of one of the

neighborhoods that met the

SES/SEG characteristics
defined by this study

An attendee of one of the
neighborhood elementary
schools defined by this
study, each of which
matched the SES/SEG
characteristics of the
surrounding neighborhood.
Black or African American

United States citizens with

ancestral ties to the African

slave trade
Residence in targeted

community for the previous

five or more years
Normal
Normal

A resident of a
neighborhood that did not
meet the SES/SEG
characteristics defined by
this study
An attendee of an
elementary school not
defined by this study.

Biracial; Not Black or
African American
Child of African, Caribbean
or other immigrant to the
US.

Residence outside of
targeted community within
the previous five years
Impaired
Impaired

Note. Given the difficulties with participant recruitment for this study, the eight children
Who met all selection criteria, except one, as identified above in Table l, were still
included in the participant pool. They were distributed across the three community
groups compared (two each in the LoSS/HiSG group and the HiSS/LoSG group, and four
in the LoSS/LoSG). However, none of their individual performance significantly
depressed the community-group outcomes; so they were included as participants after it
Was shown that none of their scores on any task fell below the normal range of scores for
participants who met all criteria. That is, their scores were higher than 2SDs above the
mean. This level was chosen because it is the common cut-off for normalcy in the field

Of Speech-Language.

4l

having classroom teachers complete screening questionnaires, which identified children
who potentially met the selection criteria listed in Table 2. See Appendix E for the
ABILITIES index (Simeonsson & Bailey, 1984) the global screening tool used in this
study. This index screens for the follbwing abilities: audition, behavior and social skills,
intellectual function, limb movement, intentional communication, tonicity of muscles,
integrity of physical health, eye health and structural status of the body. Potential
participants continued to be considered for inclusion if they received high marks (i.e.,
scores of 1 or 2) on the screener in the areas of audition and intentional communication.
To insure confidentiality at the time of screening, the names of potential participants
remained unknown to the researcher until after permission to participate was granted.
After screening, either the principal of a school, or his/her representative sent permission
forms to the parents of each potential participant. When permission forms were returned,
the participants were assigned a number, and placed on a master list, which was kept in a
locked file cabinet, in the secured office of the primary investigator. After global
screening was completed, the potential participants underwent more specific screening to
verify their appropriateness for the study.
Language and Hearing Screening Procedures
Potential participants were screened in order to verify their normal hearing and
language status. Both language and hearing screenings were conducted by undergraduate
research assistants in Speech-Language Pathology who were blind to the goals of the
Study. They conducted screening under the supervision of the primary investigator, an
ASHA-certified speech-language pathologist. Those who passed these hearing and

language screenings were considered appropriate candidates to participate in the study.

42

Language screening procedures. Participants were screened for normal language
skills using the Diagnostic Evaluation of Language Variation (DEL V), Screening T 001
(Seymour, Roeper, & de Villiers, 2003). Only those with “the lowest risk and low-
medium risk” for language disorder (as determined by scores on the DEL V-Screening
Tool) were considered as optimal candidates for this study. See Table 1, note d.

A semantics screener was devised by the investigator and given to all participants
to ensure their familiarity with the lexical items to be administered within the lexical
semantics task. In order that the participants were not primed to recognize the task items,
the screener contained the ten lexical semantics task items and ten items that were not
actual task items. The ten items that were not part of the actual task were taken from 6-
0/6-11 age level of the Expressive One Word Picture Vocabulary T est-R (EOWPVY)
(Gardner, 1990) and were considered to be among the words likely to be known by the
potential third-grade participants. All twenty lexical items were listed and then
randomized (Urbaniak, 1997-2005) as shown in column 5 of Appendix H. If participants
did not know target lexical items, this fact was taken into account in interpreting data.
See Appendix G for Lexical Semantics Screener. The number of lexical items known by
participants averaged 9 of the 10 total items. Thus, less than one item per child was
unknown. The least recognizable item was “tote,” which was unknown to half (18 of 36)
of the participants.

Hearing screening procedures. Standard hearing screenings were guided by
American Speech-Language-Hearing Assocation (ASHA) procedures (2002). Pure tones
were delivered with a portable, Bell Tone model # 120 audiometer to each ear separately

at 1000, 2000, and 4000Hz at 20dB HL. Children raised a hand to indicate perception of

43

each tone delivered in each car at the designated amplitude needed to pass the hearing
screening. See Table 1, note I”. Three participants failed the hearing screening but were
still included in the study See Table 1, notes ° and f.
Operationalization of the Dependent Variables
One dependent variable, register shift, was operationalized as a change in the way
one meaningfully speaks in a peer-centered compared to a teacher-centered context. The
second dependent variable was operationalized as the difference in community type. The
research question was, “Is extent of register shifting related to membership in a larger
community?”
Register Shift between Peer-centered and T eacher-centered Contexts
A basic assumption regarding the first dependent variable is that differences in

interlocutors create context differences. In the current study, there were two distinct
interlocutors with whom each participant interacted: a White teacher and an Afiican
American peer. These differences in interlocutors create distinct contexts in which there
were predicted to be at least two main differences in registral meaning: (1) a difference
in role/status relationships and (2) a difference in available/appropriate conversational
topics. The role and status difference involved between the participants and the teacher
were expected be relatively large due to assumed discrepancies in age, knowledge, skill,
and authority levels. In contrast the role and status difference between the participants
and the peer were expected be relatively small, due to similarities along the same
continua mentioned above: age, knowledge, skill and authority levels. Differences in role
and status were expected to be signaled by differences in registral tenor, specifically, the

use of politeness forms across interlocutors.

44

An additional difference was the race of the teacher compared to the peer. That is,
peer puppets were similar in racial/ethnic background (i.e., the participants and peers
puppets were both African American) and could be assumed by the children to share the
African American cultural experience. The teacher puppet was White and could be
assumed by the children to share a different cultural experience from them. The shared
cultural experience of the peer-centered context was expected to give cues to the
participants about the available and appropriate topics of the registral field, as signaled by
lexical semantic use, to be discussed with their peers.

This methodological approach reflects the theory of Systemic Functional
linguistics (Halliday, 1970). In particular, registral theory informed the specific instances
in which speakers differentiated meaning given the contexts of use. Given that changes in
context modify the array of available meanings from which a competent language
speaker has to choose, this study focused on two contextual differences. In particular, the
status/role differences created by the dependent variable were expected to create
differences in registral tenor. The racial/ethnic difference in interlocutor was expected to
result in differences in registral field.

Each child engaged in four language tasks in which he/she had the opportunity to
shift between a peer-centered context and a teacher-centered context. These two distinct
contexts allowed observation of the participants’ “register-shifting” skill. Register shift
was calculated by determining each individual’s percent change, or Percent Register Shift
(PRS), in use of lexical items and percent change (PRS) in use of request forms. The

average percent change was then compared across the various independent variables to

determine group differences.

45

Self-reflective Report of Register Use

Interview data were used to provide individual perspectives through participants’
reports of their own awareness of their use of register-shifting skill. The level of self-
reflection was measured by the amount and type of verbal responses to pre-selected
questions about registral use. See the self-reflective report of register use questionnaire
in Appendix L.

The assumption that guided the selection of this variable was that the ability to
talk about one’s own register shifting reflects the extent to which one applies register-
shifting skill on a conscious level. Therefore, there may be a difference between what
people report directly about their values and beliefs as compared to their actual
performance on registral tasks, which presumably reflects underlying values and beliefs.
The extent to which they can report on their own register use offers insight into the
conscious availability of their registral knowledge. It was assumed that conscious
availability is likely correlated with environmental demands on using different registral
principles. A strong contrast in life experiences between the home and school
environments, was expected to result in evidence of registral shifting skill both in
language tasks and in the reflective self-report. In contrast, if there were not strong
Contrasts in life experience between home and school environments, there would likely
be less evidence of knowledge of registral shifts across language and self-report tasks.

Because of differences in the types of life experiences, certain community types
Were expected have more conscious use of register—shifting knowledge than others. Those
in the middle groups (HiSS/HiSG and LoSS/LoSG), with life experiences that expose

them to more contrasting experiences, were expected to show more awareness of their

46

own shift than other groups, viz., HiSS/LoSG and LoSS/HiSG neighborhoods, with
smaller home and school experiential differences.
Operationalization of the Independent Variables

In the current study, there were two types of language task variables: one related
to registral tenor and the other related to registral field. Each of these task variables were
investigated in both the receptive and expressive performance modalities. Finally, the
group variable, “community type” was operationalized in terms of SES and segregation
levels.

Language Task Variables

Two language tasks were administered to each participant: lexical semantics and

request forms.

Lexical Semantics Task

The nature and type of content words (nouns, verbs, prepositions, and so on) used
within a text reflect the aspect of register, which is called field in systemic functional
theory. The assumption of the present study is that the Afiican American experience
provides a constellation of words with culture-specific meanings. These words convey
topics that presumably are available to children because of their cultural experiences.
Neither these Afiican American experiences nor their corresponding lexical labels apply
t0 mainstream speakers. Those with access to SMAE may have an awareness of the
appropriate use of these items within an intracultural context (i.e., when speaking with
Someone who is also African American) but inappropriate use within an intercultural

Context (i.e., when speaking with someone who is not African American).

47

This knowledge of appropriate use relates to registral field. Thus, the use of AAE
or SMAE lexical semantics will reflect speakers’ conceptualization of field, that is,
whether SMAE or AAE items seem appropriate to use with either a White school teacher
or an African American child.

The words of Afiican American English are more than just slang (Smitherman,
2000). There is a stable set of lexical items that is “familiar to and/or used by all groups
in the Black community” (Smitherman, 2000, p. 2). Some of these lexical items show
evidence of West African and/or creole roots while others are English words that have
been borrowed in order to encapsulate a distinctively Afiican American meaning. Two
examples of such borrowed words are: (1) kitchen, (meaning the place on the back of the
neck where hair grows short and curly and, (2) bright (meaning light skinned). So, even
though the words are homophonous with SMAE words, they have very distinct, culture-
specific meanings; that is, each one has a distinctly, African American meaning that has
been encapsulated by a distinctly African American word. Kitchen refers to a place on
the neck that is experienced only by people whose hair grows in small, tight curls and
bright refers to lightness of skin-color only in a community with many variations of skin

color.

One question addressed by the present research was whether AAE/SMAE lexical
items are used as markers of peer-teacher register shift across children from communities
With different constellations of SES and segregation (SEG) levels. Current research
(Smitherman, 2000) acknowledges that some AAE lexical items are context—specific

(e-g., church talk, Hip Hop talk, women’s language); so we may expect a general peer to

48

teacher shift. It is not yet known whether there is a difference in the extent of this shift

across people from different community types.

Request Task

The level of directness of request forms reflects speakers’ conceptualization of the

aspect of register, which is called tenor. Tenor relates to the role relationships shared by

participants and their respective statuses. In this case, tenor is reflected in the directness

of request forms used, as indicated by the number and type of modals used in making

requests. The classification scheme used by Ervin-Tripp (1976, I977) identifies six basic

imperative types:

1.

2.

3.

4.

Personal need/desire statements [e.g., I want x.]

Imperatives [e.g., Give me a x.]

Imbedded imperatives [e.g., Could you x, would you x.]

Permission directives [e.g., May I x, Can I x.]

Question directives [e.g., Do you have the time? What are you doing?]
Hints [e.g., It’s hot out here. I’m the teacher]

These imperative types become increasingly more complex and as a result occur

generally in the developmental order shown above. (Ervin-Tripp, 1976, 1977; Ervin-

TI‘ipp & Gordon, 1986) The current research regarded personal need/desire statements

and imperatives (1 and 2 above) as direct requests. All others were considered indirect

r equests.

Performance Modality Type

Each language task described above was observed in the receptive and expressive

performance modalities. The assumption was that children can know something,

49

recognize it and understand it receptively. But this does not mean that they can express
those meanings. Expression is more complex in that it involves both understanding and
using particular forms to articulate that receptive understanding. Given the complexity of
the relationship between the expressive and receptive modalities, it is not surprising the
receptive language develops before expressive language in both first and second language
acquisition. Consequently, there may be differences between receptive and expressive
knowledge of register shifting across the research participants in the present study. For
example, one group might perform similarly in the receptive and expressive modalities
while another group might show registral shifts in one performance modality but not the
other.
Community Type

The group variable used in the current study was “community type”, which was
distinguished by characteristics of SES and SEG. SES has been shown to correspond with
vernacular language use and was expected to influence the manner in which children shift
registers. SES was operationalized by using the most common indicators (viz., income
level and education level). Racial-residential segregation and SES were both used to
Characterize neighborhoods. Recent studies have confirmed that it is the relationship
between SES and segregation that affects the life chances of people within given
Communities (Briggs, 2002). In this sense, communities that are defined by their SES and
SEG levels may yield a clearer picture of Afiican American children’s language use than

One which is defined by SES alone.

50

Community type was operationalized according to the combined SEG (high or
low) and SES (high or low) of neighborhood residences. Such an Operationalization
resulted in four experimental groups:

(I) High Social Status/High Segregation or HiSSHiSG,
(2) Low Social Status/High Segregation or LoSSHiSG,
(3) High Social Status/Low Segregation or HiSSLoSG,
(4) Low Social Status/Low Segregation or LoSSLoSG.

These four groups can be visualized, as illustrated in Figure 5.

 

 

 

 

 

 

HiSG 2 1

L080 4 3
m U)
U) U)
.3 :E

 

 

 

 

Figure 5. Community Type as Defined by High and Low SES and SEG

\

Note: HiSG = High Segregation, LoSG = Low Segregation, HiSS = High Socio-
Economic Status, LoSS = Low Socio-Economic Status.

51

Socioeconomic Status (SES)

SES was defined as either “High” or “Low” by using the 2000 Census data for
income (Median Household Income, MHI) and education (Percent population over 25
years-old with a Bachelor’s degree). In the year 2000, the Median Household Income
(MHI), the customary measure for central tendency regarding income, was $61, 905 in
Oakland County, MI. Given this figure, “high” income was defined as a multiple of the
median. Thurow (1985) defined middle class as those with incomes between 75 and
125% of the MHI. In this study, income level for the “high” SES cohorts was set at a
level above the uppermost end of Thurow’s range, that is, at a level 3 1.25 x MHI, or
$77, 384 (or greater) MHI. Pressman (2001) also used this formula to calculate “upper
class”, which was “greater than 1.25 MHI.” The income level for the “low” SES cohorts
was set at a level 5 1/3 the MHI or $20, 428 MHI. This level is well below the middle-
class cutoff designated by Thurow and was meant to capture the income characterized by
those living at or near poverty. In fact, to designate two, low SES groups for this study
(one labeled “high segregation” and one labeled “low segregation”) it was necessary to
slightly elevate the income level for the low SES group from 5 $20, 428 to _<_ $21, 900. At
this income level it was possible to locate two “low SES” neighborhoods that varied by
segregation level. The need for this adjustment was likely due to the general affluence of
Oakland County, which was one of the most affluent counties in the state of Michigan
according to the 2000 US. census. Its MHI of $61,905 compared to the Michigan, state-
Wide MHI of $44,667. Poverty levels for the county and state were 5.5% and 10.5%,

respectively; twice as high for the state as a whole than for Oakland County.

52

To designate SES, education level was considered in addition to income. Both are
commonly used SES indicators. For this study, education level was operationalized for
neighborhoods as a whole. The education level of the low SES groups was
operationalized as 15% or fewer residents over age 25 with a Bachelor’s degree or
higher. Education level of the high SES groups was operationalized as greater than 50%
of residents over age 25 with a Bachelor’s degree or higher. Low SES neighborhoods
were also characterized by poverty level of people living in the same neighborhood.
Poverty was defined operationally in this research as: greater than 20% of the population
of the neighborhood living in poverty. It should be noted that the traditional definition of
concentrated poverty or CF, which is greater than 40% living in poverty (Massey et al.,
1991), was not used due to Oakland county’s status as one of the wealthiest counties in
the state. In fact there were only two neighborhoods in Oakland that could be categorized
as possessing CP according to the traditional definition. Furthermore, neither of these
neighborhoods could be categorized as “low segregation”. In order to locate both high
and low segregation communities with relatively similar levels of poverty for the current
research, it was reasonable to use a figure that was '/2 that of the customary figure for CP.
This characteristic was labeled “high poverty”. Overall, in order to be categorized as low
SES a neighborhood was required to meet the Operationalization for all three
designations: low income, low education level and high poverty.

Likewise, in order to be categorized as high SES a neighborhood was required to
meet the Operationalization for three characteristics, in this case, high income, high
education level, and high concentrated affluence. Concentrated affluence (or CA), is the

phenomenon in which many wealthy residents live in the same neighborhood. CA was

53

defined in this research as 40% or greater number of households having incomes greater
than 1.6 times Oakland’s MHI.

Qualification for reduced price lunches (RPL) also factored into the
Operationalization of SES for this study. To be eligible for free meals, the National
School Lunch Program (N SLP) requires children to come from families with incomes at
or below 130 percent of the poverty level (currently $21,710 for a family of four). Those
between 130 percent and 185 percent of the poverty level (currently $30,895 for a family
of four) are eligible for reduced-price meals. In the current study, Low SES participants
attended schools in which 72% (2 x 36%, the state average in number of students per
school with RPL) or more students received RPL. High SES participants attend schools
in which just 18% (half of the state average number of students per school with RPL) or
fewer students received RPL. Reduced price lunches were also used on an individual
basis to determine if each particular child matched the SES Operationalization of his/her
larger community. Those that did not match the Operationalization of their communities
were excluded as participants. See Reduced Price Lunch Screener in Appendix I.
Segregation (SE G)

SEG or racial-residential segregation was determined by the racial composition
of the African American and White populations within individual census tracts. “Race”
was operationalized as two major groups: African American (those who self-identified as
Black or African American on the 2000 US. census) and White (those who self-
identified as White on the 2000 US. census). Segregation levels of census tracts were
computed by deriving the percentage of African American to White residents in a

neighborhood. That is, the ratio of African American to White resident per hundred

54

residents of a census tract was computed. High segregation was defined as 50% or more
Afiican Americans in residence. Low segregation was defined as 30% or fewer African
Americans in residence.3 To control for the effects of additional racial/ethnic groups on
this study’s results, neighborhoods and/or schools that included greater than 1/3 of any
one additional racial/ethnic group (e. g., Asian or Hispanic) were excluded from the study.
See Appendix B for descriptions of the specific tracts used in the current study.

Stimulus Construction

Data were collected from participants for three separate tasks, as described below.
The first two tasks involved “controlled improvisation” (Andersen, 1990), a method of
language data collection that involves role-playing with puppets. Two different puppets
were used (a teacher puppet and a peer puppet) in order to create a teacher—centered and a
peer-centered context. Each of the first two tasks required participants to complete an oral
expression, cloze task and a receptive, pointing task. The expressive and receptive
subtasks resulted in four total tasks: 1) a receptive lexical semantics and 2) expressive
lexical semantics task as well as a 3) receptive requests and 4) expressive requests task.
The final task was a brief survey that required examiners to ask participants about their
self-reflection on the use of register shifting.

Stimuli were constructed for both of the lexical semantics tasks, both of the
request tasks, and the one self-reflective task. Stimulus items were randomized, as
described in Appendix H. Each child was administered the same randomized list using
uniform procedures. The four, lexical semantics and request tasks were presented to each

participant in random order. The self-reflective task was consistently presented last.

 

3 The average percentage of African American residence for Oakland County was 10% according to the
2000 US. census. Low segregation was defined as 3x the county average or 30% or fewer African
Americans in residence.

