1 MEASURING ATTITUDINA L CHANGE: A SOCIOLINGUISTIC AN D PSYCHOLINGUISTIC INV ESTIGATION INTO PERC EPTIONS OF AFRICAN AMERICAN ENGLISH AND ACADEMIC ENGLISH By CAROLINE KENNELLY LATTERMAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Caroline Kennelly Latterman
3 To Jeremy, with love, and to all of my students in Louisiana who sparked my interest in affecting education through linguistics
4 ACKNOWLEDGMENTS I have many people to thank for helping me along this journey. I would like to thank my committee: Dr. Diana Boxer, Dr. Wind Cowles, Dr. Helene Blondeau, and Dr. Dorene Ross for guiding me through the dissertation process. Diana and Wind especially gave of their time, helping me set up the experiment, reading and commenting on drafts of chapters, and offering advice all along the way. The departmental chair at the college where I collected data was a great help in offering me the ideal location to execut e my experiment, and I cannot thank enough the two professors who gave me not only class time but also encouragement and professional connections. I would also like to thank all of the participants who made this experiment possible, as well as the two sp eakers who provided the speech samples. This study would not have taken place had I not taught my wonderful students in Baton Rouge, Louisiana, and I would like to thank them for believing in me as a teacher and for sparking my interest in this topic. Al ong the way I developed carpal tunnel syndrome and tendonitis in my wrists, and without the hand therapy care of Anne Marie Muto and John Wyler I would not have been able to continue to type this work. I would also not have been able to complete this work without the financial support from my mom, Jan Truslow, who so lovingly and generously allowed me to have babysitters come while I wrote after the birth of my son, Parker Latterman. My mom supported and encouraged me all along the way, and I could always count on her to listen to new ideas and share in my excitement with the work. I am also very grateful to my friend Jennifer Faerber who helped me with the psycholinguistic statistics. She was great to work with and she made the process fun. Finally, I would like to thank my loving husband, Jeremy Latterman, most of all. Even before we were married he supported my academic
5 pursuits and believed in me and in my work. During the final writing stages he took our write in a quiet apartment, and he did all that he could, all along the way, to help me reach my goals. I cannot thank him enough.
6 TABLE OF CONTENTS ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 13 LIST OF ABBREVIATIONS ................................ ................................ ........................... 14 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 17 Backgr ................................ ................................ ............. 17 Purpose ................................ ................................ ................................ .................. 18 The Study ................................ ................................ ................................ ............... 20 Miscommunication Due to Differing Codes ................................ ............................. 23 Racial Overtones: I am White ................................ ................................ ................. 24 Overview of Chapters ................................ ................................ ............................. 25 2 LITERATURE REVIEW ................................ ................................ .......................... 27 Racial Achievement Levels ................................ ................................ ..................... 27 Te acher Attitudes and Student Achievement ................................ .......................... 28 Teaching Teachers About Language Diversity ................................ ................. 29 Educating Pre Service Teachers ................................ ................................ ...... 30 Measuring Attitudes ................................ ................................ ................................ 32 Explicit Attitudes ................................ ................................ ............................... 32 Implicit Attitudes ................................ ................................ ............................... 33 Implicit Association Test ................................ ................................ ................... 33 Whether to Administer an IAT ................................ ................................ .......... 35 Motivational and Attitudinal Change ................................ ................................ ....... 36 Matched Guise ................................ ................................ ................................ ........ 38 Academic Engli sh ................................ ................................ ................................ ... 40 African American English ................................ ................................ ........................ 41 Problematizing the Nomenclature ................................ ................................ .... 41 African American English as a Rule Governed Variety ................................ .... 42 Syntax ................................ ................................ ................................ .............. 43 Aspectual Be and Auxiliary Be ................................ ................................ ... 43 BIN (been) ................................ ................................ ................................ .. 44 Double/multiple ne gation ................................ ................................ ........... 44 Existential It ................................ ................................ ................................ 45 Question formation ................................ ................................ ..................... 45 Phonology ................................ ................................ ................................ ........ 46
7 Education Language Laws ................................ ................................ ...................... 48 Ann Arbor ................................ ................................ ................................ ... 48 The Oakland Controversy ................................ ................................ .......... 50 Strategies for Aiding in the Acquisition of Academic English ................................ .. 52 Negotiating Language and Identit y ................................ ................................ ... 54 Acting White ................................ ................................ ................................ ..... 55 Noticing Hypothesis ................................ ................................ .......................... 56 Explicitly Teaching Differences ................................ ................................ ......... 57 Contrastive Analysis ................................ ................................ ......................... 59 Valuing African American English ................................ ................................ ........... 62 Literacy and African American English ................................ ................................ ... 64 Dialect Readers ................................ ................................ ................................ 65 Dialect Differences Versus Reading Comprehension Errors ............................ 66 The Necessity of Acquiring Academic English ................................ ........................ 68 The Need for This Study ................................ ................................ ......................... 70 3 SOCIOLINGUISTIC METHODOLOGY ................................ ................................ ... 76 Selecting the Participant Population ................................ ................................ ....... 76 Securing a Research Site ................................ ................................ ....................... 76 Clas s Schedules and Data Collection ................................ ................................ ..... 77 ................................ .... 78 Participants ................................ ................................ ................................ ............. 78 Researcher Introduction ................................ ................................ .......................... 80 Assi ................................ .......................... 82 Pre Assessment ................................ ................................ ................................ ..... 83 Demographics Questionnaire ................................ ................................ ................. 84 Attitude Questionnaire ................................ ................................ ............................ 85 Sociolinguistic Lessons ................................ ................................ ........................... 86 Lesson and Experimentation Schedule ................................ ................................ ... 87 Lesson One ................................ ................................ ................................ ...... 87 Lesson Two ................................ ................................ ................................ ...... 91 Lesson Three ................................ ................................ ................................ ... 93 Lesson Four ................................ ................................ ................................ ..... 97 Bulletin board ................................ ................................ ............................. 98 Formal speaking event ................................ ................................ ............... 98 Jeopardy ................................ ................................ ................................ .... 99 ................................ ........................... 99 Lesson Five ................................ ................................ ................................ .... 100 Lesson Six ................................ ................................ ................................ ...... 102 4 SOCIOLINGUISTIC RESULTS ................................ ................................ ............. 104 Post Assessment Results ................................ ................................ ..................... 113 Favo rable/Accurate Answers by Race of Students ................................ ........ 117 Female Versus Male Participants ................................ ................................ ... 118 Post Assessment Responses by Race of the Participant ............................... 118
8 Switching Strategies to Use with Students ................................ ................................ ............................... 121 Attitude Questionnaire ................................ ................................ .......................... 124 Cont ent Category 1: The Structure and Inherent Usefulness of Nonstandard and Black English Dialects ................................ ................................ .......... 128 Content Category 2: Consequenc es of Using and Accepting Nonstandard and Black English in the Educational Setting ................................ .............. 131 Content Category 3: Philosophies Concerning the Use and Acceptance of Non Standard and Black English Dialects ................................ ................... 134 Growth by professor ................................ ................................ ................. 136 Growth by gender ................................ ................................ .................... 137 Growth by race ................................ ................................ ......................... 137 End of Experiment and Class Sessions Survey ................................ .................... 138 Pre Experiment Responses ................................ ................................ ............ 140 Po st Experiment Responses ................................ ................................ .......... 141 Sociolinguistic Conclusions ................................ ................................ ................... 142 5 PSYCHOLINGUISTIC METHODOLOGY ................................ ............................. 144 Experiment: Reaction Time Task ................................ ................................ .......... 144 Hawthorne Effect and Affiliative Social Tuning Hypothesis ................................ ... 145 Timing Responses ................................ ................................ ................................ 146 Method ................................ ................................ ................................ .................. 146 Participants ................................ ................................ ................................ ..... 146 Design ................................ ................................ ................................ ............ 147 Materials ................................ ................................ ................................ ......... 147 Stories ................................ ................................ ................................ ...... 147 Speakers ................................ ................................ ................................ .. 151 Adjectives ................................ ................................ ................................ 154 Procedure ................................ ................................ ................................ ....... 155 Logistics ................................ ................................ ................................ ......... 155 Apparatus ................................ ................................ ................................ ....... 157 The Experiment ................................ ................................ .............................. 157 6 PSYCHOLINGUISITC RESULTS ................................ ................................ ......... 160 Pre Test Results ................................ ................................ ................................ ... 161 Regression 1: Overall Results for the Pre Test ................................ .............. 161 Regression 2a: Pre Test Positive Adjective Chosen ................................ ...... 162 Regression 2b: Pre Test Negative Adjective Chosen ................................ ..... 163 Regression 3: Story Two (AA AE) Versus Story Four (W AE) ........................ 164 Regression 4a: Stories Three and Four Versus S tories One and Two: Number of Positive Adjectives ................................ ................................ ..... 165 Regression 4b: Stories Three and Four Versus Stories One and Two: Speed of Positive Adjectives ................................ ................................ ....... 165 Regression 5: Speed of Response When Choosing a Negative Adjective with All Stories Compared to Story Four (W AE) ................................ ......... 166
9 Regression 6: Speed of Choosing a Negative Adjective with All Stories Compared to Story One (AA AAE) ................................ .............................. 167 Overview of Pre Test Results ................................ ................................ .. 167 Post Test Results ................................ ................................ ................................ 168 Regression 7: Overall Percentages of Positive Results ................................ 168 Regression 8: Overall Story Order Results for the Post Test ......................... 169 Regression 10a: Overall speed of adjectives chosen for Story One ........ 172 Regression 10b: Story One: C omparison of speed of positive adjectives chosen between Pre and Post Test ................................ ..................... 172 Regression 10c: Story One: Comparison of speed of negative adjectives chosen between Pre and Post Test ................................ .... 173 Regression 11: Story Two (AA AE) Versus Story Four (W AE): Positive Adje ctive Selection ................................ ................................ ...................... 173 Regression 12a: Stories Three and Four Versus One and Two: Number of Positive Responses ................................ ................................ ..................... 174 Regression 12b: Stories Three (W AAE) and Four (W AE) Versus One (AA AAE) and Two (AA AE): The Relationship Between Positive and Negative Responses ................................ ................................ ................................ .. 174 Regression 13a: Stories Two (AA AE), Three (W AAE), and Four (W AE) Versus Story One (AA AAE): Positive Responses ................................ ...... 175 Regression 13b: Stories Two (AA AE), Three (W AAE), and Four (W AE) Versus Story One (AA AAE): Negative Responses ................................ .... 176 Regression 14a: Stories One (AA AAE) and Two (AA AE): Speed of Positive Responses in the Post Test Versus Pre Test ......................... 177 Regression 14b: Stories One (AA AAE) and Two (AA AE): Speed of Negative Responses in the Post Test Versus Pre Test ........................ 177 Regression 15a: Stories Three and Four: Speed of positive responses in the Post Test versus Pre Test ................................ .......................... 178 Regression 15b: Stories Three and Four: Speed of negative responses in the Post Test versus Pre Test ................................ .......................... 178 Overview of the Post Test responses ................................ ...................... 179 Percentages of positive and negative adjectives in the Pre and Post Test ................................ ................................ ................................ ....... 179 Discussion ................................ ................................ ................................ ............ 180 Demographics ................................ ................................ ................................ ....... 187 Pre Test Results ................................ ................................ ............................. 188 Post Test Results ................................ ................................ ........................... 188 Professor ................................ ................................ ................................ .. 188 Gender ................................ ................................ ................................ ..... 189 Race ................................ ................................ ................................ ........ 189 Age groups ................................ ................................ ............................... 190 East Coast versu s West Coast ................................ ................................ 190 New York versus elsewhere ................................ ................................ ..... 190 Teacher versus non teacher ................................ ................................ .... 191 Length of teaching ................................ ................................ ................... 191 Race of students ................................ ................................ ...................... 191