55

Lexical Semantics Task Stimuli

The lexical semantics stimuli were chosen from sources that list definitions of
AAE words. They were either borrowed from English (Smitherman, 2000) or have a
historically West African origin (Turner, 1949). Words were selected for their cross-
generational, cross-regional use as well as their probable salience to 3rd grade speakers of
AAE or SMAE.

This section identifies the stimulus items that were used to investigate differences
between the receptive and expressive use of AAE/SMAE homonyms, as measures of
registralfield. Hand-drawn, color pictures (Houck & Higley, 2006) were used for both
the receptive and expressive tasks. Table 3 shows the items used in the lexical semantics
task.

Request Task Stimuli

The request tasks were developed in accord with Susan Ervin-Tripp’s (1976,
I977) description of request types. Receptive tasks presented an array of direct and
indirect request forms, and the expressive tasks required participants to analyze scenarios
in which requests were made of a peer or a teacher.

This section identifies the stimulus items that were used to investigate differences
in use of request forms as measures of registral tenor. As with the lexical semantics task,
each item was designed to observe both the receptive and expressive responses. Again,
hand-drawn, color pictures were used for both expressive and receptive tasks. See Tables

4 and 5 for specific expressive and receptive, request forms task items.

56

Table 3

Lexical Semantics Task Items

 

Lexical item

Modalilty (Language Variety)al

 

Receptive (AAE)

Receptive (SMAE)

Expressive (AAE
meaning only; No
SMAE meaning
provided)

 

1. kitchen

2. tight

3.poor

4. sugar

5. bright

a. “Who do you use
the word kitchen with
when you mean ‘the

place on the back of

the neck where hair
grows short and
curly’?”

a. “Who do you use
the word tight with
when you mean
‘really cool’?”

a. “Who do you use
the word poor with
when you mean

‘skinny or sickly’?”

a. “Who do you use
the word sugar with
when you mean

‘kisses’?”

a. “Who do you use
the word bright with

when you mean ‘light

b. “Who do you use
the word kitchen
with when you mean
‘a room in the house
where food is

cooked’?”

b. “Who do you use
the word tight with
when you mean
‘closely fitting’?”

b. “Who do you use
the word poor with
when you mean

‘having no money’?”

b. “Who do you use
the word sugar with
when you mean ‘the
thing you eat that’s
sweet, white and
powdery?”

b. “Who do you use
the word bright with

to describe a strong

57

“Look at this little
girl’s hair. You can
tell she needs a perm
because her

looks sloppy.”

“Look at this man
sitting in his brand
new car. His new car
is .”
Look at this girl. She
doesn’t eat right.“lf
she ate better she

wouldn’t look so

3,

Look at the baby

sitting with her mom.
“The baby is giving

her momma

3,

some

Look at these girls.
“The
light eyes.”

girl has

Table 3 (cont)

6. tote

7. bad

8. hip

9. nap

10. crib

skinned’.

a. “Who do you use
the word to tote with
when you mean ‘to

carry
heavy

something

9179’

a. “Who do you use

the word bad with
when you mean
‘something really
good’?”

a. “Who do you use
the word hip with
when

you mean

‘cool’?”

a. “Who do you use
the word nap with
you mean ‘a
tight curl or knot of
hair’?”

a. “Who do you use

when

the word crib with

when you mean

‘house’?”

light.

b. “Who do you use
tote with

when you mean ‘a

carry

the word

bag to

something in’?”

b. “Who do you use
the word bad with
when

‘bad’?”

you mean

b. “Who do you use
the word hip with
when you mean ‘the
place where your
body attaches to your
leg’?”

b. “Who do you use
the word nap with
when you mean ‘a

short rest’?”

b. “Who do you use
the word crib with
when you mean ‘the
bed where a baby
sleeps’?”

Look at this boy and
his Grandma.
Grandma’s groceries
are heavy. “He—
the groceries for his
grandma.”

Have you seen the
new Street Fighter?
“That videogame is

9

really .’

This
wears the new styles.

girl always

3

“She is really .’

Look at this girl’s
hair. “The in
her hair make it hard
to corn .”

This boy lives right
around the corner.
“He tells his friend
that he needs to stop
by the to
change his
before they go to the

 

shoes

park.”

 

a AAE = African American English and SMAE = Standard Mainstream American English

58

Table 4

Expressive Request F arms Task Stimuli

 

Scenario

Interlocutor

 

Teacher

Friend

 

Scenario I. Forgot pencils

at home

Scenario 2. Umbrella in

doorway

Scenario 3. Books on the

floor

Scenario 4. Need help with
heavy bag

a. You forgot all your
pencils at home but you
need one so you can do
your schoolwork. What do
you say to your teacher?

a. Your teacher left her
umbrella in front of the
door. You need her to move
it so you can get into the
classroom. What do you
say?

a. Your teacher has books
all over the floor and it’s
hard to walk around. You
want her to clean them up.
What do you say?

a. Your friend is busy doing
something else but you need
his/her help to move a
heavy bag of toys off of the
carpet so you can play
there. What do you say?

59

b. You forgot all your
pencils at home but you
need one so you can do
your schoolwork. What do
you say to your friend?

b. Your friend left his
umbrella in front of the
door. You need him to
move it so you can get into
the classroom. What do
you say?

b. Your friend has his/her
school books all over the
floor and it’s hard to walk
around. You want him/her
to clean them up. What do
you say?

b. Your teacher is busy
doing something else but
you need her help to move
a heavy bag of toys off of
the carpet so you can play
there. What do you say?

Table 4 (cont)

Scenario 5. Want cupcakes

Scenario 6. Want to play
with toy

Scenario 7. Want your

backpack

Scenario 8. Need help with
book report

Scenario 9. Want to be

pushed on swing

Scenario 10. Need help

opening the windows

a. It’s not snack time yet but
today your teacher brought
treat the
class. . .cupcakes! You

can’t wait to eat one so you

a special for

ask the teacher if you can
have one now. What do you

say?

a. Your teacher brought a
brand new toy to school.
You want to play with it.
What do you say?

a. The bell just rang and it’s
time to go home from
school. Everybody is
getting their things ready to
go and you see your teacher
holding your backpack.

What do you say?

a. You need some help on
your book report so you ask
your teacher. What do you
say?

a. You want your teacher to
push you on the swing at
recess. What do you say?

a. It’s hot the

classroom and you want

inside

your teacher to help you
open the windows. What do

you say?

b. It’s not snack time yet
but today your teacher
friend
brought a special treat for

brought your
the class. . .cupcakes! You
can’t wait to eat one so
you ask your friend if you
can have one now. What
do you say?

b. Your fiiend brought a
brand new toy to school.
You want to play with it.
What do you say?

b. The bell just rang and
it’s time to go home from
school. Everybody is
getting their things ready
to go and you see your
fiiend holding
backpack. What do you
say?

your

b. You need some help on
your book report so you
ask your friend. What do
you say?

b. You want your fiiend to
push you on the swing at
recess. What do you say?
b. It’s hot the

classroom and you want

inside

your fiiend to help you
open the windows. What

do you say?

 

60

Table 5

Receptive, Request F orms Task Items

 

Stimulus Item Direct Indirect

 

1. You forgot all your pencils at home

but you need one so you can do your

schoolwork. Who do you say this to:

Can I have a pencil? X
2. You forgot all your pencils at home

but you need one so you can do your

schoolwork. Who do you say this to:

I need a pencil. X

3. Someone left an umbrella

in front of the door. You need them

to move it so you can get into the

classroom. Who do you say this to:

Could you move your umbrella? X
4. Someone left an umbrella

in front of the door. You need them

to move it so you can get into the

classroom. Who do you say this to

Move your umbrella. X

5. Someone left books all over

the floor and it’s hard to walk around.

You want them cleaned up.

Who do you say this to:

Could you clean up your books? X
6. Someone left books all over

the floor and it’s hard to walk around.

You want them cleaned up.

Who do you say this to:

Clean up your books. X

7. You need help moving a heavy bag

of toys off of the carpet so you can

play there. Who do you say this to:

Can you help me move that bag. X
8. You need help moving a heavy bag

of toys off of the carpet so you can

play there. Who do you say this to:

I need that bag moved. X

61

Table 5 (cont)

9. It’s not snack time yet but today
somebody brought a special treat for
the class. . .cupcakes! You can’t wait
to eat one so you ask if you can have
one now. Who do you say this to:
May I have a cupcake?

10. It’s not snack time yet but today
somebody brought a special treat for
the class. . .cupcakes! You can’t wait
to eat one so you ask if you can have
one now. Who do you say this to:

I want a cupcake? X
11. Someone brought a brand new

toy to school. You want to play with it.

Who do you say this to:

Can I play with your new toy?

12. Someone brought a brand new

toy to school. You want to play with it.

Who do you say this to:

Give me your toy.

13. The bell just rang and it’s time to

go home from school. Everybody is

getting their things ready to go and you

see that someone is holding your backpack.
Who do you say this to:

Could I have my backpack?

14. The bell just rang and it’s time to

go home from school. Everybody is

getting their things ready to go and you

see that someone is holding your backpack.
Who do you say this to:

Give me my backpack.

15. You need some help on your book report
so you ask. Who do you say this to:

Could you help me with my book report?
16. You need some help on your book report
so you ask. Who do you say this to:

Help me with my book report.

17. You want to be pushed on the swing

at recess. Who do you say this to:

Will you push me on the swing?

18. You want to be pushed on the swing

at recess. Who do you say this to:

Push me on the swing.

62

Table 5 (cont)

19. It’s hot inside the classroom and you

want someone to help you open the windows.

Who do you say this to:

Could you help me open the windows? X
20. It’s hot inside the classroom and you

want someone to help you open the windows.

Who do you say this to:

Help me open the windows. X

 

Self-reflective Task Stimuli
Items for the metalinguistic survey were developed in order to get
straightforward, yet open-ended responses from participants about their own awareness
of how and why they change the way they speak to different conversational partners. The
following questions comprised the survey and they were asked in the order shown.
1. Do you have a good friend?
2. What is his/her name?
3. Do you talk the same with your teacher as with your good friend?
a. If no: Tell me all the ways you talk different with your good fiiend and
your teacher. Then go to #4.
b. If yes: stop here.
4. Tell me all the reasons why you talk different.
Also see Appendix L for a list of questions that composed the survey.
Responses to these questions were expected to provide information about the
conscious availability of participants’ register-shifting knowledge. Although self-report

methodology is not always reliable (Johnson & F endrich, 1995), the value that it places

63

011

Ian

 

(I31

uni

{Ill

31h

lasl

COT.

 

lal

 

on community members’ judgments, honors their role as expert users of their own
language variety. This metalinguistic task was presented last, after the first two
(language) tasks were completed.

Questions 3 and 4 were timed using the second hand of a watch to impose a
uniform, 60 second time limit on the task. This method limited participants’ responses to
a manageable data set. They were encouraged to say as much as they could within the
allotted time.

Data Collection Procedures
Task Description

Each participant was tested individually in four task conditions. Two separate
language tasks were presented: i.e., the lexical semantics and request forms tasks. Each
task required both a receptive and expressive response. Table 6 illustrates these task
conditions.

Table 6

Task Conditions by Task Type and Performance Modality

 

 

 

Modality Type Task Type
Lexical Semantics Request Forms
Expressive Task Condition 1: Task Condition 2:
Lexical Semantics Request forms
w/ expressive w/expressive
response response
Receptive Task Condition 4: Task Condition 3:
Lexical Semantics Request forms w/
w/ receptive receptive response
response

 

64

hm

 

l't’l'

c111

(hr

1.111

ml

811i

 

pan

Pan

   

Investigators engaged participants in language tasks with the goal of observing
how they talk to a peer puppet and a teacher puppet. The procedure used was a modified
version of the “controlled improvisation” technique developed by Andersen (1990).
According to Andersen, controlled improvisation is appealing because it 1) allows
children the creative freedom to experiment with what they know about register shifting
during role-play; 2) it controls the conversation by limiting the topics to be discussed
with each particular character, and 3) it provides children with the freedom to be creative
within the time limits imposed by the constraints of data collection.

In this study, children were asked to “do the voice” of a puppet that represented
themselves while involved in a conversation with (1) a peer puppet that matched the
participants’ gender and ethnicity and (2) a White, woman teacher puppet that contrasted
with the participants’ ethnicity. That is, the African American girls “did the voice” of an
Afiican American girl puppet, and the African American boys “did the voice” of an
African American boy puppet.

Task Administration

Registral Shift Tasks

Each participant was instructed to do the language tasks as specified on the
standardized data collection forms (1-4) found in Appendix F. Receptive tasks required
participants to point to or touch a puppet. This flexible, forced choice task gave
participants the option to point to one or both of the puppets. Expressive tasks required
participants to (l) formulate requests in response to a verbalized and pictured scenario
and (2) use a word with an AAE or SMAE synonym to fill in the blank within a cloze

task. Each task had a line drawing that represented the scenario described within the task.

65

F ill-in-the-blank sentences were read aloud with the word “blank” verbalized in the place
of the lexical item to be filled in.

First, the teacher puppet and then the peer puppet were introduced, as described in
the script. Language task stimuli were presented live by the researcher or the research
assistant in random order and in the manner described in the script. Within each task, the
participants helped to operate their puppet. The peer puppet was held in the lap of the
researcher, who faced the participant at a 90 degree angle, at a distance of approximately
two feet. On the opposite side of the participant, the teacher conversational puppet was
placed on a table approximately two to five feet away. As the language tasks were
completed, the role-playing tasks ended, and the metalinguistic survey was given to a
child.

Testing was conducted in a room in each participant’s school. Ambient noise was
minimized. Participants were generally tested individually, with only the investigator in a
room. Occasionally, when access was limited to only one room, two participants were
tested separately, at the same time, in the same room. In these cases, testing locations
were at least 15 feet apart. Examiners’ voice levels were moderated in order to be heard
only by the relevant participant.

Puppets were acquired from LetUsTeachKids.Com
(http://letusteachkids.corn/puppet%ZOp/p%20ethnic.htrn). They were approximately 2 ‘/2
feet tall, made from man-made fibers and designed with a hole in the back to allow a
hand to move the mouth. Peer puppets were dressed in child-like clothing, and the teacher
puppet was dressed in adult-like clothing. Each participant played the role of a child

puppet matched to his/her age, gender and ethnicity. The child puppet operated by the

66

 

3.:

 

participant was given the same name as the participant. In this sense, the children were
asked to use the puppets to pretend to be themselves. “Peer” puppets were given the
names (Elijah or Breanna), which were selected from the Social Security
Administration’s (http://www.ssa.gov/cgi-bin/populamames.cgi) list of most popular
names for I995. Conversational partner puppets represented a peer of the participant
(matched for race and gender i.e. it was an African American boy or girl puppet) and a
teacher of the participant (a White, middle-aged woman puppet).

The participant engaged in each of the four language tasks and the self-reflective
survey. Colored drawings (Houck & Higley, 2006) were used to further establish the
authenticity of the improvised contexts. They were approximately 5x7” in size, color and
hand-drawn by professional artists.

Participants were oriented to the tasks via scripts (Appendix D), which explained
the nature and expected requirements of each task. The directions for each task were
given on the data collection forms (Appendix F).

Task presentation was randomized by (I) task type (lexical semantics and
requests) and (2) performance modality type (receptive and expressive). A random
number was assigned to each of the four language tasks, they were then labeled Task 1-
Task4. A list of sets of four randomized numbers (i.e., 1-4 representing the numbers of
the tasks) was then generated. Finally, tasks were presented to participants in the order
dictated by the randomized list.

Lists of items within each task were also randomized, and the randomly ordered
list was delivered in standard manner to each participant. No more than one hour was

needed to administer all four language tasks. Of the ten or more subjects from each of

67

the three primary schools in the study, half of the students were tested in the first four
weeks of data collection, and the other half were tested in the latter four weeks of data
collection. Similar staggering of data collection occurred at all of the testing sites for the
different communities except the fourth site, which was excluded from the primary data
analysis. Data collection at the fourth site was conducted all on the same day.

Two examiners collected the data: the primary researcher, a White woman with
ASHA, Certificate of Clinical Competence (CCC), and a Master’s degree in speech-
language pathology (SLP) and the other, an African American research assistant working
on a Master’s degree in SLP. Approximately half the participants (48%) from the three
communities used in the study’s primary analysis were evaluated by the primary
researcher, and half by the research assistant (52%). The three participants from the
HiSS/HiSG community were evaluated by the research assistant. The research assistant
was blind to the goals of the study.

Self-reflective Task
Immediately after completing the language tasks, each participant was given the
five-question survey regarding his/her own thoughts about register shifting. The survey
was given orally and face-to—face. Participant responses were recorded on-site by hand in
addition to audio and/or video recordings, which were done to substantiate any data that
may have been unclearly perceived on-site.
On-Site Recording of Responses

Participants were audio and/or video-taped depending upon available resources.

Audio-taping was done using a state of the art audio recorder (Sony TCM-465V) and

lapel microphone (Sony ECM-3 l) and videotaping was done using a video cassette

68

recorder (Sony V-8, Model #DCR-TRV310). Standard data collection forms (Appendix
F) were used to record as many responses as possible, on-site, at the time of data
collection. Responses were rechecked by listening to audio or watching video tapes at a
later time if responses were unclear at the time of data collection or if an examiner
inadvertently neglected to record a response.

Preparation and Scoring of Data for Analysis of Register Shifting

Data were transcribed by trained research assistants who did not participate in the
data collection. Transcription was done by typing the responses from each data collection
form into an Access document. The responses to each of the five tasks were labeled
separately within the document: Lexical semantics (Tasks #2 and #4), request forms
(Tasks #1 and #4), and the self-reflective report.

Each data set was scored separately after transcription. The research assistants
who scored the data differed from those who orthographically transcribed the children’s
responses. Research assistants, who did the transcription data scoring, were blind to the
goals of the study. The research assistant, who helped to collect the data, was not
involved in data preparation or scoring.

Data Scoring for Request Task
Receptive Modality

For the receptive requests task, online scoring was done to identify the puppet(s)
to which the participant pointed. See data collection forms in Appendix F. Missed data
points during online scoring, were added at a later time, when the scorers evaluated the
audio or video recordings of the testing session. A child was given credit for one register

shift if for a given pair of items (e. g, requests to open the window ), the participant

69

pointed in the expected direction to the peer puppet for one stimulus in the pair and then
to the adult puppet for the other stimulus. Each of ten pairs of items were compared in a
similar manner and resulted in ten total possible register shifts for the receptive request
forms task. The actual number of shifts used, was counted, and a percentage of the ten
total expected shifts was then derived. The resulting Percent Register Shift (PRS) scores
were analyzed in terms of average PRSs across the different independent variables when
conducting the data analyses.
Expressive Modality

Request forms used within the expressive task were labeled as either direct or
indirect (Appendix C). The frequencies of direct and indirect forms used were tallied. As
with the receptive task, each pair of items used with the teacher and the peer within each
scenario, were compared. The comparison resulted in one register shift, if it was in the
expected direction. A shift occurred if the request form used with the teacher was
indirect, and the one used with the peer was direct. Within each task, a total of ten
possible shifts could have potentially occurred. To get a PRS for this task, the actual
number of shifts was divided by the total number possible. These preparatory tasks were
completed by research assistants. Like the receptive modality, the scores were analyzed
in terms of average PRSs across the different independent variables.

Data Scoring for Lexical Semantics Task

The procedures used to prepare data for the requests task were similar to those

used to prepare the data for the lexical semantics task. Receptive and expressive data

were scored separately.

70

Receptive Modality

The receptive lexical semantics task was scored online to identify the puppet(s) to
which the participant pointed. See data collection forms in Appendix F. Missed data
points during online scoring were recorded later by scorers who observed the recorded
testing sessions. If for a set of homophonic lexical items (e.g., kitchen meaning the back
of the neck and the room in the house), the participant pointed first to the peer and then
to the teacher puppet in the expected direction, he/she was given credit for one register
shift. Comparison of each pair of items within a scenario resulted in ten total possible
register shifts. Based on the actual number of shifts observed, a percentage of the ten
possible shifts was derived (Percent Register Shift score or PRS score). All these tasks
were completed by research assistants who were not involved in data collection. For data
analysis, mean PRS scores were calculated to investigate group differences across the
independent variables.
Expressive Modality

Instead of comparing cross-contextual use of each lexical item, only the AAE
meaning was selected for analysis of the expressive modality. To score this task, on-site
observations were used to determine whether an AAE lexical item (i.e., poor meaning
gaunt) or an SMAE lexical item (skinny meaning gaunt) was used to complete each
cloze task. See Appendix F for data collection forms. Differences in lexical item use in
the expected direction signaled one register shift. For example, the participant was given
credit for one register shift if the word poor meaning gaunt was used with a friend and
the word skinny meaning gaunt was used with the teacher. There were a total of ten

possible shifts for this task. To derive a PRS score, the actual number of shifts used was

71

compared to the ten total possible shifts. Like other tasks, scoring was done by research
assistants who were not involved with data collection. Mean PRS scores were calculated
and compared to other measures.

Data Scoring for Self-Reflective Task

Data were transcribed and then coded by research assistants. Responses were
coded for the number of topics given in response to “Tell me all the ways you talk
differently with your teacher and with your friend?”, and for the number of reasons given
in response to “tell me all the reasons you talk differently?” ’

Reliability of Data Scoring

Data collection and analysis were done by different teams of research assistants.
This procedure was expected to minimize experimental bias.

Inter-rater reliability was assessed by one of the research assistants, who
computed the scores. She recomputed all scores for 5 of 36 randomly selected
participants. These scores were then compared to those scores derived from the original
scorer; 13.89% of the total number (N=36) of the participants scores were re-analyzed for
inter-rater reliability using this procedure.

The same, randomly selected set of participant scores, 5 of 36, was selected to
compute intra-rater reliability. Two separate judges recomputed all of the scores for
these five participants. These scores were then compared to each scorer’s own original
scores; 13.89% of the total number (N=3 6) of the participant’s scores were re-analyzed

for intra-rater reliability using this procedure.

72

Data Analysis
Analysis of Registral Shift
The statistical model used was using a mixed, multivariate ANOVA

(3x2x2) with one between-groups factor (SEG/SES community type) and two within-
groups factors (language task type, i.e., lexical semantics and request forms and the
performance modality, i.e., receptive and expressive) (See Figure 6). Specifically,
register shift, as operationalized as Percent Register Shift or PRS, was analyzed in
relationship to each of the aforementioned independent variables. Repeated measures
were taken across teacher-centered and peer-centered contexts.

To answer the research questions, individual PRSs were calculated for each task.
Recall the detailed description of calculation of PRSs in the above subsections labeled as
-- Data Scoring for Request Forms Task and Data Scoring for Lexical Semantics Task.
Then, PRS means were compared across each of the independent variables to investigate
potential relationships between variables. SPSS was used to statistically analyze the data.

Analysis of Self-Reflective Report

A binomial comparison was done to determine whether overall, equal proportions
of yes as opposed to no responses were used to answer the question, “Do you talk the
same with your teacher as with your good friend?” ANOVAs were then used to analyze
the relative frequency distribution of negative responses across community types as well
as to analyze the community variation in number of utterances used to explain how and
why participants reported shifting registers. This analysis entailed a relative frequency

analysis across the four, community cohorts. The goal was to determine if these cohorts

73

 

Communi T 6
Performance Modality ty yp

Figure 6. Variation in percent register shift by community type, performance modality
and task type

 

differed in the extent to which participants talked about what they thought they actually

did in order to achieve register shift.

74

CHAPTER 3
RESULTS

This study aimed to determine whether African American children display
register-shifting skills, and if and how such skills were related to membership in a larger
social community, as defined by the combined characteristics of socioeconomic status
(SES) and segregation (SEG).

Due to the inability to recruit sufficient numbers of participants from the High
SES, High SEG (HiSSHiSG) community (n=3), it was excluded from the main statistical
analyses of the data, which were designed to explore differences among community
types. However, the small dataset for this group was useful for comparing potential
community-specific differences in shifting behavior.

Data were analyzed for register shifts in the expected direction. An expected shift
was coded for the lexical semantics items when children used SMAE, the school-valued
variety, with the teacher puppet and AAE, the peer-valued language variety, with the peer
puppet. Likewise, for the request items, an expected shift was coded when more polite
forms were used with the teacher than with the peer puppet. When the same variety was
used with both the teacher and peer, no shift was counted for either task. Finally, when
SMAE was used with the peer and AAE was used with the teacher, an unexpected shift
was counted on the lexical semantics task. When a more polite form was used with the
peer and a less polite form with the teacher an unexpected shift was counted on the

request forms task.

75

Reliability of Scoring Register-Shifting Behavior

Both inter-rater and intra-rater reliability were measured by percent agreement in
judging expected shifts for tasks 1-4 and the percent agreement in the number of
utterances tallied for the self-report of register-shifting task. Reliability estimates were
based on 5, or 13.9%, of the 36 study participants. The five participants used in the
reliability measures were chosen randomly from the entire pool of 36 participants. All
communities except the LOSS/HiSG group were represented in the set of participant
scores from which reliability measures were calculated.

Inter-rater reliability was assessed by one research assistant, who computed the
scores. She scored five of the participants whom she had not originally scored for each
task. These scores were then compared to those obtained by the original scorer. The
same set of participant scores were used to compute intra-rater reliability. Two separate
judges recomputed all of the scores for these five participants. These scores were then
compared to each scorers own original scores.

The generally high reliability shown in Table 7, which ranged from 88-100%,
indicated that the scoring procedures yielded consistent results across different examiners
(inter-rater agreement) and within the same examiner across repeated attempts (intra-rater
agreement).

Presence of Shifting Behavior

Table 8 shows that every child shifted in the expected direction. The number of
expected shifts per participant averaged 10.28 (SD = 4.64). Most or 72.2% (N = 36) of
the participants shifted in the expected direction from 6 to 15 times. No participant had

fewer than three total expected shifts or more than 21 total expected shifts.

76

Table 7

Reliability of Scoring of Tasks

 

 

 

Types of Tasks Percent Intra-Rater Percent Inter-Rater
Agreement by Rater Agreement by

Identification Number Rater Identification

Numbers

1 2 3 4 I & 2 3 & 4
Expressive Requests 98 88 _ _ 96 __
Receptive Requests 100 88 _ _ 94 _
Expressive Lexical Semantics 98 98 _ _ 98 _
Expressive Lexical Semantics 98 90 _ _ 96 _
Self-Reflective Report 98 100 96

77

Table 8

Number of Children Categorized by Total Number of Expected Shifts

 

 

Number of Possible Shifts Number of
(N =40) Children
(N=36)
0-2 0
3-5 6
6-10 13
1 1-15 13
16-20 2
21-25 2
26-40 0

 

78

Categorical Distribution of Shifting Behavior
Table 9 shows that only a small percentage (5.56%, N = 36) of the participants
showed evidence of a shift on just one of the four tasks. This group of participants can be

called undistributed shifters.

Table 9

Comparison of Undistributed and Distributed Shifters (N =3 6)

 

 

Participant Group Number Percentage

All Participants 36 100
Undistributed Shifters 2 5.56
Distributed Shiftersa 34 94.44

 

aDistributed shifters were represented significantly more than 50%
of the time (p < .001) according to a binomial comparison of

94.44% distributed shifters to the test probability of 50%.

In comparison, 94.44% of the 36 participants showed evidence of a shift on two
or more of the four tasks. This group of participants can be called distributed shifters.
This comparison supports the conclusion that there were significantly more distributed
than undistributed shifters. See footnote to Table 9. That is, their shifting skills were
distributed across two or more tasks instead of a single one.

Table 10 shows variability among the distributed shifters in the extent of shifting
behavior displayed. Participants can be divided into four categories: undistributed,

mildly, moderately, and well-distributed shifters. A chi-square analysis (22(1, n = 36) =

79

Table 10

Extent of Distribution of Shifting Behavior by Number and Percentage

 

 

Extent of Shifts by Task Type of Shifter Number Percentage
(N=36)

Shift on 1 and Only 1 Task Undistributed 2 5.56

Shift on 2 Tasks Mildly-Distributed 10 27.78

Shift on 3 Tasks Moderately- Distributed 20 55.56

Shift on all 4 Tasks Well-Distributed 4 11.11

 

28.12, p < .001) indicates that there were significant differences in the frequencies of
mild-moderately distributed shifters as opposed to either undistributed or well distributed
shifters. Overall, most participants can be placed in the middle groups, which identified
either mildly or moderately distributed shifters. Relatively few participants can be placed
on the extremes as undistributed or well-distributed shifters.

Variation in Shifting Behavior in Relation to Community, Task and Modality Types

The independent variables in this study were task type, modality, and community

type. Register-shifting performance was analyzed in relation to each of these three
variables. The examiner, an additional variable, was analyzed as a covariate. Although
the two examiners were matched for gender and relative age, they differed in race; one
was White and the other was African American.

Table 11 shows the results of a parametric analysis of variance (2 x 2 x 3

80

Table 1 1

Analysis of Variance in Percent Register Shift Scores

 

Source

Corrected Model
Intercept

Examiner (as
Covariate)

Modality
Task-type

Community-typea

Modality X Task-
type

Modality X
Community-Type

Task-type X
Community-Type

Modality X Task-
type X
Community-type
Error

Total

Corrected Total

Type III
Sum of
Squares
392.51
53.18

.943

369.69
14.05
1.65

.98

5.70

1.65

1.75

457.37

1550.00

849.88

*df

12

119

132

131

Mean
Square

32.71
53.18

.94

369.69
14.05
.05

.98

2.85

.83

.88

3.84

F

8.51*

13.84*

.25

96.19*

3.66**

.01

.25

.74

.22

.23

<00]

<00]

.62

<.001

.058

.99

.62

.48

.81

.80

E ta-
Squared

.46
.10

<01

.45
.03
<00]

<01

.01

<01

<01

 

*p < .001. **p approaches significance at a level of .058

a The analysis was based on three communities: HiSS/LoSG, LOSS/LoSG, and
LOSS/HiSG. The HiSS/HiSG community group was excluded due to paucity of

participants.

 

ANOVA). Of the three main effects, only the modality was statistically significant at the
minimum conventional probability level of .05. The moderate effect size (n2 = .45)
accounted for nearly half of the score variance. The main effect, task type and
community type, were not significant. The two-way and three-way interactions for the
three main effects were not significant either. Neither was the examiner covariate effect.
Figure 7 graphs the relationships among the three main effects observed for
modality, task and community. It shows that for the significant main effect of modality,
the mean percentage shift scores were higher in the receptive than the expressive
modality across task and community types. The main effect for task type was not
statistically significant at the maximum allowable conventional levels of probability (cf.,
.05 and .058). It was however, biased toward higher mean scores for the lexical than the
request forms task regardless of modality and community type. The means for these two
tasks and the two modalities were similar across the three community types compared in
the statistical analysis. Even though the means for the HiSS/HiSG community group were
the highest, its outcomes were excluded from the statistical analyses because they were
based on a small sample size (n = 3). Consequently, the results may be spurious. These
interpretations of the data can be seen more vividly in Figures 8 and 9, which respectively

graph the modality and task main effects by community type.

82

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80%
70%
60%

g 50%

CI)

2‘.

E 40%

.5

(I)

g 30%

'36

0)

3‘5 20%

8

0

ts

an 10%

 

 

 

 

 

 

 

O HiSS/LoSG

’8
I,” \\ E] LOSS/LoSG
/’ \
0” A LOSS/HiSG
\
\, <> HiSS/HiSG

 

 

Receptive Receptive Expressive Expressive
Request Lexical Request Lexical
Forms Semantics Forms Semantics

 

Figure 7. Percent shift by community by task where tasks are categorized as: Task 1

(Receptive Request Forms), Task 2 (Receptive Lexical-Semantics), Task 3 (Expressive

Request Forms) and Task 4 (Expressive Lexical Semantics). HiSS/LoSG == High Social

Status/Low Segregation, LOSS/LoSG = Low Social Status/Low Segregation, LOSS/HiSG

= Low Social Status/High Segregation, HiSS/HiSG = High Social Status/High

Segregation. HiSS/HiSG community is indicated by a dotted line to show its potential

relationship to other communities. It was excluded from the ANOVA outcomes.

83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

70% Receptive
Mode
60% .
______ Expressrve

:,\°\ Mode
9'; 50%
04 O
9:
5'“: 40% \Q
..C.‘
m

«"5
fire 30%

3’ p
E 20%

o I

8 /

all lO°/ El _____________ D."

0 ’ ‘ ’ ,
f a, , — ' T T ’

 

 

 

 

 

HiSS/LoSG LOSS/LoSG LOSS/LoSG (HiSS/HiSG)

Community Type

 

Figure 8. Main effect of percent register shift by modality by community (where the
HiSS/HiSG community is indicated in parentheses to show its potential relationship to
the other community types). HiSS/LoSG = High Social Status/Low Segregation,
LOSS/LoSG = Low Social Status/Low Segregation, LOSS/HiSG = Low Social

Status/High Segregation, HiSS/HiSG = High Social Status/High Segregation.

84

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

70%
60%
2 50%
e:
U)
25 40%
Q
..C‘.
2 30%
e
“a
g; 20%
a
8
s 10%
9..

 

 

 

 

Lexical
Semantics
Tasks

Request
Forms
Tasks

 

 

n=11

Community Type

HiSS/LoSG LOSS/LoSG LOSS/LoSG
n=10

(HiSS/HiSG)
n = 3

 

Figure 9. Main effect of percent register shift by task type by community (where the

HiSS/HiSG community is indicated in parentheses to show its potential relationship to

the other community types). HiSS/LoSG = High Social Status/Low Segregation,

LOSS/LoSG = Low Social Status/Low Segregation, LOSS/HiSG = Low Social

Status/High Segregation, HiSS/HiSG = High Social Status/High Segregation.

85

Post Hoc Analysis of Shifting Behavior across Community—Type

As already stated, the parametric analysis revealed that neither the community
main effect, nor its interaction with the modality and task type variables were
significant. Therefore the following poSt hoc probes of the data were done to determine
whether differences among the community groups could be revealed on other measures
of register-shifting behavior that were not explored in previous analyses. Specifically, the
community differences were analyzed using (1) categorical as opposed to continuous
measure of response shifts and (2) alternative linguistic devices for expressing shifts in
politeness.

Categorical Distribution of Register-Shifting by Community

Recall that it has been shown that that most participants could be placed along a
continuum of mild to moderate shifters (Table 10). The following analysis differed from
the analysis of variance because it focused on distribution of people as shift-ers and not
on their shift scores. Table 12 shows this analysis of community type as related to
distribution of shifting behavior. It reveals that significantly more (p = .04, compared to a
test probability of .50) strong than weak shifters were found in the LOSS/HiSG group
than in the other groups (p = .04, p = 1, and p = 1 respectively). That is, the other two
groups showed nearly equal numbers of strong compared to weak shifters.

Alternative Linguistic Means for Indicating Register—Shifts by Community Type
These analyses assumed that the children may have used linguistic devices other than the
ones sampled by the experimental stimuli, such as modal verbs used to express politeness
of requests. Measures of shifting behavior were selected for the expressive as opposed to

the receptive tasks for two reasons. First, the children showed the least amount of shifting

86

Table 12

Nature of Shift Distribution by Community Type

 

 

 

Type of Shifter HiSS/LoSG LOSS/LoSG LOSS/HiSG HiSS/HiSG
(n=10) (n=1 1) (n=l2) (n=3)

Undistributed l O 1 ...........
Mildly Distributed 4 5 l ...........
Moderately 5 5 8 ...........
Distributed

Well Distributed 0 1 2 ...........
Weak Shiftersa 5 5 2 ...........
Strong Shiftersb 5 6 10 ...........

 

Note. HiSS/LoSG = High Social Status/Low Segregation, LOSS/LoSG = Low
Social Status/Low Segregation, LOSS/HiSG = Low Social Status/High Segregation,
HiSS/HiSG = High Social Status/High Segregation. The HiSS/HiSG community is not
considered within this analysis due to its relative paucity of numbers. aWeak shifters are
categorized as those who are either Undistributed or Mildly Distributed. bStrong shifters

are categorized as those who are either Moderately Distributed or Well Distributed.

87

on them when using the planned outcome measures. So the question can be raised as to
whether the pre-planned measures adequately captured shifts that did occur on some
other level. Second, two, well known, standard measures of expressive language could be
justifiably and easily applied, namely, utterance length and the frequency of using
specific routinized forms of politeness and address, such as “please” and “Mrs”. The
procedures used and the results obtained are separately described below.
Mean Length of Utterance by Community Type

Utterance length was chosen because it is used routinely to measure the volume of
verbal output, and there are well defined procedures for computing it (Brown, 1973).
Conceivably, children talk more or less to a teacher depending on social distance and
comfort level felt. They may experience different degrees of comfort and willingness to
shift registers depending on their community type. Given that the amount of shifting
behavior observed was smaller on the expressive than the receptive tasks, firrther probes
were done of the expressive task in order to expose whether alternative linguistic means
were used to indicate politeness shifts. In particular, the Request Forms Task (Task #3)
was chosen because participant responses were nearly always in the form of complete,
sentence-length utterances; a form available for a standard analysis of mean length of
utterance (MLU) and the subsequent analysis of the use of the rote, lexical politeness
terms, “please,” and address terms to be described.

Method of analysis. MLU was computed by tallying the total number of
morphemes used in the response to each stimulus item and dividing by the number of
items. This denominator was usually ten but occasionally was reduced by the number of

“no responses” in order to ensure that participants’ MLUs were not unduly deflated when

88

no response was given to a particular task item. Denominators of less than ten were used
for 3 of 36 or 8.33% of participants. MLU scores were then averaged across participants
in each of the three community cohorts (HiSS/LoSG (n=10), LoSS/LoSG (n=1 1) and
LOSS/HiSG (n=12)).

Results by community type. As seen in Table 13 below, community differences
were evident when an MLU analysis of the Expressive Request Forms Task (Task #3)

was conducted.

Table 13

Community Differences between Teacher and Peer— Centered Mean Lengths of Utterance
(Afl. Us)

 

Community MLU MLU Average
Type

w/Teacher w/Peer MLU t-test d
HiSS/LoSG 8.87 9.25 9.06 t (9) = -.52 , p = .31 -.13
LOSS/LoSG 8.13 8.17 8.15 t (10) = -.42 , p = .34 -.06
LOSS/HiSG 8.25 7.58 7.92 t (1 l) = 1.93 , p = .04 .32

 

Note. HiSS/LoSG = High Social Status/Low Segregation, LOSS/LoSG = Low Social

Status/Low Segregation, LOSS/HiSG = Low Social Status/High Segregation.