10 Demographi cs: Pre and Post Test Comparisons ................................ ................. 192 Demographics Discussion ................................ ................................ .................... 194 Psycholinguistic Conclusions ................................ ................................ ................ 198 7 CONCLUSIONS AND IMPLICATIONS ................................ ................................ 202 Conclusions ................................ ................................ ................................ .......... 202 Summary of Sociolinguistic Results ................................ ............................... 202 Summary of Psycholinguistic Results ................................ ............................. 206 Exploring the Sociolinguistic and Psycholinguistic Results Together ............. 207 Implicit Versus Explicit Measures ................................ ................................ ... 208 Is New York Ready for African American English? ................................ ............... 209 ................................ ................................ 211 Implications ................................ ................................ ................................ ........... 212 APPENDIX A PRE ASSESSMENT ................................ ................................ ............................. 215 B DEMOGRAPHICS QUESTIONNAIRE ................................ ................................ .. 217 C ATTITUDE QUESTIONNAIRE ................................ ................................ .............. 219 D END OF TASK RESPONSE (AFTER PRE PSYCHOLINGUISTIC EXPERIMENT) ................................ ................................ ................................ ..... 225 E END OF TASK RESPONSE (AFTER POST PSYCHOLINGUISTIC EXPERIMENT) ................................ ................................ ................................ ..... 226 F POST ASSESSMENT ................................ ................................ .......................... 227 G READING EXAMPLE ................................ ................................ ........... 229 LIST OF REFERENCES ................................ ................................ ............................. 230 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 237
11 LIST OF TABLES Table page 1 1 Story, speaker, and variety ................................ ................................ ................. 22 3 1 ................................ ................................ 79 4 1 Percentages of Pre Assessment favorable/accurate responses ...................... 107 4 2 Favorable/accurate responses by gender of participant ................................ ... 112 4 3 Total favorable/accurate responses ................................ ................................ .. 113 4 4 ................................ ......... 114 4 5 Pre and Post Assessment favorable/accurate responses ............................... 114 4 6 Favorable/accurate responses by teacher status ................................ ............. 116 4 7 Favorable/accurate responses by race of students ................................ .......... 117 4 8 Favorable/accurate responses by gender of participant ................................ ... 118 4 9 Means of Pre and Post Attitude Questi onnaire responses .............................. 126 5 1 Story, speaker, and variety ................................ ................................ ............... 149 5 2 Positive, negative, and neutral adjectives ................................ ......................... 154 6 1 Story, speaker, and variety ................................ ................................ ............... 161 6 2 Overall Pre Test speed of results ................................ ................................ ..... 161 6 3 Pre Test positive adjective chosen ................................ ................................ ... 162 6 4 Pre Test story order ................................ ................................ .......................... 163 6 5 Speed of Pre Test negative adjective chosen ................................ .................. 163 6 6 Overview of Pre Test responses ................................ ................................ ...... 168 6 7 Percentage of positive adjective selected for Pre and Post Tests ................... 169 6 8 Differences between the mean numbers of positive adjectives ........................ 169 6 9 Average Post Tes t response times ................................ ................................ ... 170 6 10 Post Test positive adjective response times ................................ ..................... 171
12 6 11 Post Test negative adjective response times ................................ ................... 172 6 12 Overview of the Post Test responses ................................ ............................... 179 6 13 Percentages of positive and negative adjectives in the Pr e and Post Test ..... 179 6 14 ................................ ................................ ............ 180 6 15 Pre and Post Test comparisons ................................ ................................ ...... 193
13 LIST OF FIGU RES Figure page 4 1 Pre Assessment favorable/correct responses by race ................................ ..... 112 4 2 Post Assessment favorable/correct responses by race ................................ .... 118 4 3 Pre Assessment favorable/correct responses by race ................................ ..... 119
14 LIST OF ABBREVIATION S AAVE African American Vernacular English: What I am referring to as African American English BEV Black English Vernacular IAT Implicit Association Test L 1 First language acquired L 2 Second language acquired SE Standard English: What I am referring to as Academic English
15 Abstract of Disse rtation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MEASURING ATTITUDINA L CHANGE: A SOCIOLINGUISTIC AN D PSYCHOLINGUISTIC INV ESTIGATION INTO PERC EPTIONS OF AFRICAN AMERICAN ENGLISH AND ACADEMIC ENGLISH By Caroline Kennelly Latterman May 2013 Chair: Diana Boxer Cochair: H. Wind Cowles Major: Linguistics and Academic English. Participants were graduate students of Education at a college in New York City. They completed a paper and pencil questionnaire that assessed their explic it attitudes towards the two varieties, as well as a Psycholinguistic Experiment that was designed to capture their implicit attitudes. They completed these measures at the beginning of the semester and again at the end. In between the Pre and Post Tes ts they participated in six lessons that gave an overview of the syntax, phonology, and social issues surrounding African American English. After the six lessons the participants completed the same attitudinal measures as before, and the change in attitud thinking that African American English was a slang, lazy manner of speaking, to believing that it is a rule governed variety that has social value. The Psycholinguistic Experi ment was designed within the matched guise methodology, such that participants heard four speech samples ostensibly from four different speakers, yet there were in fact only two speakers who each gave two separate samples. One speaker was an
16 African Ameri can woman and the other was a White woman. Each speaker told a story in African American English and another story in Academic English, for a total of four different stories. Participants heard one story and then completed an adjective discrimination tas k in which they assigned eleven sets of opposing adjectives to the speaker they just heard, using a response box that timed their responses down to the millisecond. They completed this activity for all four speech samples. After the Pre Psycholinguistic Experiment the participants strongly preferred the speech of the White speaker, regardless of what variety she was using. By the Post Psycholinguistic Experiment, however, the only significant interaction was that participants strongly preferred the White speaker using African American English over the African American speaker using that same variety. The results are considered positive in keeping with attitudes towards Afri can American English.
17 CHAPTER 1 INTRODUCTION This research grew out of several negative phrases I heard teachers saying to students a decade ago while I was teaching in Baton Rouge, Louisiana, with Teach For America. school I was teaching the majority of the students were African American, and most, if wor d since the students and teachers were using it in a conversational context and it had meaning attached to it, and it is also listed in major dictionaries. When I heard about this paradoxical use. Their teachers were using it as a word and everyone understood it to be a negation and question marker with a variety of meanings, so how was it not a word? I did not know if the students understood the larger framework within which the teachers were operating, i.e. that mastery and usage of Academic English hat their teachers were saying as a personal criticism and that it might cause them to be less eager to speak in class. different from Academic English, yet I did not know tha t it was a fully fledged variety with its own set of syntactic and phonological rules that were codified in academic texts. I often heard teachers saying very negative things to the students about how they spoke, and while I understood that the teachers we re trying to help the students, I knew
18 that this could not be the most appropriate or most successful strategy. Language and identity are inextricably linked, and in my opinion, as the teachers degraded the ding the students themselves. I understood that the teachers knew that the students would be more prepared academically and career wise if they acquired Academic English, as this is a significant tool towards achieving academic, social, and economic mobil ity and equity. However, I could not shake the feeling that what they were saying to the students could have very negative repercussions. At the time I had no idea that years later I would be conducting research that grew from my discomfort with the way t he teachers were discussing language with the students. This discomfort put me on the path towards seeking information regarding how the students were speaking as well as appropriate methods to assist them in mastering Academic English. While I still reg ret not having this information to share with my students and the other teachers then, I feel fortunate to be able to share it with teachers now as I did in this current study. Purpose The purpose of this research was to ascertain whether my teaching me thods, measured not only on a paper and pencil questionnaire, but also throug h a Psycholinguistic Experiment that made it much harder to conceal potentially socially undesirable attitudes. As will be discussed in Chapter 2, research indicates that most teachers have negative attitudes towards African American English and that te aching about the
19 linguistics of African American English ought to improve these attitudes. Additional language varieties, the students perform better academically. I wou ld like to see all of these strands of research connected, which entails teaching teachers about the linguistics of African American English, changing their attitudes towards African American English and thus ultimately their interactions with their studen ts, and then seeing how this teaching has affected the students themselves and their academic progress. My larger vision is that eventually all teachers will be trained in the linguistics and social issues surrounding African American English, and that thi s training will improve Controlling only African American English and not having the ability to produce Academic English when desired greatly limits social and econ omic mobility, and I want are due to their limited linguistic repertoire. There is currently much discussion in politics and education about the disparities between m on why minority students do not perform as well as their White counterparts. I believe, however, that language is an issue that is grossly overlooked, and that closing the achieveme nt gap as well as the opportunity gap is not possible until the variety of language that students are using is taken into account and treated as an element that must be examined in order to help improve their performance. This idea of an was coined by da Silva et al. (2007), and they are referring to the idea
20 that not only is there a gap in terms of academic achievement that can be seen through test scores, but there is also a gap in terms of what opportunities are available to different must recognize that the gaps in educational achievement that we are so fond of Much research investigate s social issues that affect student performance, and while I find this research to be extremely valuable and necessary, there is little research this is a significant hole in the fight to close the achievement gap, and that the solution starts in part with teaching teachers about the systematic, rule governed nature of African American English, and with providing teachers strategies to help their students learn to code switch between their home variety and Academic English. It is imperative or how they sound. This is where the current research fits in. It examined if a six lesson te aching teachers have positive attitudes towards African American English is, from my perspective, the first step in the process of teaching students how to code switch. This ought to play a strong role in ultimately leveling the academic playing field. The Study This research sought to determine whether teaching participants about the linguistic and social features of African American English would make their attitudes to wards the variety more positive. The participants were graduate students of Education at a university in New York City, and most of them were currently teaching.
21 The research consisted of gathering pre test data for participants regarding their viewpoint s towards African American English and Academic English through sociolinguistic and psycholinguistic measures, then teaching about the linguistics of African American English for six hour long sessions over a period of about six weeks, and then administeri ng post tests that were identical to the pre tests in order to measure The sociolinguistic elements included a Background Questionnaire that ge of African American English as a linguistic variety, and it ascertained whether they knew of any methods to assist mono dialectal speakers of African American English acquire Academic English. Participants also completed a Demographics Questionnaire an d an Attitude Questionnaire that sought to determine their attitudes towards African American English as well as its place in education and in society. timed responses to eleve n sets of adjectives that they selected after hearing stories told by an African American woman and a White woman. Each of the women narrated one story in African American English and one story in Academic English. Participants heard a story, selected an adjective by pressing the associated button on a response box for each set of adjectives, and then went on to the next story until they had completed the task for all four stories. The adjectives were opposite in nature, such as intelligent/unintelligent, loud/quiet, upper class/lower class, etc. There were four different stories with two stories from each speaker in a matrix as follows below in Table 1 1.