The LOSS/HiSG group (n=12) was the only community group to show
significantly longer utterance lengths (t = l.93(l l), two-tailed p = .04) with teachers
(n=8.25) than with peers (n=7.58). A modest effect size was associated with this
difference ((1 = .32) accounted for approximately one-third of the variance. This

89

significant difference for one community but not the others is evidence of a community
based difference.

Given that the significant community group difference was distinguished by a
longer MLU with teachers than peers for Task 3, additional MLU analyses focused on the
number and percentage of individual participants with longer teacher than peer MLU
within each community group using the following procedures and criteria. Each of the
twenty, paired-task items was compared on utterance lengths used with teacher and peer
for a total of ten pairs analyzed. For each participant, those items with longer teacher
than peer MLU were scored as “l” and those with equal to or shorter teacher than peer
MLU were scored as “O”. The percentage of participants who showed a shift in the
expected direction on all ten, paired, task items was then calculated. In the HiSSLoSG
community just 10% (1 of 10) of the participants used longer MLU with teachers than
with peers across all ten task items. In contrast, 36.36% (4 of 11) and 58.33% (7 of 12)
did so in the LOSS/LoSG and LOSS/HiSG groups, respectively. A one-way analysis of
variance, which compared these mean percentages for the three community groups,
revealed no significant main effect for community type (F = 3.00 (2, 32), p = .07, 112 =
.17) at conventional probability levels. However, the LoSSHiSG community had (nearly
significantly) larger numbers of participants who used longer teacher than peer MLU
across all ten task items than did the HiSSLoSG group. Pairwise comparisons of these
two groups yielded significant differences in their estimated marginal means (mean
difference = -.48, p = .02) in terms of numbers of participants with greater peer than
teacher MLU across all ten task items. No other significant differences were observed for

pairwise comparisons that included the LOSS/LoSG group.

90

Specific Politeness Forms by Community Type

While the scores for Task 3 had been based on the use of modal verbs in the main
data analysis, the post hoc analyses focused on the use of specific politeness forms. It
singled out the use of “please,” and the use of specific address forms such as “teacher,”
“Mrs. Jones,” “Elijah,” and “friend.” Both types of forms were viewed as alternative
means of expressing politeness. They were easy to count and identify in the data sets of
each community group

Method of analysis. The numbers of instances of “please” and address terms used
in Task 3 were tallied by community and then divided by the respective number of
community members in each group in order to derive an average number per person (See
Table 14.).

Results by community type. Table 14 identifies community differences in use of
specific politeness forms. Peer-teacher differences were statistically significant with the
use of address terms (t = 2.32 (32), p = .03; d = .22) but not with the use of please (t =
1.14(32), p=.26; d = .14). The HiSS/LoSG group showed the highest rate of use of please
with teachers, relative to other community groups, although this difference was not
statistically, significantly different (F=.18 (2), p = .84; partial 112 = .01) See 3 x 2
ANOVA results in Table 16 in Appendix K).

Finally, all groups had relatively low rates of using address terms with peers however,
Table 14 shows that those in the HiSS/LoSG group (M = 2.6) were again ranked highest
in frequency of use. The LOSS/LoSG group was the lowest (M = .73); the LOSS/HiSG

group was ranked in between (M = 1.58). However, none of the community group

91

Table 14

Mean Frequency of Rate-Polite “Please ” and Address Terms Used in Expressive Request
Task by Community

 

Mean Mean Mean Mean
please please address address
Comm- w/ w/ w/ w/

unity n teacher SD peer SD teacher SD peer SD

 

Type

1 10 6.70 3.23 5.70 2.54 3.60 3.69 2.5 3.60
2 11 6.00 4.14 5.73 3.71 2.18 3.97 .73 2.41
3 12 5.83 3.16 5.67 2.46 3.33 4.03 3.33 4.03
(4) 3 (3.33) (3.66) (2.00) (0)

F 33 F = .18 (32, 2), p = .84 F = (32, 2) .93, p = .40

 

Note. 1 = HiSS/LoSG = High Social Status/Low Segregation, 2 = LOSS/LoSG = Low
Social Status/Low Segregation, 3 = LOSS/HiSG = Low Social Status/High Segregation,
(4) = HiSS/HiSG = High Social Status/High Segregation. The HiSS/HiSG community is
included in parentheses to show its potential relationship to the other communities. It was

excluded from statistical analysis. F = value resulting from peer-teacher comparison.

92

differences were statistically significant. See 3 x 2 ANOVA results in Table 17 in
Appendix K.
Conclusions from Post hoc Analyses of Community Type
The overall, post hoc analyses revealed that the LOSS/HiSG group used longer
utterances with teacher than peers when making requests and all community groups used
please more often than address forms as politeness markers.
Self-Reflective Report of Register-Shifting Use

The data regarding the self-reflective report of register use are found in Table 15 . The
survey yielded self-reflective reports of register use, as provided by the 28 respondents
who answered the question, “Do you talk the same with your teacher as with your good
friend?” Sixteen (57.14%, n = 28) respondents answered “no,” indicating that they do
speak differently with their teachers and friends (Table 15). More than half of all the
respondents reported talking differently with their teachers and peers, although the
difference between those who did and those who did not report speaking differently was
statistically insignificant using a binomial comparison of 57.14% to the test probability of
50% (p = .57).

Community Differences among Participants ' Reports of Whether They Speak Diflerently

with Teachers than Peers
Regarding community group differences, fewer participants from the HiSS/LoSG

community (4/9 or 44.44%) reported speaking differently with their fi’iends than with
their teachers as compared to each of the two other community groups (5/8 or 62.5% and
7/11 or 63.63%); however his difference was not statistically significant (F = .40 (2, 27) p

= .672 partial 112 = .03). Table 15 further reveals that the participants from the

93

HiSS/HiSG community 1) reported that they shifted more often and 2) offered more
language describing how and why they shifted than any of the other three community
groups.

Participant Reports of How and Why They Speak Differently with Teachers and Peers
Table 15 also shows some community differences in the number of utterances used by
participants to describe why and how they differed in their speech with teachers and
peers. The LOSS/HiSG group members offered more language, as determined by the
mean number of utterances, used to describe “why” and “how” they spoke differently to a
teacher as opposed to a peer (M = 7.08 compared to 5.86 and 5.57 for the other groups).
However, a one-way analysis of variance did not reveal a significant group effect (P = .76

(2, 25) p = .48, partial 112 = .06).

94

Table 15

Self-Report of Register Shifting: Participant Numbers and Means by Community Type

 

 

Community Number Who Mean Number Who Mean
Type Responded to Percentage Responded to Number of
Question 1 (Number) Question 2 Utterances
Who Reported Regarding
Shifting “Why”/“How”
Totalsa 28 57.14 (16), SD 26 6.17 (N),
= 10.77 SD = .80
HiSS/LoSG 9 44.44 (4) 7 5.86
LOSS/LoSG 8 62.50 (5) 7 5.57
LOSS/HiSG 11 63.63 (7) 12 7.08
(HiSS/HiSG) (3) (100 (3)) (3) (12.66)

 

Note. HiSS/LoSG = High Social Status/Low Segregation, LOSS/LoSG = Low Social

Status/Low Segregation, LOSS/HiSG = Low Social Status/High Segregation,

(HiSS/HiSG = High Social Status/High Segregation). The HiSS/HiSG group is included

in parentheses in order to show its potential relationship to the other community types.

aTotals include the 33 core research participants and exclude the HiSS/HiSG community

group with just three participants.

95

Summary and Conclusions Regarding General Findings

The following main results were found. Most of these 3rd graders showed registral
shifts on one or more of the tasks provided. ANOVA tests revealed that such shifts were
modality-dependent. A significant difference was found between receptive and
expressive modes with participants showing greater receptive than expressive register
shifi. Rank order differences were also found between task types. Specifically, the lexical
semantics task revealed more shifts than did the request forms task across modality and
community group, but this difference was not statistically significant. No overall,
significant main effect was observed for community type. However, post hoc probes of
the data revealed community differences in the relative frequency distribution of
participants who shifted across all four tasks. The LOSS/HiSG group had a significantly
larger number of strongly than weakly distributed shifters compared to the other groups.
Community differences also were observed on the Expressive Request Forms Task.
Specifically, the LOSS/HiSG community used longer utterances than those from the other
community groups when making requests of teachers than making requests of peers.
Please was used significantly more often than address terms as a politeness marker across
community types.

Self-reports of register use revealed that over half of the participants were able to
orally express that they do indeed speak differently with peers and teachers. Preliminary
data from the HiSS/HiSG community revealed that this group seemed to express its
awareness and explanation of shifting behavior to a greater extent than the other

community groups.

96

It is concluded that African American third graders do show evidence of registral
shifts. This shifting behavior indicates awareness of the differences between the rules of
teacher and peer-centered interactions. Most seem to be consciously aware of making
such shifts. These shifts are modality and task dependent but vary more with the
modality of presentation than with the linguistic task condition or community type.
Community differences in shifting behavior were evident, yet not in the expected

direction.

97

CHAPTER 4
DISCUSSION
Research Problem and Theoretical Framework Revisited
This study investigated children’s skill in shifting between peer-centered and

teacher-centered registers. In particular, it examined African American 31rd grader’s
register shifts during tasks that involved receptive and expressive use of lexical and
request forms. Language is one of the factors that has been implicated in the academic
underachievement of African American children. Another factor is social context. The
definitions of both language and social context were expanded in the current study.
Instead of focusing solely on structural language features, this study focused on register
(Halliday, 1978), or, the meaningful use of language in social context. Rather than SES
alone, the focus on social context incorporated both SES and SEG in a definition of
community type.

This research was guided by several theoretical assumptions. The first was that
register-shifting skill is an area of communicative competence. It necessarily develops
throughout one’s lifetime because of the changing requirements to communicate in
different types of social situations. Furthermore, register can be viewed as an element of
linguistic capital (Bourdieu, 1977) that co-varies with access to other sorts of cultural
capital. SES and SEG were aspects of cultural capital hypothesized to co-vary with
register-shifting ability in this study.

According to Systemic Functional Linguistic theory (Halliday, 1978), register
Shifting reflects specific aspects of the social environment. This study investigated two

linguistic behaviors with theoretical links to social context. The first was lexical

98

semantics, which reflected the theoretical notion of registral “field” (Halliday, 1978). The
second one was the use of politeness forms, which reflected registral “tenor” (Halliday,
1978). These two forms of registral expression were expected to be related to one’s
access to linguistic capital (Bourdieu, 1977), as defined by the combined SES and SEG
characteristics of school/community neighborhoods. Register-shifting skill was viewed as
a potential contributor to the relatively low academic success of African American
children, given its relevance to academic success (Delpit, 1995). Various methods have
been proposed for teaching African American children the language of school instruction.
But only recently have scholars suggested that these children cannot be expected to
passively absorb a code without being explicitly exposed to the social context in which
language learning occurs. Recall that Delpit (1995) stated that:

“Students should be taught the codes needed to participate fully in the mainstream

of American life, not by being forced to attend to hollow, inane, de-contextualized

subskills, but rather within the context of meaningful communicative

endeavors. . .and that even while students are assisted in learning the culture of

power, they must also be helped to learn about the arbitrariness of those codes and

about the power relationships they represent.” (p. 44)
This perspective suggests that the academic success of African American children may be
limited by not only restricted access to linguistic “subskills,” but also by their lack of
access to the social codes that underlie language. These social codes entail information
about relative social status, elements of which are reflected in register use. Culture-
specific word-meaning and use of politeness forms, for example, reflect knowledge about

underlying social codes. It follows, then, that differential access to these codes of

meaning may, in turn create differential use of register-shifting skill.

99

The following discussion focuses on what was learned about African American
children’s register-shifting skill in the present study, and on the implications of these
observations for registral theory, practical application and the future study of pragmatics.

Register Shift

African American children do, in fact, show evidence of peer-teacher register
shifting. They also have meta-linguistic awareness of peer/teacher, register shifting as
evidenced in the questionnaire data obtained in this study.

Such findings ought not be surprising, given that register shifting has been
documented as early as 3-5 years of age across cultures and languages (Andersen, 1991;
Clark, 2002; Donahue, 2000). However, observing it among African American children
is important because of its educational implications. The current study’s research findings
imply that a lack of register-shifting ability is unlikely to account for African American
children’s academic underachievement. The conclusion that these children exhibit
register-shifting skill is important because it may have been previously assumed that
African American children do not shift from peer to teacher-centered registers. This
result is especially important to refuting the false, historical assumption that African
American children have deprived linguistic skills in general. The results of this study
provide evidence that shifting from peer to teacher-centered register is certainly a skill
that they possess.

Register Shifting: A Modality- and Task-Dependent Skill

All of the participants in this study provided quantitative, positive evidence of

register shifting, but their display of peer-teacher, register shifts varied with the task type

and modality of response. They varied with whether an expressive or receptive response

100

modality was required and to a lesser extent with whether a lexical semantics or request
forms task was used. Register shifting in the expressive modality is often used to gauge
absolute, register-shifting ability; however, this research provides evidence that receptive,
register shifting is a separate skill. Furthermore, it remains to be seen which additional
factors may mediate this variation.
Register Shifting: A Modality Dependent Skill

The participants’ more frequent display of register shifts in the receptive than the
expressive modality should be expected when considering the typical patterns of both
first and second language learning. Receptive language skills often outweigh expressive
ones (ASHA, 1997; Bloom, 1974). Between 18 and 23 months of age a child is typically
able to use receptive language to follow simple commands, point to basic body parts and
enjoy storybook reading. At this same age expressive language is characterized by use of
only 8-10 words (NIDCD, 2007). A similar type of receptive/expressive mismatch occurs
in second language learning (Gethin & Gunnemark, 1996; Wells & Nichols, 1985).

Additionally, the processing load for the receptive and expressive modalities is
different (Bloom & Lahey, 1978). Specifically, more processing is involved in the latter
than the former. First, more processing is needed to translate linguistic representations
into oral gestures. Additionally, immediately after expressing a thought, speakers think
about what they’ve said. In this sense, there is an additional feedback loop involved in
oral expression in which what one has expressed is also processed receptively. Overall,
expressive language involves more processing than receptive language in at least two
ways: expressive language requires both motor planning and understanding one’s own

Verbal output. The extra processing involved in expression makes the expressive tasks

101

involved in this research more difficult, in terms of processing load, than the receptive
ones. For this reason better performance on the receptive than the expressive tasks may
be partly due to reduced cognitive demands.

Register Shifting: A Task Dependent Skill

Furthermore, the lexical semantics task resulted in more shifts than did the task
involving politeness of requests. Even if this outcome lacked the statistical power
observed for modality, it has validity within Systemic Functional Linguistics (Halliday,
1978). Recall that within this theoretical framework, lexical semantics, was
conceptualized as reflective of registral field. Politeness of request forms was
hypothesized to reflect registral tenor. Thus, qualitative differences in register-shifting
skill across task type may reveal something about participants’ prioritization of shift in
one registral aspect over another. Greater shifting with the lexical semantics task reflects
shift priority for culture-specific word meanings used in AAE as opposed to SMAE. This
result indicates that for the African American child participants in this study, registral
field was a relatively important tool for conveying peer-teacher register shift.

Within Systemic Functional Linguistics (Halliday, 1978), register theory is useful
for understanding how children apply differential significance to ethnicity-based word
meaning and authority-figure, role-based politeness. At least in the third grade, African
American, peer as opposed to White, woman teacher signals a need for greater registral
shift in field as opposed to tenor. Since participants used modal verbs in request forms
nearly equally as often with peers and teachers, registral tenor seemed (at least in the
initial analysis) to be a relatively less important tool than registral field for conveying

peer-teacher shifts. Post hoc analyses revealed that one of the main shifts in registral

102

tenor displayed by this pool of participants was the greater use of address forms with
teachers as opposed to peers. The forms of “teacher,” a general role assignment, and
“Mrs. Jones,” a specific role assignment composed of a title and a surname, were used
predominantly in this study.

At the same time, the observed task differences may reflect other task differences
related to (1) processing load, (2) familiarity with task items and (3) level of knowledge.
Regarding processing load, field, as instantiated by lexical knowledge, may be easier to
process than tenor, which was instantiated by knowledge of differential politeness of
request forms. The cognitive demands necessary to process the linguistic information
involved in each of these tasks were quite different. The lexical semantics task required
processing the meaning of a single word, i.e., semantic mapping. In contrast, the request
forms task required processing an entire utterance, i.e., syntactic processing. There is a
greater processing load (Bloom & Lahey, 1978) involved in processing syntactic
information as opposed to semantic information. Thus, there is assumed to be a greater
processing load for the request task than for the lexical semantics task. This mismatch in
processing requirements is likely to have contributed to an outcome that showed more
register shift for the lexical semantics than the requests task.

Regarding task familiarity, lexical semantics shifting was stronger than request
forms shifting even though some of the lexical items were not known by individual
children. Recall that the average number of lexical items unknown by participants was
.91 (or less than one lexical item per child). The lexical item with the least recognizability
was “tote” which was not known by half (18 of 36) of the participants. Only one

participant did not know five of the target lexical items. So, it is possible that, on those

103

items that showed no lexical semantic register shift, a participant did not know the target
word as opposed to not knowing how to make a registral shift using the word.

Regarding learning completeness, it is likely that, the expressive request forms
tasks in particular were affected by incomplete learning of this particular register-shifting
skill. This was evident in the general use of modal verbs with both peers and teachers and
the use of over-generalized forms. In the case of the present study the most apparent
over-generalized form used was “May you. . .” as a request to get a second person to
perform an action. The grammatical use of “may” requires the use of the first person and
thus a permissive interpretation. In contrast, the over-generalized form used by the
participants in the current study indicate that they have not yet learned the specific
requirements for using the polite, permissive form; namely that its use with the first
person does not extend to requests made of others. Task and modality-dependent registral
shifts have implications for assessment practices as discussed later in this chapter.

Relationship of Community Type to Register Shifting

The current study defined community type in terms of the levels (high and low) of
racial residential segregation (SEG) and socioeconomic status (SES). Community type
was defined in this manner because the levels of cultural capital in general and access to
the SMAE school register in particular were hypothesized to be related to the cultural
capital access afforded by SES (i.e., education, employment, finances) and SEG (i.e.,
physical proximity to others with access to the SMAE school register). Unlike many
studies that have focused solely on the relationship between low SES and African

American language use, this study expanded social context to include different levels of

104

SEG, which may interact with SES to differentially influence access to peer-teacher
register-shifting skill.

The main, parametric analysis of Percent Register Shift (PRS) by community type
failed to yield significant results. This study revealed little difference among community
types in the extent of shifting behavior displayed. The communities may not have been
differentiated because the language variables targeted were not the critical ones that
actually differentiated communities, as the post hoc analyses implied. Instead, it revealed
community differences in l) the number of strong shifters and 2) the use of MLU as a
signal of registral shift.

Community Differences in Number of Strong Shifters

It was surprising to observe that more of the participants from the LOSS/HiSG
group were strong shifters compared to the other community types. Evidently, these
participants are more willing to display their expressive knowledge than were those from
the other communities.

This finding was unexpected because the LOSS/HiSG was expected to understand
and speak in one register only. That is, they were expected to use the casual, vernacular
register with both teachers and peers because of the fewest SES resources and least
access to those who speak SMAE.

Community Differences in MLU as a Signal of Register Shift

It was surprising that the LOSS/HiSG community group used longer MLU with
teachers than peers, which suggested that utterance length was used as a politeness
marker. This may be the case because verbosity can increase with social distance. Clark

(1995) suggests that communicative interaction is needed to establish common ground.

105

Once common ground has been established communicative partners can more easily
predict or “interpredict” each other’s attitudes and actions. It can be inferred then, that
before common ground is established, one might need to be relatively verbose. After
common ground has been established the inter-prediction process can begin and verbosity
will likely decrease.

For this reason, more background information must be given to a stranger than a
peer. Communication with a stranger, on the other hand, requires verbosity to establish
background knowledge and common ground. With a peer, verbosity is not required in
that background knowledge and common ground already have been established.

This result is the opposite of that found by Labov (1972b) in which African
American, child participants were much more verbally forthcoming with friends as
opposed to an adult, White, male examiner. But in the current study, an MLU difference
was observed only for the LOSS/HiSG community. A more definitive interpretation of
these results could have been given had it been possible to compare the LOSS/HiSG and
HiSS/HiSG groups. Recall that the number of participants in the HiSS/HiSG community
group was drastically smaller than that for other community groups.

Definition of Community Type Revisited

Failure to observe systematic differences across the different types of
communities may have been caused by not only the type of linguistic measures used but
also by the paradigm used to define them. That is, categorization of communities by SES
and SEG levels alone may not account for the greatest variability in register-shifting
behavior. Previous investigations have described community differences in other ways

that have yielded language differences. Despite the fact that most of these studies focused

106

on “code switching” instead of register-shift, they are instructive; they have shown that
the alternate use of a language variety was influenced by other ways of defining
community types. For example, Craig and Washington (2004a) defined a community by
its size and location relative to a central city. They labeled their two communities as
“mid-size, central” and “urban-fringe”. The characteristics which distinguished these
communities were their population size and relationship to an urban center. SEG was
described as percent African American student body composition, which was 16% in the
former and 70% in the latter. SES was free to vary. Children in the mid-size central city
used fewer AAE morphological features and were judged to be the most adept “dialect
shifters.” The authors also acknowledged that those in the mid-size central city had fewer
African American peers in their classrooms than did those in the urban-fringe
community. As a result, they were likely to be systematically exposed to “SAE” or
“Standard American English” than those from the urban-fringe community.

Charity et al (2004) also suggested that exposure to speakers of different dialects
varies with type of community residence and influences language use. The important
community variables were 1) the population size of a city and 2) the community’s SES,
as determined by eligibility for free or reduced price school lunch. The authors concluded
that the reading levels of African American children were related to their familiarity with
“SE” or “School English.” SwainsonI showed that increases in SES led to higher
familiarity with SE. While increased SES generally led to higher familiarity with SE,
variation in familiarity with SE was rather high. Furthermore, Swainson found that the
addition of SEG accounted for much of the variation in familiarity with SE. When
findings were extended to a high SES/low SEG community, familiarity with SE was very

_

' Data reported in Charity et al. (2004), based on personal correspondence.

107

high and variation was greatly reduced. This research suggested that racial segregation in
addition to SES should be acknowledged when defining community effects on language.

Connor and Craig (2004) studied the language of a group of Afiican American
Head Start students from the aforementioned (Craig & Washington, 20043) urban-fringe
and midsized cities. Participants were all of low SES but differed in their community
membership. The researchers found more variation within than between groups. This
negative result for community differences points to the need for more specific definitions
of community type in relation to African American language variation.

Taken together the studies mentioned above supported the notion that there is
something about community type that relates to the language variation found among
African American child residents. African American children who live in larger, urban
areas appear to be less adept at code-switching than those from more mid-sized, suburban
areas.

Despite other ways to identify community types, the approach used in the current
study, which combined SES and SEG levels to describe communities should not be
abandoned for two reasons. First, it is possible that a larger participant sample would
have revealed registral shift differences in community types, as defined by the SES and
SEG levels used in the current investigation. As alluded to earlier, the lack of a stronger
relationship between register-shifting skill and community type as defined by SES and
SEG may be partly due to the difficulty with participant recruitment. The issue of

participant recruitment is addressed later in this section.

108

A second reason not to abandon the current framework relates to its potential
conceptual relevance. Results like those of Swainson’s2 indicate that the combination of
low and high SEG and SES has the potential to explain a significant amount of variance
in African American children’s familiarity with the structural components of “School
English.” The approach used in the current study, which created four different groups by
combining SEG and SES had the advantage of targeting two rarely studied, HiSS/LoSG
and LOSS/LoSG, groups of children. The most widely studied community groups in
terms of language variation is the Low Socioeconomic Status/High Segregation
(LoSSHiSG) group. The others (Low Socioeconomic Status/Low Segregation
(LOSS/LoSG), High Socioeconomic Status/High Segregation (HiSS/HiSG) and High
Socioeconomic Status/Low Segregation (HiSS/HiSG» have received less attention. The
identification of the HiSS/HiSG group, in particular, does not appear to have been
identified or described in past research.

The preliminary data on its small sample size (n=3) showed similar performance
patterns as the other three community types across tasks and modalities, yet bigger
differences in shifting behaviors. While the sample size was too small to render a reliable
statistical test of its differences, they were large and consistent enough across tasks to
encourage further study with increased numbers of participants. If the preliminary
differences observed in the current study are supported in future research, they would
indicate that children from HiSS/HiSG communities are different than the other
community types in terms of the extent to which they understand and express register

shifts across task type.

 

2 Data reported in Charity et al. (2004), based on personal correspondence.

109

The results of the current study showed that in at least some respects, SES and
SEG affect the presentation of register shifting in Afiican American children. The
implication is that we should be cautious about viewing SES as a single determining
variable of language variation. Additional community characteristics that affect language
use should be added to the discussion.

Implications
Theoretical Issues

Systemic Functional Theory

This study largely confirmed that African American children, a group that is
seldom studied in terms of register, do engage in register shifting. By applying Systemic
Functional Linguistic theory, one is able to qualify the type of register-shifting
knowledge displayed. This theory makes it possible to expose the relative greater use of
shifts in registral field (as displayed by differences in lexical semantics) as opposed to
registral tenor (as displayed by differences in politeness). The difference between
receptive and expressive competence was essential to interpreting the current data in that
receptive register-shifting skills were significantly stronger than expressive skills.
Systemic Functional Linguistic theory ought to be expanded to include receptive and
expressive components. In this manner, register shifting can be viewed broadly, as more
than just an expressive skill.

The interpretation of the data in the current study also was limited to register-
shifting knowledge as displayed in the areas of lexical semantics and politeness of
request forms. Consequently, all potential areas of inquiry regarding African American

children’s register-shifting skill were not explored in this study. For example, registral

110

mode was not addressed, and only two behaviors representing participants’ knowledge of
re gistral field and tenor were addressed. Registral mode could be addressed in future
studies by focusing on written as opposed to spoken communication. Written
communication would increase both'spatial (i.e., physical distance between interlocutors)
and experiential distance (i.e., distance between language and the social process). The
study of registral tenor could be expanded to encompass other rote politeness terms,
which function to modify politeness of requests, namely “excuse me” and “thank you” as
well as a means of analyzing “hedges” (e. g., “if you want me to”) and explanations or
justifications (i.e., “it’s hot in here so...” could you help me open the window?), which
function to modify the politeness of requests.

A major problem with the specific conclusion that third-grade African American
children value field over tenor is that for the most part, children were polite. They used
modal verbs, can, will, could, would, and may, with both their teachers and their friends.
In other words, their register shifts did not involve shifts in the use of modal verbs to
indicate politeness. Overall, they were exceptionally and equally polite in both peer-
directed and teacher-directed registers. Either this finding is real and most had not learned
to be more direct with their friends and less direct with their peers, or there was
something about the research task that did not allow them to express their real world
knowledge. For example, it is possible that the wording of the task stimuli could have
unknowingly prompted more-polite responses. The question, “What do you say?” may
have carried the implication, “What are you supposed to say?” as opposed to “What do
you actually say?” A prompt specifically designed to elicit a more authentic responses

would likely elicit more valid data in future research. As an alternative, one might

’lll

consider using a question such as “What would you really say to your very best friend

?” Using the participant’s true, best friend’s name to fill in the blank would

 

add an additional element of authenticity to the context.

Controlled Improvisation

In addition, experimental data were limited to what participants did within a
controlled improvisation task. So, behaviors were easily measured by using a controlled
standardized stimulus yet the behaviors and contexts were merely role-plays of actual
contexts.

Relationship of .Language and Community

The effects of community type were not as direct as those of modality and task
type, however, the differences that were observed indicate that community parameters are
related to language use and deserve future, systematic investigation in terms of their

effects on register use. It is likely that SES and SEG are relevant but that there are other
variables that account for even more of the within group variance generally found in
studies of this kind. One variable that has the potential to add explanatory value is the
social network. The strength of social networks appears to be positively correlated with
the use of vernacular language varieties (Milroy, 1987). For example, speakers of AAE
who have strong community or ethnic network ties tend to use more AAE features
(Milroy, 1987). Furthermore, recent research has shown that social networks are also
negatively correlated with mainstream language behaviors. For example, Jones &
Preston (in press) showed that AAE speakers who had stronger community and ethnic

network ties did not participate to as great an extent in the North Cities Chain Shift as

those who had weaker social network ties.

112

Applied Issues

Three types of applied issues emerged from this research: 1) research methods, 2)
educational and clinical assessment, 3) implications for educational policy.
Implications for Research Methods

This study exposed five main issues that should be considered when designing
future studies of register shift in African American child language. They include issues
related to natural versus experimentally controlled contexts of observation, minority
participant recruitment, participant recruitment using a top-down definition of community
type, the definition of community type in relation to language use, and the general
framing of studies of pragmatics.

Natural versus experimentally controlled contexts of observation. Any research
involving human participants requires the investigator to navigate between two types of
contexts for observation. Natural, unmanipulated behaviors have the advantage of being
authentic. Consequently, any research that relies on observation of natural behaviors has
the additional advantage of being generalizable to functional contexts. In contrast,
experimentally controlled research aims to increase the validity of the observed
relationship between the variables in question; separating extraneous effects from the true
effects of interest. The imposition of control, necessarily limits the naturalness of the
context. This dilemma creates special challenges for experimental work in the area
pragmatics. Pragmatics can be defined as language use in a real-world context. As a
result, attempts to decrease a context’s naturalness in order impose experimental control,

interferes with the very object of inquiry. Therefore, careful development of standardized

113

procedures should be used to both control for random variation and create maximally
authentic stimuli.

To carefully control the influence of random variation on the current study, task
types and items were chosen carefully and the items were delivered in a standardized
way. Experimental control was necessary to increase the study’s validity and as a result,
realism was sacrificed to some extent.

Despite this study’s relatively controlled method of data collection, the use of
natural language data as opposed to more “rigidly prescribed” methods would have been
beneficial for several reasons (Stockman, 1996b). First, use of a natural language sample
is more culturally sensitive because it is likely to elicit language that truly reflects one’s
community language variety; participants are relatively free to choose the words they will
use. Second, natural samples are more valid than those that are experimentally controlled
because they more accurately reflect the holistic, real-world use of language than do

narrow, de-contextualized, language-elicitation methods. The spontaneous language
needed for language sample analysis also is readily accessible; it does not require the pre-
planned data collection procedures that controlled elicitation methods do. Finally,
language sample analysis is also flexible in that it can be applied across speakers from
various language communities.

Use of a natural language sample may have increased the cultural sensitivity and
validity of this study’s outcomes. By restricting the task items, the language structures
from which the participants had to choose were limited. This limitation may have
prevented participants from fully displaying their real language behaviors. For example,

children’s expressive scores could have been negatively impacted by an aversion to play-

114

acting the assigned roles. But this negative outcome may be eliminated if children were
asked to talk to real friends and teachers. Yet all of the participants were willing to
participate in the simulated activity. It is unknown whether their simulated behaviors
represented what they would do in actual situations of like kind.

The realism of item responses could have been enhanced in this study by asking
for responses in a way that did not one a more polite and/or mainstream form. Task 3, the
Expressive Request Forms Task illustrates this. When the participants were given
examples of scenarios in which they had to make a request of either a teacher or peer to
get a problem fixed, the researcher asked, “What do you say to your teacher/peer?” The
children used a relatively equal and generally high number of modal verbs to respond to
both teachers and peers. Therefore, it is possible that the open-ended question “What do
you say?” was given a prescriptive interpretation by the children. That is, they may have
interpreted the question to mean, “What are you supposed to say?” or “What is the polite
way to ask?” Instead of “What do you as the speaker say normally?” Future research
could use alternative questions such as, “What would YOU say to your friend/teacher?”
or “What words would YOU use to ask your friend/teacher?”

Authenticity may have been increased by using a natural context, or a more
naturalistic experimental task. As a result, shifting behavior might have increased. It is
not known how increases would have affected the variation in register shifting.
Furthermore, it is unknown whether this variation would have been effected by the
studies independent variables: task, modality and community type. Likely, willingness to
display express shifts would increase due to decreased pressure to play-act, as might

willingness to display receptive knowledge.

115

Minority participant recruitment for research studies. It was fairly difficult to
recruit participants for this study. For this reason, the total number of participants was
smaller than had been hoped. To obtain even this number of participants, the selection
criteria had to be modified to include children who met all except one selection criterion,
albeit one that was judged to not compromise the ability to respond to the tasks, e.g.,
failing a hearing screening at one frequency.

The difficulties with participant recruitment in the current study may have
reflected the skepticism that some African Americans had about the research on their
group. This sentiment is likely rooted in the historic abuse of participant rights in
experiments such as the infamous Tuskegee—Syphilis Experiment; “. . .an outrage to our
commitment to integrity and equality. . (Clinton, 1997). In addition, the sensitive,
political nature of non-mainstream language use within the African American community
can further alienate potential Afiican American participants. While the linguistic
community agrees that all language varieties are equally valid and rule-govemed systems,
there continues to be considerable debate and division among members of the African
American community as to the value of African American English. The most recently,
publicly outspoken member of the African American community on this issue has been
Bill Cosby. On the 50'h anniversary of Brown vs. Board of Education, Cosby was quoted
as saying:

"They're standing on the comer and they can't speak English. I can't even talk the

way these people talk: 'Why you ain't?' 'Where you is?‘ . . . And I blamed the kid

until I heard the mother talk. And then I heard the father talk. . . . Everybody

knows it's important to speak English except these knuckleheads. . . . You can't be
a doctor with that kind of crap coming out of your mouth!"

116

This quote can be categorized as a very public expression of disdain for AAE. This
message reached the ears of Americans along with some retorts but none as clear as that
initial statement made by someone so generally revered as Bill Cosby. The “anti-ebonics”
viewpoint has been given more clout than the alternative view that African Americans
need to learn how to shift their way of speaking in different situations.

This study was initiated at a time in history when linguists agree that African
American English is valid although people in general, including African Americans, do
not appreciate its value. Some of the difficulties with participant recruitment in the
current study can be interpreted with that knowledge. In the HiSS/HiSG community, for
instance, shortly before data collection began for the current study, a high-level school
administrator, who was not African American, made public comments about the
ungrammatical language use of even the brightest of the district’s high school students.
Despite her resignation of her position, certainly due in part to the negative publicity
surrounding her comments, a study that focused on such a divisive issue was not likely to
be palatable to a district undergoing such drastic administrative changes. Ultimately I was
not given permission to conduct research in the public school system of that community
and as a result, was able to recruit only a few participants from a private school on the
fringes of the neighborhood.

It is a pity that a critical mass of participants could not be recruited from a
HiSS/HiSG group. In fact, most research focuses on low SES groups instead of High SES
groups of Afiican Americans. Those high-end groups in highly segregated areas, in
particular, might tell us something about the distinct effects of segregation on AAE

language use. In terms of register shifting, it may be that this group is prone to the

117

greatest shifting due to access to both economic capital and a high concentration of
African American community members.

Participant recruitment using top-down definition of community type. This study
relied heavily on US. Census Statistics to determine neighborhood characteristics. That
is, neighborhoods were identified from massive lists of county-based, census-tract
figures; then each potential neighborhood was mapped as were particular elementary
schools within each neighborhood. Individual students were then recruited from eligible
schools in these neighborhoods. As might be inferred, this process was one of “top-
down” honing from the macro-level census information to the micro-level student
participant. This process, however meticulous in its potential to describe the SES and
SEG characteristics of each neighborhood, severely limited the pool of potential
participants. The alternative would be a bottom-up model in which participants could be
selected from any school in a given area (a county or city for example), census tract
information could then be collected; afterwards, participants could be grouped according
to neighborhood characteristics. Ultimately, a multi-level model might be applied in
order to distinguish relevant variables. However, a major problem with utilizing such a
bottom-up model is that the particular neighborhoods of interest as differentiated by High
and Low SEG and High and Low SES, would not likely receive adequate representation.
This is a largely a matter of probability based on the actual occurrence of neighborhood
types in the real world. For example, in selecting the neighborhoods for this study, there
were many HiSG/LOSS and LoSG/HiSS neighborhoods from which to choose. In
contrast to these two extremes, there were very few HiSG/HiSS and LoSG/LOSS

neighborhoods from which to choose. This is a result of the continuing trend for wealthy

118

city neighborhoods to be predominantly White and poor neighborhoods to be
predominantly African American.

Definition of community type in relation to language use. Community type was
defined for the purposes of this study in terms of high and low SES and high and low
SEG. As stated in the discussion of community effects above, recent research has pointed
to community variables thought to have an effect on African American children’s ability
to switch from AAE to SMAE. These variables have tended to be more descriptive than
those used in this study. For example, other studies have taken into account: (1)
community SES, (2) city size, (3) geographical relationship to the central city, and most
recently (4) segregation level (Craig & Washington (2004a), Charity et al (2004), Connor
& Craig (2004), Swainson3). A more systematic analysis of the most relevant community
features to the language variation of African American children is needed.

An alternative to correlating community characteristics to language features is the
notion of social networks (Milroy, 1987). It may be that there is a systematic, yet likely
complex relationship, between the linguistic variation found within communities that
differ in (l) SES and SEG levels and (2) the strength of intra- and extra-community ties
of their individual members. Social networks data could be added to studies of register-
shifting skill by surveying each participant or each participant’s parent(s) as to the
strength of his/her ties to other geographic and ethnic community members. In this
manner one could determine not only community-type of each participant but the strength
of community ties, of each participant’s ties to his or her community. Social networks
studies generally have concluded that the stronger the intra-community ties, the stronger

the use of the vernacular language variety (Milroy, 1987). One could expand this notion

 

3 Swainson’s data reported in Charity et a1. (2004), based on personal correspondence.

ll9

to include the study of register-shifting and an additional dimension of social networking
that accounts for the strength of the bond with external communities. It seems logical to
hypothesize that the strength of ties with the external, mainstream society would lead to
more familiarity with SMAE, the language of choice for school register.

Framing future studies of pragmatics. The results of this research clearly showed
that it is useful to socially contextualize the study of language use (i.e., pragmatics). A
theoretical framework that contextualizes language allows one to interpret differences in
meaning based on various uses of particular language forms. Such theoretical grounding
is preferable to methods that acknowledge linguistic variation outside of a theoretical
model within which to interpret situated meaning. For example, if one were to name a
linguistic form or set of forms (e. g., pronunciation of -in’ for -ing or use of habitual be),
which vary in terms of a given experimental dimension, the observed variation would
remain hollow unless there is some theoretical guideline for its interpretation.
Implications for Clinical and Educational Assessment

The controlled improvisation procedure used in this study could be firrther
developed to assess pragmatic knowledge in a standardized way. This method is needed
because there do not appear to be a broad range of models for assessing pragmatic
language skills. Few standardized methods exist for assessing pragmatics. The other
commonly used option involves analyzing spontaneous language samples for pragmatic
features such as comments, questions, answers, and so on (Lahey, 1988).