22 Table 1 1. Story, speaker, and variety Story Speaker Variety One African American African American English Two African American Academic English Three White African American English Four White Academic English they selected in reaction to the four different speech samples. There is agreement among researchers that when respondents take a longer time to respond to a cue in this type of experiment it indicates more difficulty in making that decision. Likewise, when participants responded more quickly it indicated th at they had less difficulty with making that decision, or in this case, adjective selection. Therefore, this experiment not only much difficulty or ease they had in making that decision. The sociolinguistic methodology included paper and pencil questionnaires, however participants may shift their responses towards what they believe is more socially acceptable or what the researcher is looking for in their responses when completing a paper and pencil measure. Because of this, the psycholinguistic element of the experiment was that they may not have realized that they have. Between the pre and post testing I taught six lessons about the syntax, phonology, history, and other social issues surrounding African American English. I also taught participants methods for assisting their students who speak African American English or anoth er non standard variety to acquire Academic English without
23 Chapter 3. Miscommunication Due to Differing Codes Miscommunication is often a result of speakers who command different linguistic varieties, yet when these varieties are very similar to each other, such as with Academic English and African American English, interlocutors are often unaware that these insta nces of miscommunication are actually happening. Most Americans are unaware of the existence of African American English and therefore do not consider that there would be language differences which could create confusion or a miscommunication between two speaking Americans. For example, if Person A only speaks English and she is speaking with Person B, who speaks French as an L1 and has limited ability in English, Person A would likely be mentally prepared for potential communication issu es. On the other hand, if Person A is part of a conversation with Person C, who is a mono dialectal African American English speaker, and Person A is unaware that Academic English and African American English are two distinct varieties, she might not be p repared for communication gaps between Person C and herself because she is unaware that she and Person C are operating within different codes. Because she is unaware of the differing codes, she is unaware that a miscommunication may occur since she assume s that she and Person C are operating within the same linguistic system. In a similar vein, Shirley Brice Heath (1983, reprinted in 2009) provides an -remote past BIN -that led to a miscom munication and misunderstanding between a teacher and student. The teacher is a mono dialectal speaker of Academic English. Brice Heath recounts:
24 the teacher indicat had not understood her follow up question, and she was not certain what Teegie had meant. I explained that his use of been in the first answer indicated that the taking and keeping of his dog at his g happened in the distant past, and Teegie reaffirmed that point in his follow up answer. (p. 277) This is a clear example of a teacher who was not familiar with the syntactic patterns of African American English and who misunderstood the It is clear that it would be beneficial for teachers to understand the language patterns of their students not only for the sake of comprehension, yet also in terms of enabling their students to acquire code switching ski BIN she would have been able to use that moment to show Teegie and the other students the differences and similarities between that construction and the Academic English equivalent. Racial Overtones: I a m White I have considered and have often been asked about how I think my race -I am White -affects my teaching about African American English and how participants respond to me. This is a question about which I am unsure of my response, and I reflected a bout race consistently throughout all phases of the research. My perceived negatives of being White in relation to my goals are that participants might assume that trying speaking students speak. Also, I wonder if participants and people in general will listen long enough to find out that I am very aware of the myriad social and racial undertones underlying the issues of
25 language and how language, identity, and power are intertwined. Language (and especially African American English) is a sensitive topic, and I never know if when beginning a conversation a listener is going to assume that I am working within a racist fra mework once they realize that I am talking about the intersections of race and language. My perceived positives of being White are that I am not trying to promote a language variety that I speak natively. I think that this might lend some credibility to the notion that what I am teaching is valid and that I am not trying to promote the use of a non standard variety that I speak natively. I believe that teachers should be aware that African American English is a rule governed, logic based variety, but whe n I am presenting these beliefs the color of my skin does not suggest that I am attempting to promote African American English because it is my personal linguistic variety. Overview of Chapters Following this introduction is Chapter 2, where I provide an overview of the literature that underpins this study. Research is culled from several domains including sociolinguistics, psycholinguistics, education, and social psychology. Because this research intertwines several different domains, I attempted to pr esent relevant research from all fields in a manner that informs the reader of the major tenets within each field that support the current research. Chapter 2 reviews the methodology I used for the sociolinguistic aspects of the research. Following that is Chapter 4, which details the results from those measures. Chapter 5 lays out the methodology for the Psycholinguistic Experiment, and Chapter 6 then presents the results from that experiment. I conclude with Chapter 7 in which I tie
26 together the soci olinguistic and psycholinguistic results as well as provide implications for this study as well as for research that could be born from this study in the future.