In terms of educational assessment this study legitimizes the general use of
dynamic assessment models. Using this principle, children could be expected to be

assessed on the “same” information in more than one context. That is, students might be

120

judged on whether they can flexibly apply the same information across various social
contexts. This competence would implicitly require registral components of language use.
Just as a speech-language pathologist or an English instructor may include both receptive
and expressive language measures in acomprehensive language evaluation, any
assessment of register-shifting skill should also include both modalities.
Implications for Clinical and Educational Instruction

The question arises as to whose job it is to teach registral competence. Delpit
(1995) argued that this is an obligation of school teachers, and that “language” instruction
should entail register-shifting skill. In fact, it was Delpit’s opinion that the primary
educational deficiency of non-mainstream students results from an incomplete
pedagogical framework that focuses only on linguistic codes. Such pedagogy neglects to
incorporate information about the social context in which language learning occurs.

Instruction in register-shifting skill could potentially fall into the scope of practice
of SLPs (Adler, 1988). According to ASHA’s Position on Social Dialects (ASHA, 1983)
the SLP’s primary and traditional role is to diagnose and treat language disorder. The also
can provide client-elective, instruction in accent modification. The latter type of service
delivery is meant to provide clients who speak non-mainstream (e.g., regional, social,
international) dialects, to acquire alternative language code such as SMAE. Part of this
elective service should include specific instruction on the socially-embedded meaning
underlying the use of a less-valued as opposed to a more-valued language variety.
Despite the potential for SLPs to provide accent modification services, institutional

constraints may prohibit them from doing so. Typically school-based SLPs are employed

121

by Special Education programs, which exclusively focus on the clinical diagnosis and
treatment of communication disorders.
Educational Policy

There is certainly value in continuing this line of research in terms of what it
might contribute to educational policy. Currently, most public school systems do not
acknowledge the structural differences between AAE and SMAE. Some have attempted
to do so, but have generally been met with very public disapproval. For example, in
December of 1996 the Oakland Unified School District, in an effort to improve their
students’ academic success, adopted the now famous “Ebonics Resolution,” (Amended
Oakland Resolution on Ebonics, 1997). This resolution acknowledged that the language
variety that African American children spoke at home was different than the variety
spoken at school and that it was the responsibility of educators to help them bridge this
divide. Largely due to confusion about the intentions of the resolution, it was met by an
intense media firestorm. This negative publicity contributed to the generally negative
public sentiment about acknowledging AAE in the education of African American
children.

It also created a policy backlash exemplified by the federal govemment’s ruling
that AAE-based education was ineligible for federal bilingual education funding. The
assumption was that that AAE (i.e., Ebonics) was not a language (CNN, December 24,
1996). Even though a negative public sentiment remains and policy is contradictory, there
is counter-evidence that educational programs, which bridge the gap between AAE and
SMAE structure, have been shown to be effective. Simpkins, Holt, & Simpkins’

“Bridge” program (Labov, 1993) has been shown to improve reading levels and Harris-

122

Wright’s “Bidialectal Communication” (CNN, December 19, 1996) has been shown to
increase test scores in reading and language arts.

The results of this research extend the concept underlying traditional bridging
programs which focus solely on language structure. The current research suggests that
resources should be made available to expose students to more than merely the structural
aspects of SMAE. English teaching policy should acknowledge the relevance of not only
the structural knowledge of language but also what is entailed in its pragmatic use as
reflected partly in register-shifting skill. In this sense, issues of relative meaning and
relative social appropriateness can be addressed directly. This type of policy requires
investigation of not only language forms (i.e., rules of grammar and pronunciation) but
also social context issues that help to interpret the socially embedded meaning of
linguistic structures. Once the social structures are laid bare, language users should be
able to make an informed decision about what it means to use one language variety or
register as opposed to another in a given context of use.

Future Outlook

Few, if any studies, have investigated African American child language from the
perspective of Systemic Functional Linguistics registral theory. This study made an initial
attempt to do so using one group of 3rd grade students. The context investigated was the
school setting, within which teacher and peer centered registers were investigated.
Though field and tenor were systematically investigated in the current study, both aspects
could be studied further in addition to mode, the third aspect of register which was not
addressed in this study. Efforts should be made to maximize authenticity of each test

condition in order to increase the validity of response values.

123

Further research could extend the current type of research design in order to study
other than African American cultural groups. In this manner one could investigate the
general applicability of registral theory to the use of school registers. Another potential
application is in the area of register development. That is, it would be fruitful to know
whether this theoretical model would apply across ages. Some previous research has
documented the general development of register skill in children (Andersen, 1990), but it
was not theoretically grounded in Systemic Functional Linguistics. Although Anderson
described the development of registral structures, a theoretical framework was not
provided to allow for a systematic, socially embedded explanation of these structures.
Furthermore, previous research on the development of register shift did not include
children from various cultural backgrounds.

The most promising reformulation regarding future research might be the
investigation of the relationship between community type, however defined, and social
networks (Milroy, 1987). The latter would add a systematic measure of the participants’
relationship with their communities. This type of analysis could provide a broader
framework for capturing the community dynamics that are relevant to individuals’
willingness to display register-shifting knowledge. Such a model could reveal how
children simultaneously evaluate (l) socio—linguistic resources and (2) community bonds
to make metalinguistic judgments about register choice. Results of such research would
likely have implications for diversification of residential, racial composition. Decreased
residential, racial segregation in combination with access to other forms of cultural
capital most likely increase the diversity of linguistic resources to which one has access.

Implications might also be made regarding the relationship between improved education

124

and improved cross-community communication. As cross-community relationships are
strengthened and multiplied it is likely that access to different language varieties and

registers will also increase.

125

APPENDICES

126

Appendix A
lnforrned Consent Procedures

Dear Parent/Guardian,

This study is being conducted in order investigate the manner in which third grade African American
children have learned to use the language of school instruction. This topic will be addressed by
investigating how children speak to peers as compared to teachers. A role playing task in which child
participants talk to realistic-looking peer and teacher puppets will be conducted. Your child is being asked
to participate in this study because he/she fits the participant eligibility criteria. That is, he/she is 8-9 years
old, African American, and a resident of the larger Metro Detroit Area. Because you are the adult
parent/guardian of your child you are responsible for giving or denying consent for participation in this
study.

As part of the study, you and your child will receive two main benefits: 1. free speech-language and
hearing screenings and 2. after participation in the study, you will receive a $50 check, to be sent to your
mailing address. Please indicate mailing address below:

 

 

 

In terms of general, social benefits, your participation in this study may contribute to the understanding of
how African American children from different types of neighborhoods learn to use language. In addition,
there may be implications for educational policies as they affect the academic success of African American
children. There are no known risks associated with participation in this study.

Data collection will require no more than one hour total. Speech/language and hearing screenings to be
conducted on a day prior to data collection and will last no longer than 45 minutes total. Regarding results
of the speech/language and hearing screenings, results will remain strictly confidential and will be shared
only with you, the parent. No benefits will results from the screenings in that you are responsible for any
follow-up services that you may wish to obtain.

Your participation in this research project will not involve any additional costs to you or your health care
insurer.

Your participation in this study is completely voluntary. You are free to refuse to participate and at any
time after you have agreed to participate are free to discontinue participation without penalty. Your child
may also refuse to answer any questions, participate in any aspect of the study.

In addition to your (the parent/ guardians’) signed consent as obtained by means of this consent form, the
researchers will also ask your child for their assent to participate. Specifically, before beginning our
conversation the investigator will say, “I brought some puppets so that we can play and talk together.
Would you like to play and talk with me?” Then specific reference will be made to the activities of the
study in which the investigator will ask whether it’s “okay that we first get an idea of how well you are
hearing and talking and then take a tape recording of your speech.”

If your child indicates in any way (either verbally or nonverbally) that he/she does not want to participate
the indication will be taken as a refusal to participate in the study and the examiner will not initiate the
research tasks.

If you have any questions about this study, please contact the primary investigator (Dr. Ida Stockman, (517)
353-6764, e-mail: stockmaléizmsuedu, mailing address: do Michigan State University, Department of
Audiology and Speech Sciences, East Lansing, MI 48824) or the secondary investigator (Johanna Boult,
(517) 355-9721, e-mail: boultjoh@msu.edu). If you have any questions or concerns regarding your child’s
rights as a study participant, or are dissatisfied at any time with any aspect of this study, you may contact —
anonymously if you wish —Peter Vasilenko, Ph.D. Chair of the University Committee on Research
Involving Human Subjects (UCRIHS) by phone: (517) 355-2180, fax: (517) 432-4503, e-mail:
ucrihs@msu.edu, or regular mail: 202 Olds Hall, East Lansing, MI 48824.

Your signature below indicates your voluntary agreement to allow your child to participate in this study.

 

 

Signature Date

127

Language data will be audio and video taped in order to be transcribed at a later time. By signing this
consent form you will give permission for the researchers to use audio and video taping procedures. In L
order to protect confidentiality, participant names will not be written on tapes. Rather, tapes will be

assigned a number (e.g. 1-20) and the list matching participants with participant numbers will be kept in a

locked file cabinet in the primary investigator’s locked university office. The list will be destroyed

immediately following the conclusion of the study. Tapes will be stored indefinitely in the possession of the

primary examiner. If you choose not to be audio or videotaped you may deny consent by refusing to sign

your permission on the line below. If you choose to withdraw from the study, tapes and data relevant to

your child will be destroyed. Regarding all aspects of this study, your and your child’s privacy will be

protected to the maximum extent allowable by law.

 

 

 

Signature (regarding audio- and videotaping) Date

128

 

1.

Appendix B

Schools/Neighborhoods Designated by SES/SEG groups

HiSS/LoSG (Farmington Hills, Oakland County)

Census Tract #: 1667

Census Tract segregation: 20% Black/African American, 72% White
Census Tract SES: Education:58% with Bachelor’s or higher; Median
Household Income: $95, 772

“School Name: Wood Creek Elementary [361]: 26% African American, 64%
White, 8% Asian.

RPL: 16%

Principal: Gloria Chapman

Address: 28400 Harwich, F armington 48334-3208

Phone: 248.785.2077

Evaluations Contact: Kris Gekeir

Phone: 248.489.3402

HiSS/HiSG (Southfield, Oakland County)

Census Tract #: 1600

Census Tract segregation: 50% Black/African American, 47% White.
Census Tract SES: Education: 52% with Bachelor’s or higher; Median
Household Income: $89, 303

”School Name: McIntyre Elementary [450]: 86% African American, 13%
White

RPL: 18%

Principal: Greg Kubasiewicz

Address: 19600 Saratoga, Southfield 48076

Phone: 248.746.7365

Evaluations Contact: Lynda Wood

Phone: 248.372.2505

LOSS/LoSG (Pontiac, Oakland County)
Census Tract #: 1417
Census Tract segregation: 27% African American; 49% White;
11% (Non-White) Hispanic
Census Tract SES:
Education: 47.9 High School; 3% Bachelor’s or higher
Median Household Income: $27, 641*
CF: 27.2%
**School Name: Whitman [352] 28% AA, 35% W, 30% H, 7% Asian
RPL: 84%
Principal: Iraida Garcia
Address: 125 Montcalm
Pontiac, MI 48342
Phone: 248.451.7950

129

Evaluations Contact: Dr. Brian Yancey
Phone: 248.451.6868; 248.431.3987

4. LOSS/HiSG (Pontiac, Oakland County)
Census Tract #: 1425
Census Tract segregation: 74% African American, 20% White, 5% Hispanic
Census Tract SES: '
Education: 66.5% High School; 6.1% Bachelor’s or higher
Median Household Income: $14, 898
CP: 30%
**School Name: Longfellow Elementary [328]; 83% AA, 10% W, 5% Asian
RPL: 870/0
Principal: Donna Delaney
Address: 31 N. Astor
Pontiac, MI 48324
Phone: 248.451.7810
Evaluations Contact: Dr. Brian Yancey
Phone: 248.451.6868; 248.431.3987
* Oakland County is one of the most affluent counties in the US. For this reason, tract
l4l7 was chosen as the “low SES, low SEG” tract even though the MHI exceeds the
operationalized cutoff level. In fact, this tract was the only one to match the
Operationalization for SEG and come close to matching the Operationalization for income.
This tract meets the other SES operationalizations.
** All specific school characteristics obtained from Great Schools.net at the web-site:
http://www.greatschools.net/modperl/browse_school/mi. Information obtained April 7,
2005.

Operationalizations

1. Low Socioeconomic Status or LoSS

A. below bottom third of Oakland MH income = $20, 635.

B. fewer than 15% have BA. or higher.

C. >20% live in poverty = '/2 that of traditional CP measure (i.e. >40% pop. in poverty).
D. RPL: > 2x state average = 72%

2. High Socioeconomic Status or HiSS

A. 1.25x or greater than Oakland MH income = $77, 384.

B. greater than 50% have BA. or higher.

C. 40% or > have incomes greater than 1.6 times Oakland’s MHI.
D. RPL: < or = state average = 36%

3. High Segregation or HiSG
50% or greater African American population.

4. Low Segregation or LoSG

Less than three times the county average % African American population
i.e. < 30% African American.

130

Oakland (2000) US. Census Stats

89.3% with high school education

38.2% with Bachelor’s or higher

10.1 % African American

82. 8% White

$61, 907 MHI

CP: >20% live in poverty (1/2 that of traditional CP measure)
CA. 40% or > have incomes greater than 1. 6 times Oakland’ 3 MHI.
RPL: State Average = 36%

131

Appendix C
Direct and Indirect Requests According to Modal Type

Direct Request

1. [No Modal Used] Direct imperative.
e.g. Give me a pencil.

2. [No Modal Used] Need statement.
e. g. 1 need my toothbrush.

Indirect Requests.
3. Use of modal auxiliary: can, could, will, would. [more indirect than the
imperative] e. g. Can/Could I have a pencil; Will/Would you give me a pencil?

4. Use of embedded modal auxiliaries. [more indirect than simple aux use]
e. g. Do you think I could borrow a pencil?

132

Appendix D
Data Collection Scripts
1. Read Assent Procedures: 1 brought some puppets so that we can play and talk
together. Would you like to play and talk with me? Can we take a tape recording of

your speech?

Props: puppets, line drawings.

[M'—

2. lntroduction of puppets. All participants will receive the same introduction of puppets.

Mrs. Jones: Hello . I am Mrs. Jones. I am a teacher who works at your school.

Let’s pretend that I am your teacher. We are at school having a lesson.

[Switch to peer puppet]

Elijah/Breanna: Hi . I am Elijah/Breanna. I am a kid in your class. Let’s
pretend that I am your best friend. We’ll sit next to each other on the carpet and play
together during free time.

2. Introduction of stimulus items [Tasks and modalities to be presented in random

order.] See Appendix F for Data Collection Forms with specific instructions for delivery

of each task.

1. Receptive Lexical Semantics Task (Speaker Identification through Pointing)
You are going to look at a picture then touch to the puppet who you ’d say each
thing to. Your job is to touch the correct puppet.
[Puppets are presented in plain view directly in front of the participant and within
reach (about l/12 feet) from the participant for pointing tasks. Give the direction:
Touch the puppet who you ’d say each thing to. ” Let’s try one: 1. That’s a cool
shirt. Let’s try one more: 2. When is the homework due?
(Probes may include repetitions or additional queries “Who would you say it
to?” or requests for action, “Touch the one who you ’d say it to. ”)]

2. Expressive Lexical Semantics Task (Cloze Task)

Now, I will read you a sentence with a word left out. Your job is to think of the
word your teacher or your friend would use to fill in the blank. Let’s practice one. I
really your new outfit. What word would your friend say? What word
would your teacher say? [Line drawing of scenario provided; written version of
sentence also provided]

3. Receptive Request Forms Task (Speaker Identification through Pointing)

Now I will tell you something about a picture. There is a problem in
each picture. The problem is either your teacher’s fault or your friend ’s fault. You
are going to look at a picture then touch to the puppet who you ’d say each thing to.
Your job is to touch the correct puppet.

133

 

[Puppets are presented in plain view directly in front of the participant and within
reach (about 1/12 feet) from the participant for pointing tasks. Give the direction:
Touch the puppet who you ’d say each thing to. ” Let’s try one: 1. That’s a cool
shirt. Let’s try one more: 2. When is the homework due?
(Probes may include repetitions or additional queries “Who would you say it
to?” or requests for action, “Touch the one who you ’d say it to. ”)]

5. Expressive Reguest Forms: What do you say (Reguestsl

Now I will tell you something about a picture. There is a problem in each
picture. The problem is either your teacher’s fault or your friend ’s fault and you
want the problem fixed. Remember your job is to listen then tell me what you say to
get the problem fixed. [Line drawings of scenarios are provided] Let’s try one. Your
friend spilled some milk. What do you say?

134

 

Appendix E
ABILITIES Index

(See following two pages.)

135

..x0: 2: 0. 2:: 0:0 80:: :0 ::0:

0:. :0 :::: .:0:0::_: E 5:> ::0 0:05:00:

:0 20:59. .5650 2:8 E .00Em 0:2:

0: 200:0. 23:2 03 H ::0: :0 0:05:00:

:0 30:59. :0 00:w0: 0.: :0::0_:E :::.

:8: :0:0 E __x._ :: 00:2: 0:52 .250: 0:.

E .0: E 0:0: 20:2: m_:_50 0.: :: .0000: :0:
5m: :0::0: 0:: 00.002: ::::0 0:. :::: :::2:
2:52 0:0: 0.00.2: m.:_50 0:. 30:: 055:.

$00k. 0.02.2: 552.0%

0%: 2:2.

0:. :0 :0::__:0 :0:0 0. :5. 0:::E00 :::
.20. ::3 :52: ::: 2:058: 0030: :::0
0:. 2:3 0:: 30:: 55:. .:00:0: ::: 055:
0. :0::5: ::::0 0.: 300:0: w:::: 05:.
302.0000“

M: 9.35:: 002055 5:30:02.—

.0m: 05:: 2: :0

5550 :0:.0 0: :5. 0:23:00 ::: 0:050 ::3 080508800
::: :::.::0::: 0: 5:5: A::::0 2: 30:: 055:.

209:0: 0:203 .w0:::m0m .293 mE0=:. :2: .050 02:3 E
0.:0_::EE00 0: 3:802: :0:0—0E w:::: :::. 0:050 :23
0.:050EE00 0. 5:5: ::::0 0:. :0: 0:0 ::: £050 :::2
.:0::.. 0. 5:5: 0.250 2: :0: 0:0 0:2: 0:: mw:::: 03:.

3.0.20 :23 95:02:08.3“:
0w w:_::2m..0::3: :0_.:0_..........00 22.22.02.—

.0.0 :0:

.32: .wE:::c :::: .wEx00: .wEE:0:0m .wEE: .wEEm:
0:.._0E 52: 8320: 530:: 0:. 08:52:93: .:0::2:
33:5:2: : E £050 0. 02:0: 0. 5:5: m.:_50 0.: 0. :0:0:
::0: _:_00m .5320: 252:: :0 0:350:32: :0: 0:0

::: 0:50 :::00: :0: 0:0 .:0:: 0.5: E 0:2: 0:: 23:2 03:.

3.3: 3:80: 3.5.00

.:E: : :0 0m: 0: 0:: :::0
0:. :::. :::0E 05305:: :::0:0:n::
: :0 0:00: < 50.280: 0:5:

59: ::: ::0_ 0:00m 0053.0: 5::
E 000. ::: :::: .m:::: :0: :0 :5
0m: 0. 5:5: 0.250 0.: 30:: 055:.