27 CHAPTER 2 LITERATURE REVIEW Racial Achievement Levels African American students have consistently fared worse academically than their White or Academic English speaking counterparts (Wolfram, 1970; Craig et al., 2004; Lee, 2002; Taylor, 1991; Singleton & Linton 2006; Ladson Billings 2000, among numerous other s), and children who primarily or only speak African American English are far behind White students in reading and most ot her areas of academia (Rickford, 2005). Rickford notes that performing poorly in English class generally indicates poor performance i n all other subject areas (1999), and he has shown that for every year African American students are in school they progressively perform worse (2005). This decline in academic performance continues as will be discussed shortly. It is abundantly clear t hat there is a need for dramatic educational change. According to the most recent data from the National Center for Education Statistics (2010), there remains a gap between White, African American, and Hispanic Students. The report categorizes students into three achievement levels: Basic, Proficient, and Advanced These achievement levels are performance standards showing what students should know and be able to do. Basic denotes partial mastery of knowledge and skills that are fundamental for profic ient work at a given grade. ( Below Basic therefore, denotes less than this level of achievement.) Proficient represents solid academic performance. Students reaching this level have demonstrated competency over challenging subject matter. Advanced sign ifies superior performance. In terms of academic proficiency for eighth graders in 2011, while 30% of White students were categorized as Proficient, only 10% of Black students and 13% of Hispanic students were Proficient. Additionally, while only 13% of W hite students were categorized as Below Basic, 35% of Black students and 31% of Hispanic students were
28 Below Basic. The results for the twelfth graders are almost identical in terms of those labeled Proficient, yet there are even more Black and Hispanic s tudents who are categorized as Below Basic: 39% of Black students and 35% of Hispanic students, compared to 13% again for White students. White students, on average, outper form poor Black students, and this pattern persists at the middle and upper income levels as well. Even more alarming are data that indicate poor White students may outperform middle These authors believe race and racism to be at the heart of the achievement gap, and I am in agreement with them about these major causes for the academic disparities between the races. Certainly language is deeply intertwined with the issues of race and racism, though I believe it pl ays a stronger role than is currently being discussed. It is because of this belief that in this current study I am investigating attitudes towards different language varieties with the hope that in the future, the results of this study will be able to be tter inform teacher development and ultimately increase student achievement, and many teachers have negative attitudes towards African American English. It is imperative that w e ensure that students are being judged based on their true academic performance and not on the language variety they possess or the way they sound. Teacher Attitudes and Student Achievement language, the student will perform more poorly (Rickford, 2005). It is thus imperative
29 that attitudes can be mitigated. significant role in the experiences of students, and furthermore, predictions of stud other studies link student achieve ment with teacher attitudes (Ladson Billings, 2004; Ladson Billings 2000 ; among others), and many indicate that teachers in gene ral have negative attitudes towards African American st Billings, 2000; Bowie & Bond, 1994; Blake & Cutler, 2003). For example, Fogel & Ehri (2000) note specifically that African American English purpose of improving student achievement. Although measuring student achievement is not part of the scope of this study, the guiding purpose is to investigate whether teaching teachers about the rule governed nature of African American English and the fact that it is a recognized linguistic variety will affect their attitudes, given that the research so clearly shows attitudes. Teaching Teachers About Language Diversity Blake & Culter (2003) insist that teachers be taught about language diversity both for students who speak English as an L2 as well as for s peake rs of African American English. They state that even in such a diverse setting as New York City, teachers are not linguistically informed. According to them, it is imperative to have educational laws which, in this case, are targeted at categorically tr aining teachers about language diversity. The individuals creating such training programs should be cognizant of the needs of non
30 native English speakers, while also addressing the needs of students who are nonstandard English speakers, this including Afr ican American English speaking students. In diverse metropolitan centers like New York City, one cannot address bilingualism without addressing bidialectalism, and vice versa, and yet this continues to be the case. (pp. 189 190) Likewise, Baugh (1995) st understanding for language minority students among educators can have negative consequences for African American st udents Even when linguistic diversity training is provided, Adger (1995, as cited in Wiley & Lukes, 1996) says that it does not meet the demands of its population (p. 528). Baugh (1995) also reports that some urban educators have remarked that language development is a low academic priority for poor minority students because of the myriad of other social problems they face in daily life. The paradox at hand lies in the fact that a good education is critical to breaking perpetual cycles of pove rty, and linguistic skills are a strong indicator of the likelihood of eventual academic success (p. 88, italics mine) Educating Pre Service Teachers achievement, one might expect p re service teachers to be educated about linguistic diversity. Wynne (2008) gave a survey to pre service teachers in attempts to assess their opinions about Academic English and African American English. She was genuinely surprised by their negative resp onses to African American English and general lack of awareness of the social prestige that Academic English holds, and she we in schools of education give so little time, effort, and attention to teaching our pre 211).
31 To investigate how pre service teachers are being educated about linguistic issues, Bowie & Bond (1994) conducted a study with seventy five pre service teachers. Their purpose was to assess teacher attitudes towards African American English. Unsurprisingly, their results indicated negative attitudes, and they noted that the most negative responses in relation to Afric an Americans and African American English stemmed from pre service teachers who reported little to no exposure to African is a paper and pencil measure of explicit attitude s, and I used this same measure in the participants rated African American English as inferior to Academic English. English is a legitimate and rule Notably, only 19% of the pre service teachers said that they had had in depth discussions regarding African American English. This clearly dem onstrates the need to expose pre service and other teachers to African American English as a systematic variety. Bowie & Bond (1994) specifically state that teacher education students need to be more culturally aware and cognizant of differences, including linguistic difference s Charity et al. (2004), discussing both teachers and pre service teachers, assert abilities and social statues than have speakers of AAVE or other (p. 1352). Because of this, it is imperative that we teach pre service and currently teaching teachers about linguistic diversity and encourage them to see that the variety a
32 student speaks does not necessarily correspond with how in telligent that student is or how well that student will fare academically. The research that has been presented thus far indicates that teachers, along with other community members, have negative attitudes towards non standard varieties, including African American English Beca use of these negative attitudes that filter down into student assessments, I believe students must become bi dialectal and able to switch between codes. Ideally, teachers would be educated in linguistic differences and made aware of differing varieties and rules gov erning them in order to assist students in becoming bi dialectal, though in the mean time much of the burden rests on the students themselves. Measuring Attitudes Researchers in a variety of fields, including psycholinguistics, psychology, and import attitude is broken down into either an implicit or explicit attitude. Explicit Attitudes Explicit attitudes are those that are defined as attitudes for which people have conscious control a nd are those that people are able to express to an audience. Measures of explicit attitudes include scales such as a multiple item rating scale, semantic differential, feeling thermometer, likability rating, evaluative rating scale, and the modern racism scale, as identified throughout the literature but summed up in Gawronski & Bodenhausen (2006), as well as the Internal and External Motivation to Respond without Prejudice (Plant & Devine, 1998). While there is debate regarding