395 w 2...}. £2.03 :0 00.3: 00...:

.E:0: 0.: :0 :0:: 0:. :0
m5. 0.:0_:E .:_: wE::0: : 0.0:: :::0
0:. :: ::: wE::0: : 30:23 wE::0:
0.250 0:. 0::m .wE::0: 0: :::

:::0 0:. :::. :::2: 3.00: :::0:0:n::
c :0 0:00: < 50:22:00. ::0 :0:0 :0:
mE::0: 0:00m 0053.0: :::»:00 E
::0: 0. 5:5: 9:50 2: 50:: 055:.

35:80: 5:63.

09.29. 2.0::

00:05:02.: 3:02:50 0:302: 0: .....0: 0:... :0 0.0:: 0.: ..0 000% 0... 0m: 5:... 00> 9:2... :0:0 9.2:... E :0» .020: 0. 30:0: 00:20:30 .09.
0.0:: 0:: 00.6.20 .850 0: :0:0..500 :::0 0:: 2.0:: 0.5:: .95.... :0:0 9.22.. 5 .525: :0 0.0.: 2.00:0...— ..0 02.0.2.0 922.0%... : ::: $2.5:
“.250 0:: 2.0:: 000.20.... 0.00.: 9.20080: 30:02:33 N £2.50 _:...:0.. 95:02.0: _ :23 0 0. _ :0 0.00.: : ..0 0::... 0...: :0:: :0:0 E 09:23—

136

3.0.00. :0 :::... .:0>08

250 0:. 30: :23 0:0::0.E :00:0:0:.2: 0:2:
:02: 30: 2:062 200:: :w:::~. .5.8:0:0:
::0. :0 8:: ::: 0:5: :0>:..0 00.: ::0...::00
0:20E 05.02.: E :00:0:0::.: ”.00: :20

:0 0.22. 20.0 00.: ::0...::00 0:20E 8:0:

E :00:0:0::.n. :22. ::0: 0:. :0 0:20.02:

:0 .2.::: 8:0: :23 :0.:.00::: :00:0:0::.: 0:
0:: 0:05 :::. :::08 .:8:0Z :60: :_250 2:
:0 0:20.22: ::: 8:0: 2: ::00.:0: w:::: :5 ...

0:30.25
::: ......0... ::0:— .0..:..@ ::23m ......2022m

8:0: 0:. :0 0.0::

0.: :0 :5. 0::0.:E .:0:::.w :0:: 250 0:. :. .:0:::.w 50:23
:0.:.> :250 0.: 0.3. .:0.:.> 0: ::: 250 0:: :::. :::08
3:0. :::0:0:..: 0 :0 0:00: < 0:0 59: w ::0. 0.: ::0: 0:00m
.:0.2>..0: :::»:03 E 00: 0. 5:5: :_250 0.: 2.0:: 0.2:...

.833 ... u : m

.80 .:00::0 52.02:»: ::_:0::8
.:0:0::.: 302.20: 0:20E ::8
:8050:.. :2:0: chm:O .:0.2>:0:
28: :80502. :2:0: :053 0.
00:m0: 0:. w:::0.:E 0::8 0: 200::
:w:::: 8050:: :._:0: : :. 0:2: :.
::: :5 2::

: :0: .:0...5 :0::02: a. :8050:..
:2:2. .:::: 0:. :::08 .::::07.

.23: :58: p22: 2.: as: .52

2:8: ...:30:
5.3.. .82.... :o 2.3....

3:00: m 2:20.23:

[37

Appendix F
Data Collection Forms

(See following pages)

138

 

 

Participant Name l
Participant Number
Location

Date of Data Collection

 

 

Data Collection Form #1: Receptive Request Forms

Directions: Both puppets should be sitting directly in front of the participant at a distance
of about 1 '/2 feet away. Show each drawing (with a number which matches the stimulus
item) as you say each stimulus item aloud to the participant. Each item should be
prefaced by the carrier phrase “Who would you say this to?” Put an “X” in the column
which indicates where the participant pointed in response to each stimulus item.
Directness will be calculated only by scorers.

Trial items:
Trial Item #1: Who would you say this to? Your friend, your teacher, or both? Wipe

that stuff off your shirt.
Trial Item #2: Who would you say this to? Your friend, your teacher or both? Could I
have my homework back?

Now let’s try some more.

 

Where Does Participant Directness
Point?

Stimulus Item Teacher Peer Direct Indirect

puppet Puppet

 

 

13. Could I have my backpack?

20. Help me open the windows.

 

 

3. Could you move your umbrella?

 

17. Will you push me on the swing?

 

8. I need that bag moved.

 

11. Can I play with your new toy?

 

14. Give me my backpack.

 

4. Move your umbrella.

 

18. Push me on the swing.

 

2. I need a pencil.

 

5. Could you clean up your books.

 

9. May I have a cupcake?

 

12. Give me your new toy.

 

 

 

 

 

1. Can I have a pencil?

 

 

 

I39

 

7. Can you help me move that bag.

 

6. Clean up your books.

 

 

16. Help me with my book rmm.

 

10. I want a cupcake.

 

windows?

19. Could you help me open the

 

 

report?

15. Could you help me with my book

 

 

 

 

 

 

Items 1-20 have been randomized and should be delivered to all participants in the,
standard order shown on the Data Collection Form. Scoring should be based on
comparison of items 1 & 2, 3 &4, 5 &6, etc. Place an X in the appropriate column.

 

Item
Numbers

Both teacher

Both Peer

Expected
Shifts: Teacher
Direct to Peer
Indirect

Unexpected

Shifts: Peer
Indirect to
Teacher Direct

 

1&2

 

3&4

 

5&6

 

7&8

 

9&10

 

ll&12

 

l3&l4

 

15&l6

 

l7&18

 

l9&20

 

Actual
Shifts

 

 

 

 

 

Possible
Shifts

 

 

1,0

 

 

Percent
Shift

 

 

 

 

 

 

 

 

 

 

 

140

Participant Name

 

Participant Number
Location

 

Date of Data Collection
Data Collection Form #2: Recegtive Lexical Semantic Form

Directions: Both puppets should be sitting directly in front of the participant at a distance of
about 1 '/2 feet away. Show each drawing (with a number which matches the stimulus item) as
you say each stimulus item aloud to the participant. Put an “X” in the column which indicates the
participant pointed in response to each stimulus item. Shift will be calculated only by scorers.

Trial items:

Trial Item #1: Who would you say “jam ” to when you ’re talking about good music?
Your friend, your teacher, or both?

Trial Item #2: Who would you say “jam” to when you’re talking about jelly? Your
friend, your teacher, or both?

Now let’s try some more.

 

Where Does Participant Shifi

 

Point?
Stimulus Item Teacher Child Y N
ptmpet PUPPet

 

" 20. Who do you use the word crib with when

ou mean the bed where a baby sleeps?

 

8. “Who do you use the word sugar with when
you mean the thing you eat that’s sweet, white

and powdery?”

 

10. “Who do you use the word bright with to

describe a strong light?”

 

7. “Who do you use the word sugar with when

ou mean kisses?”

 

12. “Who do you use the word a tote with when

on mean a bag to carry something in?”

 

6. “Who do you use the word poor with when

you mean havingno money?”

 

2. “Who do you use the word kitchen with
When you mean a room in the house where food

is cooked?’

 

 

 

 

 

 

 

 

who do you use the word tight with when

 

l4l

 

you mean closely fitting?”

 

13. “Who do you use the word bad with when

you mean something really good?”

 

I. “Who do you use the word kitchen is the
place on the back of the neck where hair grows

short and curly?”

 

17. Who do you use the word nap with when

you mean a tight curl or knot of hair?

 

16. Who do you use the word hip with when
you mean the place where your body attaches to

your leg ?

 

11. “Who do you use the word to tote with

when you mean to cagy something heavy?”

 

3. “Who do you use the word tight with when

you mean really cool?”

 

14. “Who do you use the word bad with when

you mean bad?”

 

19. Who do you use the word crib with when

you mean house?

 

5. “Who do you use the word poor with when

. you mean skinny or sickly?”

 

15. Who do you use the word hip with when

you mean cool?

 

9. “Who do you use the word bright with when

you mean light skinned?”

 

l8. Who do you use the word nap with

 

 

when you mean a short rest?

 

 

 

 

Items 1-20 have been randomized and should be delivered to all participants in the,
standard order shown on the Data Collection Form. Scoring should be based on
comparison of items 1 & 2, 3 &4, 5 &6, etc. Place an X in the appropriate column.

 

Item Numbers Both Teacher Both Peer

 

 

 

 

Expected
Shifts: Teacher
SMAE to Peer
AAE

 

Unexpected
Shifis: Peer
SMAE to
Teacher
AAE

 

142

 

 

Actual Shifts

 

Possible

 

 

Percent Shift

 

143

Participant Name
Participant Number
Location

Date of Data Collection

 

 

 

Data Collection Form #3: Expressive Request F orn_i§

Directions: The puppet which matches participant identity (boy or girl) should be sitting
on the child’s lap. Help the child adjust his/her hand in the puppet’s mouth so he/she can
easily move it. Interlocutor puppets should be hidden behind screen and brought out in
front of the participant as each stimulus item dictates. When brought out, the puppet
should sit directly in front of the participant at a distance of about 1 '/2 feet away. Show
each drawing (with a number which matches the stimulus item) as you say each stimulus
item aloud to the participant. Put an “X” in the column which indicates the participant
pointed in response to each stimulus item. Directness will be calculated only by scorers.

Trial items:

Trial Item #1 : Your friend left some crumbs on the table after snack. What do you say
to your friend?

Trial Item #2: Your teacher left some crumbs on the table afler snack. What do you say
to your teacher?

Now let’s try some more.

 

Directness

 

Stimulus Item Participant Direct Indirect
Response. Record
word-for-word

online

 

12. Your friend brought a brand
new toy to school. You want to

play with it. What do you say?

 

20. It’s hot inside the classroom
and you want your friend to help
you open the windows. What do
you say?

 

15. You need some help on your
book report so you ask your
teacher. What do you say?

 

5. Your teacher has books all over
the floor and it’s hard to walk

 

 

 

 

 

 

 

144

 

around. You want her to clean

them up. What do you say?

 

16. You need some help on your
book report so you ask your friend.

What do you say?

 

3. Your teacher left her umbrella
in front of the door. You need her
to move it so you can get into the

classroom. What do you say?

 

4. Your friend left his/her umbrella
in front of the door. You need
him/her to move it so you can get
into the classroom. What do you

say?

 

6. Your friend has his/her school
books all over the floor and it’s
hard to walk around. You want
him/her to clean them up. What do

you say?

 

2. You forgot all your pencils at
home but you need one so you can
do your schoolwork. What do you
say to your friend?

 

17. You want your teacher to push
you on the swing at recess. What
do you say?

 

10. It’s not snack time yet but
today your friend brought a special
treat for the class. . .cupcakes! You
, can’t wait to eat one so you ask
your friend if you can have one

now. What do you say?

 

 

9. It’s not snack time yet but today
your teacher brought a special treat

for the class...cupcakes! You

 

 

I45

 

 

 

 

 

can’t wait to eat one so you ask the
teacher if you can have one now.

What do you say?

 

8. Your friend is busy doing

something else but you need.

his/her help to move a heavy bag
of toys off of the carpet so you can
play there. What do you say?

 

14. The bell just rang and it’s time
to go home from school.
Everybody is getting their things
ready to go and you see your
friend holding your backpack.
What do you say?

 

19. It’s hot inside the classroom
and you want your teacher to help
you open the windows. What do

you say?

 

11. Your teacher brought a brand
new toy to school. You want to

play with it. What do you say?

 

18. You want your friend to push
you on the swing at recess. What

do you say?

 

13. The bell just rang and it’s time
to go home from school.
Everybody is getting their things
ready to go and you see your
teacher holding your backpack.
What do you say?

 

 

7. Your teacher is busy doing
something else but you need her
help to move a heavy bag of toys
off of the carpet so you can play
there. What do you say?

 

 

I46

 

 

 

 

 

1. You forgot all your pencils at
home but you need one so you can

do your schoolwork. What do you

 

say to your teacher?

 

 

 

 

Items 1-20 have been randomized and should be delivered to all participants in the,
standard order shown on the Data Collection Form. Scoring should be based on
comparison of items 1 & 2, 3 &4, 5 &6, etc. Place an X in the appropriate column.

 

Item Numbers

No Shift: Both
Direct

No Shift: Both
Indirect

Expected
Shifts: Teacher
Direct to Peer
Indirect

Unexpected
Shifts:

Peer
Indirect to
Teacher
Direct

 

1&2

 

3&4

 

5&6

 

7&8

 

9&10

 

ll&12

 

l3&l4

 

15&16

 

l7&18

 

l9&20

 

Actual Shifts

 

 

 

Possible

Shifts

 

 

 

Percent Shift

 

 

 

 

 

 

 

 

 

 

147

 

 

Participant Name

 

Participant Number

 

 

Location

 

Date of Data Collection
Data Collection Form #4: Estsive Lexical Sqnyantics Form

Directions: The puppet which matches participant identity (boy or girl) should be sitting
on the child’s lap. Help the child adjust his/her hand in the puppet’s mouth so he/she can
easily move it. Interlocutor puppets should be hidden behind screen and brought out in
front of the participant as each stimulus item dictates. When brought out, the puppet
should sit directly in front of the participant at a distance of about 1 '/2 feet away. Show
each drawing (with a number which matches the stimulus item) as you say each stimulus
item aloud to the participant. Give each stimulus item remembering to say Blank in the
position of the blank space. Immediately following each item say What word would you
say to your teacher(friend? Record the word the participant uses for each concept. Shift
will be scored only by scorers.

Trial items:
Trial Item #1 : This music is . What word would you say to your friend?

Trial Item #2: This music is . What word would you say to your teacher?

Now lets try some more.

 

Participant Response. Record Shift

word-for-word online

 

Stimulus Item Yes No

 

16. Look at this boy and his Grandma.

‘ Grandma’s groceries are heavy.

“He is going to the groceries for

his grandma.”

What word would you say to your

teacher?

 

ll. Look at this little girl’s hair. “You

can tell she needs a perm because her
looks uncombed.”

What would word would you say to your

teacher?

 

 

 

 

 

5. Look at these girls.

 

I48

 

 

“The girl has light eyes.”

What word would you say to your
friend?

 

4. Look at the baby sitting with her mom.

“The mom is giving her baby some

9,

What word would you say to your
friend?

 

19. Look at this girl’s hair.

“The in her hair make it hard to
brush.”

What would word would you say to your

teacher?

 

13. Look at this lady. She doesn’t eat
right.

“If she ate better she wouldn’t look so

9,

What would word would you say to

your teacher?

 

 

 

 

10. This boy is going to play soccer. He

, lives around the comer.

“He needs to stop by the to

change his shoes.”

What would word would you say to
your friend?

 

 

149

 

 

 

 

6. Look at this boy and his Grandma.

Grandma’s groceries are heavy.

“He is going to the groceries for

his grandma.”

What word would you say to your
friend?

 

8. This boy always wears the new styles.

,3

“He is really .

What word would you say to your
friend?

 

15. Look at these girls.

“The girl has light eyes.”

What would word would you say to your
teacher?

 

7. Have you seen the new Street Fighter?

,3

“That videogame is really .

What word would you say to your
friend?

 

3. Look at this lady. She doesn’t eat
right.

“If she ate better she wouldn’t look so

9’

What word would you say to your
friend?

 

 

 

18. This boy always wears the new

 

 

ISO

 

 

 

 

styles.

9’

“He is really .

What would word would you say toyour

teacher?

 

1. Look at this little girl’s hair.

“You can tell she needs a perm because

her looks uncombed.”

What word would you say to your
fi’iend?

 

2. Look at this man sitting in his brand

new car.

9,

“His new car is

What word would you say to your
friend?

 

20. This boy is going to play soccer. He

lives around the comer.

“He needs to stop by the to
change his shoes.”

What would word would you say to your

teacher?

 

 

12. Look at this man sitting in his brand

new car.

“His new car is

What would word would you say to your

teacher?

 

 

151

 

 

 

 

 

17. Have you seen the new Street
Fighter?

,9

“That videogame is really .

What would word would you say to your

teacher?

 

14. Look at the baby sitting with her

mom.

“The mom is giving her baby some

,9

What would word would you say to

your teacher?

 

 

9. Look at this girl’s hair.

“The in her hair make it hard to

brush.”

What word would you say to your
friend?

 

 

 

 

 

Items 1-20 have been randomized and should be delivered to all participants in the,
standard order shown on the Data Collection Form.
Calculate shift by comparing items 1 and 11, 2 and 12, 3 and 13, 4 and 14, 5 and 15, 6
and 16, 7 and 17, 8 and 18, 9 and 19 and 10 and 20.
Rewrite words used in the first two blank columns. Then, place an X in the appropriate,

subsequent columns.

 

 

 

 

 

Item Word Word No Shift: No Shift: Expected Unexpected

Numbers used used Both Both Shifts: Shifts: Peer
with with AAE SMAE Teacher SMAE to
Peer Teacher SMAE to Peer Teacher

AAE AAE

1 & ll

2 & 12

3 & 13

 

 

 

 

 

 

 

152

 

 

 

4&14

 

5&15

 

6&16

 

7&17

 

8&18

 

9&19

 

10&20

 

Actual
Shifts

 

 

Possible
Shifts

.10

 

 

Percent
Shift

 

 

 

 

 

 

 

153

 

 

 

 

 

 

Data Collection Form: Task #5
Self-reflective Survey. Immediately after completion of the language tasks, each
participant will be given the following five question survey regarding his/her own
thoughts about register shifting. The survey will be given orally, face-to-face and
questions will be asked by the primary investigator or a research assistant. Participant
responses will be audiorecorded and analyzed at a later time.
Questions regarding language use:
I. Do you have a good friend?
2. What is his/her name?
3. Do you talk the same with your teacher as with your good friend?
a. If no: Tell me all the ways you talk different with your good friend and
your teacher. Then go to #4.
b. If yes: stop here.

4. Tell me all the reasons why you talk different.

154

 

Appendix G
Lexical Semantics Screener

Directions:
Ask potential participants: Do you know this word?

 

Screen Item Response

 

Yes No

 

l7. stool

 

15. furniture

 

20. mermaid

 

10. crib

 

16. instruments

 

12. ostrich

 

13. pineapple

 

1. kitchen

 

l9.dentist

 

7. bad

 

18. tractor

 

9. nap

 

1 1. binoculars

 

14.drinks

 

5. bright

 

6. tote

 

4. sugar

 

2. tight

 

3.fl)or

 

 

 

 

 

8. hip

 

155

 

Appendix H

Randomized Number Sets Used For Construction of Data Collection Lists (Urbaniak,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1997-2005)
8&1 8&2 8&3 8&4” 8&5

13 20 12 16 17
20 8 20 11 15
3 1O 15 5 20
17 7 5 4 10
8 12 16 19 16
11 6 13 12
14 4 10 13
4 4 6 1
18 13 8 19
1 17 15 7

5 17 10 7 18
16 9

12 11 18 11
1 3 14 1 14
7 14 19 2 5
6 19 11 20 6
16 5 18 12 4
10 15 13 17 2
19 9 7 14 3
15 18 1 9 8

 

 

 

 

 

156

 

Appendix 1

Reduced Price Lunch Screener

Does qualify for free or reduced price lunch?

 

yes 110

157

 

Appendix J
Hearing Screening

Please indicate a pass with an X and a fail with an F.
Screening conducted at 20dB.