33 how conscious these explic it rating scales are, there is general agreement within the research community that these scales measure attitudes that participants are able to express to others. There are differences in when participants will express their attitudes, in that a White pa rticipant is more likely to respond in a less biased way when the experimenter is Black than when she is White (Lowery et al., 2001), yet simply the fact that she can control the expression of her attitude indicates that it is explicit. Explicit attitudes have been assumed to be less truthful or insightful as implicit attitudes, their attitudes in order to conform to expected social norms (Plant et al., 2003, p. 187). Im plicit Attitudes Researchers disagree about the amount of awareness a participant may have regarding their implicit attitudes, yet in general it is assumed that implicit attitudes are those over which the participant does not have conscious control, thus t hey have been referred to as unconscious attitudes available evidence indicates that self reported and indirectly assessed attitudes are to depend on a variety of different variables, such as motivational factors, the degree of deliberation during self report, conceptual correspondence between measures, and measurement account different ways in which participants may be primed for certain attitudes. Implicit Association Test The Implicit Association Test (IAT) was designed by Greenwald, McGhee, and Schwartz in 1998, and it is referenced in nearly every article that di scusses measures of implicit attitudes. The IAT can be used to judge implicit attitudes towards a variety of
34 groups, such as Black and White, or Korean and Japanese participants, as originally created in Greenwald et al. (1998). They specifically state t automatic associations even for subjects who prefer not to ex 1465). For explanatory purposes, I will focus on the measure in terms of Black/White attitudes. The measure works as follows: There are five trials within the test, and each of the trials associates certain emotions and names w ith which hand the participant uses to press the key. The first differentiate between Black and White American names. In the second trial, called the participants to combine their decisions between a Black or White name and a Pleasant or Unpleasant attrib forces participants to change keys, such that Black and White names are on opposite Trial three, wh ere on one key participants associate Black names with Unpleasant attributes or White names with Pleasant attributes. Response times are scrutinized on the third and fifth trial, and the IAT effect defined as the difference in mean latency between the se two conditions (non compatible trial is Trial Three in which participants are associating Black names with Pleasant attributes, and the
35 compatible trial is Trial Five in which participants are associatin g White names with Pleasant attributes. The results show that participants are able to categorize what they perceive to be compatible associations faster than non compatible associations. Thus, Trial Five is generally faster than Trial Three, revealing towards the two racial groups, with faster response times when White names are paired with Pleasant adjectives and slower response times when Black names are paired with Pleasant adjectives. Whether to Admi nister an IAT I spent considerable time debating whether it would be appro priate to administer an IAT, and I chose not to because of what it measures and the time it would take to administer it. For this specific research I was interested in attitudes tow ards language varieties and how these attitudes change over time due to explicit instruction regarding the linguistic features of the variety along with social and cultural factors. I thus measured not only attitudinal change that resulted solely from li nguistic education but from activities in which the participants engage d that were specifically targeted at changing their attitudes towards African American English The IAT, which measures in Black/White names), does not measure language attitudes. Although the IAT is not used in the present experiment, the results indicating a latency towards Black names and Pleasant words will be drawn upon in the experimental design, namely in the question regarding whether participants will respond more slowly to positive adjectives paired with African American English. As will be described in Chapter 5, pressing the buttons slower during the Psycholinguistic
36 Experiment indicates more conflict for agreemen Additionally, in this study a comparison between the implicit measure and explicit measure will be made to investigate whether participants respond in more socially desirable ways on an explicit measure. Motivation al a nd Attitudinal Change Motivational and attitudinal change have been explored in a variety of ways. Plant & Devine (1998) investigate the differences between internal and external ond without prejudice as resulting from internalized and personally important non prejudiced standards. In contrast, external motivation to respond without prejudice is conceived of as resulting from social pressure to comply with non 812). Through several self report measures they classify participants as either high or low in both external and internal motivation to respond without prejudice and consider how people who are motivated by different factors respond to different situati ons. address this attitude. However, some of the literature regarding expressed attitudes indicates that when people are directly confronted with their attitudes the y may go against what they perceive to be what society expects of them and actually further very real concern is that people who are primarily externally motivated may g row to resent the infringement on their freedom and feel increased frustration and reactance, Instead, I
37 attitudes, especially during the first few lessons. Plant & Devine (2001) a degree that people view pro Black pressure as a constraint on their freedom of response and anticipate that noncompliance will result in punishment as well as future restrictions on behavioral freedom, they are particularly likely to respond with reactance teaching lessons I did not directly confront race. I did not believe that there was enough time to address it in a way and I did not want participants to react negatively to the information that would be presented in the future lessons. In conjunction, Monteith et al. (1998) produce findings indicating that if pa nable with responses timed to the millisecond, indicating that the responses are not under conscious control. For example, Blair et al. (2001) find that participants were less likely to judge a woman with stereotypical traits when primed by a counter ster eotypical individuals in counter effective in reducing unwanted stereotyping than attempts to negate an exi sting were incorporated into the current experimental design. For example, while I did tell
38 of Academic governed variety that is just as grammatical as Academic English, simply with different grammatical and sound Matched Guise The matched guise methodology is a measure that was created within the discipline of social psychology and is used within the realm of sociolinguistics and psychology, though not generally used within the discipline of psycholinguistics. As this current research is combining sociolinguis tic and psycholinguistic methodology, data will be examined that was created in part through this methodology. It is difficult to determine whether the matched guise technique should be categorized as an explicit or implicit attitudinal measure. In a matc hed guise experiment one speaker is recorded speaking in two different registers or languages. For example, in the seminal work by Lambert (1967), Canadian participants listened to a speech sample supposedly from a French Canadian and another from an Engl ish Canadian. Participants rated the English Canadian speaker more favorably, indicating preference for this manner of speaking, and according to Lambert, this group of people. In reality, the participants only heard one speaker who used a French Canadia n guise in one speech sample and an English Canadian guise in the other. Participants thus made linguistic judgments and displayed their attitudes towards the different language tic identity. (p. 94). Researchers to this day use this technique, as it is believe d to tap into