 

 

 

Right Ear

lOOOHz 2000Hz 4000Hz
Left Ear

IOOOHz 2000Hz 4000Hz

 

 

 

158

 

Appendix K
Additional Statistical Analyses

Table l 6. Analysis of Variance in Please Used with Teacher across Community Groups

 

 

Sum of Mean
Source Squares df Square F p =
Between Groups 4.48 2 2.24 .18 .84
Within Groups 375.77 30 12.53

 

Table 1 7. Analysis of Variance in Address Forms Used With Peers across Community
Groups

 

 

Sum of Mean
Source Squares df Square F p =
Between Groups 16.53 2 8.27 .934 .40
Within Groups 265.35 30 8.85

 

Table I 8. Analysis of Variance in Self-Report of Register Use across Communities

 

 

Sum of Mean
Source Squares df Square F p =
Between Groups .214 2 .107 .404 .672
Within Groups 6.643 27 .266

 

Table [9. Analysis of Variance in MLU of Task 5 Responses across Communities

 

 

Sum of Mean
Source Squares Df Square F p =
Between Groups 12.40 2 6.20 .76 .48
Within Groups 187.49 25 8.15

 

159

Appendix L

Self-Reflective Report of Register Use Questionnaire

I. Do you have a good friend?
2. What is his/her name?
3. Do you talk the same with your teacher as with your good friend?

a. If no: Tell me all the ways you talk different with your good friend and

your teacher. Then go to #4.

b. If yes: stop here.

4. Tell me all the reasons why you talk different.

160

 

REFERENCES CITED

The American Speech-Language Hearing Association (ASHA) (1997). Retreived March
16, 2007. http://www.asha.org/public/speech/development/child hear talk.htm

The American Speech-Language Hearing Association (ASHA) (2002). ASHA 2002
Desk Reference, 4, Audiology and Speech-Language Pathology. Chapter 5:

Guidelines for screening for hearing impairment—School-age children 5 through
18 years, pp. IV, 364-IV, 366.

The American Speech-Language Hearing Association (ASHA) Committee on the Status
of Racial Minorities (1983). Position Statement on Social Dialects, 3, 77-82.

Amended Oakland Resolution on Ebonics. (1997). Retreived August 1, 2007.
http://linguist.emich.edu/topics/ebonics/ebonics-resZ.htrnl .

Andersen, E. (1990). Speaking with Style: The Sociolinguistic Skills of Children, New
York: Routledge.

Austin, J. L. (1962). How to Do Things with Words. New York: Oxford University Press.

Ayers, W. (2004). Teaching toward Freedom: Moral Commitment and ethical action in
the classroom. Boston, MA: Beacon Press.

Baratz, J. C. (1969). Teaching reading in an urban Negro school system. In J .C. Baratz
and R. Shuy (Eds), Language Differences: Do they interfere? (92-116). Newark,
DE: International Reading Association.

Bates, E., & MacWhinney, B. (1979). A functionalist approach to the acquisition of
grammar. In E. Ochs & B. Schieffelin, (Eds), Developmental pragmatics. New
York: Academic Press.

Bloom, L. (1974). Talking, understanding, and thinking: Developmental relationship
between receptive and expressive language. In R. L. Schiefelbusch and L. Lloyd
(Eds), Language perspectives-Acquisition, retardation, intervention (pp. 285-
312). Baltimore, MD: University Park Press.

Bloom, L. & Lahey M. (1978). Language Development and Language Disorders. New
York: John Wiley and Sons.

Bourdieu, P. (1977). Outline of a Theory of Practice. Cambridge, MA:
Cambridge University Press

161

 

Bourdieu, P. (1983, 1991). Language and Symbolic Power. Cambridge, MA:
Harvard University Press.

Bourdieu, P. (1993). The Field of Cultural Production: Essays on Art and Literature.
New York: Columbia UP.

Bourdieu, P. & Wacquant, L. J. D. (1992). An Invitation to Reflexive Sociology.
Cambridge: Polity.

Briggs, X. (2002, February). Social Capital and Segregation: Race, Connections and

Inequality in America. JFK School of Government, Harvard University Faculty
Research Working Papers Series.

Brown, P. & Levinson, S. C. (1987). Politeness: some universals in language usage.
New York: Cambridge University Press.

Brown, R. (1973). A First Language: the Early Stages. Cambridge, MA:
Harvard University Press.

Charity, A. H., Scarborough, H. S. & Griffin, D. M. (2004). Familiarity with school
English in African American children and its relation to early reading

achievement. Child Development. 75 (5), 1340-1356.

Chomsky, N. (1965). Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.

Clark, E. (2002). First language acquisition. New York: Cambridge University Press.
Clark, H. (1996). Using language. Cambridge: Cambridge University Press.

Clinton, W.J. (1997, May). An Apology for Stuay Done in Tuskegee. Speech presented in
the East Room of the White House, Washington, DC.

CNN. (Dec. 19, 1996). Ebonics Taught as ’Home Speech' in Georgia School. Retrieved
August 1, 2007. http://www.cnn.com/US/9701/l7/home.speech/ .

CNN. (December 24, 1996). Education Secretary: N0 Federal Funds for 'Black English '.
Retreived August 1, 2007. http://cnn.com/US/9612/24/briefs.
pm/black.english.html.

Connor, C. M. & Craig, H. K. (2006). African American preschoolers’ language and
emergent literacy skills, and use of African American English: a complex

relation. Journal of Speech, Language, and Hearing Research. 49, 771—792.

Corsaro, W. (1979). Young children’s conception of status and role. Sociology of
Education. 52, 15-79.

162

 

Craig, H. K. & Washington, J. A. (2004a). Grade related changes in the production of
African American English. Journal of Speech, Language, and Hearing Research,
47, 450-463.

Craig, H. K. & Washington, I. A. (2004b). Language variation and literacy learning. In
C. A. Stone, E. R. Silliman, B. J. Ehren, & K. Apel (Eds), Handbook of language
and literacy development and disorders (pp. 228-247). New York: Guilford Press.

Currie, J. & Thomas, D. (1995). Does Head Start make a difference?, American
Economic Review. 85 (3), 341-364.

Delpit, L. (1995). Other People ’3 Children: Cultural conflict in the classroom. New
York: The New Press.

Dillard, J. L. (1972). Black English: Its History and Usage in the United States. New
York: Vintage Books a division of Random House.

Dillard, J. L. (1977). Lexicon of Black English. New York: The Seabury Press.

Donahue, M. (2000). Influences of school-age children’s beliefs and goals on their
elicited pragmatic performance: Lessons learned from kissing the Blarney Stone.
In L. Menn & N. Bernstein Ratner (Eds) Methods for Stuaying Language
Production (pp. 353-368). Mahwah, NJ: Erlbaum.

Duchan, J. (1984). Language assessment: The pragmatics revolution. In R. Naremore
(Ed.) Language Science. San Diego, CA: College-Hill Press.

Eggins, S. (1994). An Introduction to Systemic Functional Linguistics. London, Pinter
Publishers.

Ervin-Tripp, S. (1970). Discourse Agreement: How Children Answer Questions.
Cognition and the Development of Language. J. R. Hayes. New York: John Wiley
& Sons: 79-108.

Ervin-Tripp, S. (1976). Is Sybill There? The structure of some American English
directives. Language in Society. 5, 25-66.

Ervin-Tripp, S. (1977). Wait for Me, Roller Skate. Child Discourse. S. Ervin-Tripp and
C. Mitchell-Keman. New York: Academic Press Inc: 165-188.

Ervin-Tripp, S. & Gordon, D. (1986). The development of requests. In R Schiefelbusch

(ed.), Language Competence: Assessment and intervention. New York: College
Hill Press.

163

Ervin-Tripp, S. & Mitchell-Keman, C., Eds. (1977). Child Discourse: Language,
thought and culture, advances in the study of cognition. New York: Academic
Press Inc.

Ferguson, C. A. (1959). Disglossia. Word, 15, 325-400.

Gardner, M. F. (1990). Expressive One- Word Picture Vocabulary T est-R. Novato, CA:
Academic Therapy Publications.

Gethin, A. & Gunnemark, E.V. (1996). The Art and Science of Learning Languages.
Oxford, England: Intellect Books.

Goffman, E. (1967). Interaction Ritual: Essays on F ace-to-F ace Behavior. New York:
Doubleday Anchor.

Goldfield, B. & Snow, C. E. (1992). Who's cousin Arthur's daddy?: The acquisition of
knowledge about kinship. First Language, 12, 187-205.

Green, L. (2002). African American English: A Linguistic Introduction. New York,
Cambridge University Press.

Gregory, ST. (2000). Strategies for Improving the Racial Climate for Students of Color
in Predominately White Institutions. Equity & Excellence in Education.
33(3), 39-47.

Grice, H. P. (1968). Utterer's meaning, sentence-meaning, and word-meaning.
Foundations of Language, 4, 225-242.

Grice, H. P. (1969). Utterer's meaning and intentions. The Philosophical Review, 78,
147-177.

Halliday, M. A. K. (1970). Language structure and language function in J. Lyons (ed.)
New Horizons in Linguistics, Harrnondsworth, England: Penguin.

Halliday, M. A. K. (1978). Language as Social Semiotic, London, Edward Arnold.

Hart, B. & Risley, T. R. (1995). Meaningful Differences in the Everyday Experience of
American Children, Baltimore, MA: Brookes Publishing.

Heath, S. B. ( 1983). Ways with Words. Bath, Great Britain: The Pitman Press.
Heath, S. B. (1986). What no bedtime story means: narrative skills at home and

school. Language Socialization Across Cultures. B. B. Schieffelin and E. Ochs.
New York: Cambridge University Press: 97-126.

164

 

Homberger, N. H. & Chick, J. K. (2001). Co-Constructing School Safetime: Safetalk

Practices in Peruvian and South African Classrooms in Voices of Authority:
Education and linguistic difference, (pp. 31-55) M. Heller and M. Martin-Jones
Eds. Westport, CT: Ablex.

Houck, B. & Higley, N. (2006). Color drawings picture stimuli.

Hymes, D. (1974). Foundations in Sociolinguistics: An ethnographic approach.
Philadelphia: University of Pennsylvania Press.

Jencks, C & Phillips, M. (1998). The Black-White Test Score Gap. Washington, DC:
Brookings Institute.

Johnson, T. & Fendrich, M. (1995). Modeling sources of self-report bias in a survey of
drug use epidemiology. Annals of Epidemiology. 15(5), 381-389.

Jones, J. & Preston, D. (in press). AAE and Identity: Constructing and deploying
linguistic resources. In D. Dwyer and B. Derman, eds. Afiican Language
Teaching and Learning, F estschrift for David Dwyer.

King, J. E. (1991) Dysconscious Racism: Ideology, Identity, and the Miseducation of
Teachers. The Journal of Negro Education. 60(2) 133-146.

Labov, W. (1966). The Social Stratification of English in New York City. Washington,
DC: Center for Applied Linguistics.

Labov, W. (1972a). Language in the Inner City. Philadelphia: University of
Pennsylvania Press.

Labov, W. (1972b). Sociolinguistic Patterns. Philadelphia: University of Pennsylvania
Press.

Labov, W. (1993). Peut-on combattre l'illettrisme? Aspects sociolinguistiques de
l'inegalite des chances a l'ecole [Is It Possible to Combat Illiteracy?
Sociolinguistic Aspects of the Inequality of Opportunity in School], Actes de la
Recherche en Sciences Sociales, 100, December, 37-50.

Labov, W. (1995). Can reading failure be reversed: A linguistic approach to the question.
In V. Gadsden and D. Wagner (Eds) Literacy among African-American Youth:
Issues in learning, teaching and schooling. (pp. 39-68). Cresskill, NJ: Hampton.

Lahey, M. (1988). Language development and language disorders. New York:
MacMillan.

Lareau, A. (2003). Unequal Childhoods: Class, race and family life. Berkeley, CA:
University of California Press.

165

 

Lawrence, S.M.& Tatum, 8D. (1997). Teachers in Transition: The Impact of Anti-

Racist Professional Development on Classroom Practice. The Teachers’ College
Record. 99(1), 162-178.

Lemke, J. (1990). Talking science: Language, learning and values. Norwood, NJ:
Ablex.

Malinowski, B. (1935). An Ethnographic Theory of the Magical Word. Coral Gardens
and Their Magic vol II. London: Allen and Urwin..

Massey, D. 8., Gross, A. B. & Eggers, M. L. (1991). Segregation, the concentration of
poverty and the life chances of individuals. Social Science Research 20(4), 397-
420.

Merriam-Webster Online Dictionary. Retrieved July 14, 2005. http://www.m-w.com/cgi-
bin/dietionary?book=Dictionary&va=idio1ect&x=8&y=17

Milroy, L. (1987). Language and Social Networks. (2nd Edition) Oxford: Basil
Blackwell.

Mitchell-Keman, C. & Keman, K. T. (1977). Pragmatics of directive choice among
children. Child Discourse. S. Ervin-Tripp and C. Mitchell-Keman. New York:
Academic Press Inc: 189-210.

Morgan, M. (2002). Language, Discourse and Power in African American Culture.
New York, Cambridge University Press.

Nettles, M. (2000). Statement on the NAEP 1999 Trends Report, The Nation’s Report
Card, National Assessment Governing Board, National Assessment of
Educational Progress (NAEP).

NIDCD. (National Institute of Deafness and Other Communication Disorders). Retrieved
July 10, 2007. http://www.nidcd.nih.gov/health/voice/ speechandlanguage.
asp#mychild

Ninio, A. & Snow, C. (1999). The development of pragmatics: Learning to use
language appropriately. in T. K. Bhatia & W. C. Ritchie (Eds), Handbook of
Language Acquisition (pp. 347-383). New York: Academic Press.

Ochs, E. (1988). Culture and Language Development: Language acquisition and

language socialization in a Samoan village. New York: Cambridge University
Press.

Ochs, E. & Schieffellin, B. B., Eds, (1979). Developmental Pragmatics. New York:
Academic Press.

166

Orfield, G. & Lee, C. (2005). Why Segregation Matters: Poverty and Educational
Inequality. Harvard University: The Civil Rights Project.

Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an
individual difference. Journal of Personality and Social Psychology, 77, 1296-
1312.

Pollock, S., Bailey, G., Berni, M., Fletcher, D., Hinton, L., Johnson, L, Roberts, .1. &
Weaver, R. (1998). Retrieved 9/11/2005. Phonological Features of African
American English, http://www.rehabmed.ualberta.ca/spa/phonology
/features.htm.

Poynton C. (1989). Language and Gender: Making the Difference. Oxford: Oxford
University Press.

Pressman, S. (2001). The Decline of the Middle Class: An International Perspective.
Luxumbeg Income Study Paper No. 280. Syracuse University: Maxwell School
of Citizenship and Public Affairs.

Prutting, C. A. (1982). Pragmatics as Social Competence. Journal of Speech and Hearing
Disorders 47: 123-134.

Reid, T. B. W. (1956). Linguistics, structuralism and philology. Archivum Linguisitcum,
8, 28-37.

Rickford, J. R. (1999). African American Vernacular English: Features, evolution,
educational implications. Language in Society, 26, Oxford, UK: Blackwell.

Rickford, J. R. & Rickford, R. J. (2000). Spoken Soul: The story of Black English. New
York: John Wiley and Sons Inc.

Saussure ’s Third Course of Lectures on General Linghuistics (1910-191 1) publ.
Pergamon Press, 1993.

Schieffelin, B. (1990). The give and take of everyday life: Language socialization of
Kaluli children. New York: Cambridge University Press.

Schieffelin, B., & Ochs, E. (Eds) (1986). Language Socialization Across Cultures.
New York: Cambridge University Press.

Searle, J. R. (1975). Indirect speech acts. In P. Cole & J.L. Morgan (Eds), Syntax and
Semantics. Vol. 3: Speech acts. New York: Academic Press.

Sedlacek, W. E. & Brooks, G. C. (1977). Racism in American Education: A Model for
Change. New York: John Wiley.

167

 

Seymour, H. Roeper, T. & de Villiers, J. (2003). Diagnostic Evaluation of Language
Variation-Screening Test, (DELV-Screening Test) San Antonio, TX: The
Psychological Corporation, A Harcourt Assessment Company.

Simeonsson, R. J. & Bailey, D. B. (1984). The Abilities Index. Chapel Hill, NC: Frank
Porter Graham Child Development Center.

Smitherman, G. (1975). Soul ‘11 Style. The English Journal, 64 (6), l2-l3.

Smitherman, G. (1977). T alkin ’ and Testifiin The Language of Black America. Detroit,
MI: Wayne State University Press.

Smitherman, G. (2000). Black Talk: Words and Phrases from the Hood to the Amen
Corner. New York: Houghton Mifflin.

Social Security Administration. Popular Baby Names (Popularity in 1995).
http://www.ssa.gov/cgi-bin/popularnames.cgi, entries 93 (Elijah) and 75
(Breanna). Retrieved Sept. 13, 2005.

Spencer, M. (2000). Afi'ican American Males’ Academic Achievement Experiences
Assumptions About Opportunities and Facts about Normative Developmental
Needs and Inopportune Structural Conditions. Unpublished paper presented at the
US Department of Education Symposium on African American Male
Achievement, Washington, DC, December 4, 2000.

Steele, C. M. (1992). Race and the schooling of Black Americans. Atlantic Monthly,
269(4), 67-78.

Stockman, I. J. (1996a). The Social-Political Construction of Science: Evidence from
language research on African American children. W.E.B. DuBois Distinguished
Visiting Lecturer Series: The Graduate School and University Center, The City
University of New York.

Stockman, I. J. (1996b). The Promises and Pitfalls of Language Sample Analysis as an
Assessment Tool for Linguistic Minority Children. Language, Speech and
Hearing Services in Schools. 27, 355-366.

Stockman, I. J. (1996c). Phonological development and disorders in African-American
Children. In A. Kamhi, K. Pollack & J. Harris (Eds). Communication
Development and Disorders in African-A merican Children: Research,

assessement, and intervention (pp. 117-153). Baltimore, MD: Brookes
Publishers.

Stockman, I. J. (1999). Semantic development of African American children. In 0.

168

 

Taylor & L. B. Leonard (Eds), Language acquisition across North America:
Cross-cultural and crosslinguistic perspectives (pp. 61—106). San Diego, CA:
Singular.

Stockman, I. J. (2007). African American English speech acquisition. In S. McLeod (Ed. ).
The International Guide to Speech Acquisition (pp. 148-160). Clifton Park, NY:
Thomson Delmar Learning.

Thurow, L. C. (1985). The Zero Sum Solution. New York: Simon & Schuster.

Troyna, B. & Hatcher, R. (1992). Racism in Children ’s Lives. New York: Routledge,
Chapman, and Hall.

Turner, L. D. (1949). A fricanisms in the Gullah Dialect. Chicago: University of Chicago
Press.

Urbaniak, G. (1997-2005). Retrieved 9/15/2005. Research Randomizer
http://www.randomizer.org/form.htm

Washington, J. & Craig, H. (1998). Socioeconomic Status and Gender Influences on
Children’s Dialectal Variations. Journal of Speech, Language, and Hearing
Research. 41(3), 618-626.

Wells, G. & Nichols, J. (1985). Language and Learning: An interactional perspective.
Contemporary Analysis in Education Series. New York: Routledge F almer

Wolfram, W. (1969). A Sociolinguistic Description of Detroit Negro Speech.
Washington, DC: Center for Applied Linguistics.

169

 

1|lflllllljfliljlllflljlflll111111I