39 voluntarily produce upon questioning. I used the matched guise technique with the four speech samples that participants heard in attempts of uncover ing their linguistic and racial attitudes that they might not express on a paper and pencil measure or on a response where they might not want to display unfavorable social attitudes. Researchers do not seem to be questioning whether a matched guise exper iment on its own should be considered to gather implicit or explicit attitudes. Rather, they conceive of it as a tool to gather attitudes towards groups of people or dialects (Lambert, 1967; Purnell, Idsardi, & Baugh, 1999; Casesnoves Ferrer & Sankoff, 200 4) as opposed to the actual speaker specific effects that would have & Stuart Smith, 2012, p. 769). The matche d guise methodology is still widely in use and I believe that it, coupled with the IAT type response data, serves as a solid methodology with which to gather data for the current research. In the current experiment I used the matched guise technique in con junction with different types of speech and assessed their attitudes based on how long it took them to press the buttons for each particular trial. I therefore belie ve that I was accessing the explicit paper and pencil measures that are discussed in relation to the sociolinguistic element of the experiment. Their explicit att itudes were those that they displayed on the paper and pencil measures as well as the particular adjective that they
40 selected, and their implicit attitudes are revealed in the speed at which they made their adjective selections. More recent research clea rly indicates that explicit attitude questionnaires are mediated by external factors (Shelton et al., 2005; Lowery et al., 2001; Phelps et al., 2000; among numerous others), and though using a matched guise technique might not be as effective at capturing implicit racial attitudes as an IAT, it would access more unconscious attitudes than a measure that asks participants to describe or rate a speaker while knowing more about the social and linguistic background of that speaker. Although an IAT is not appro priate for the current research, the matched guise technique is, and the method by which I recruited speakers and collected the speech samples will be detailed in Chapter 5. The matched guise in the current experiment involves two speakers, each of whom speaks both Academic English and African American English. Following is a more in depth exploration of what Academic English and African American English are considered to be for this purpose. Academic English to convey the idea that African American this is not a term or concept with which the general public is familiar, and since in my research I interacted with participan ts who were not familiar with this terminology, I chose to use the term Academic English, which does not have an unsavory opposite.
41 Of note, however, is that while the term Academic English has the term st formal of language that is often used in academia. I am instead referring to what many think of as Standard, or perhaps mastery of would benefit students in terms of c areer choice and social and economic mobility. African American English Problematizing the Nomenclature The term for the non standard variety that I am referring to as African American English does not have a perfect label. I will be using the term Africa n American English and I will discuss my reasons for this choice, however it is important to note that there are many reasons why this term is not ideal. Over the decades there have been many terms for this language variety, including Negro English, Blac k English, Black Vernacular English, African American Vernacular English, Ebonics, and what is currently in use most often -African American English. An important reason for the term being not ideal is that all African Americans do not speak African Ameri can English, and one does not need to be African American in order to speak it. Children grow up speaking the language variety that is around them, so in the United States a Dominican child who lives around a group of people who speak African American Eng lish, regardless of what race she is, will naturally speak that variety. I never use the term Ebonics because it is not technically the correct term for the variety of English spoken by African Americans in the United States as it refers to languages in the African Diaspora, which includes Caribbean languages as well. I also
42 avoid using that terminology because, as will be discussed, it came into popular terminology in the United States during the Oakland Controversy in 1996. When people hear it they im mediately have a negative impression towards the variety due to a misunderstanding of the program the Oakland school board was attempting to implement in the school system, and I do not want to activate this negativity in their minds. I also do not use th e term African American Vernacular English (AAVE) because I do not like the idea of it being relegated to a vernacular. It is used as a vernacular for many speakers, yet for many it is their sole language of use. I have considered calling it a variety t hat does not include the term African American since there are many speakers in New York who speak it who are not African American, however I think there is value in sourcing the variety to whom it originates from in the name. Because millions of African Americans speak African American English, and because the largest majority of people who speak it are indeed African American, I believe it takes something away from the speakers to take their heritage out of the name. In my opinion there is no perfect na me for the variety. I have chosen to use the term African American English since this is the terminology used most often and currently in the literature, and I believe it to be the most accurate. I have no doubt, however, than within the next decade ther e will likely be a new term in use. African American English as a Rule Governed Variety African American English is regarded by linguists as a rule governed, systematized variety (Wolfram, 1970; Green, 2002; Green, 2007; Rickford 1999, among many others). systematic variety of English, we are less likely to have negative attitudes toward it and
43 indeed a rul e governed, systematized variety, linguists have also attempted to refute popular belief that African American English is a lazy, illogical form of Academic English. African American English is just as logical as any other variety, simply in different way s with different rules. An anecdote provided by Fasold (1999, p. 3) highlights the perception that many people have about African American English: Question : Certainly Ebonic is bad English, in the same sense that Syntax Green (2007) gives an extremely comprehensive account of the grammatical system of African American English, and in this discussion I will only highlight some of the features of the variety that I discussed with the participants during the six lessons. A spectual Be and Auxiliary Be African American English differentiates between Aspectual Be and Auxiliary Be The distinction between these two forms of be marks a significant difference between African American English and Academic English. For non Africa n American English speaking people, this difference is one that is often pointed out as illogical, and during Lesson Two, this was the first distinction that I presented to the participants about how African American English grammar differs from Academic English grammar. Mono dialectal speakers of Academic English, who are unacquainted with African American conjugate the verb to b
44 Aspectual be indicates an habitual act ion or occurrence. The following are some example of Aspectual B e with the corresponding Academic English gloss underneath: (a) Dee be running. Dee is usually/often running. (b) Dee is running. Dee is running. (c) Dee running. Dee is runni ng. BIN (been) The use of the stressed BIN is also frequently heard among African American dialectal Academic English speakers, who often assume that the speaker has simply omitted the form at is a crucial element in the sentence. Double/multiple negation Although multiple negation is quite stigmatized in African American English, and many Romance languages use it to this date, such as in the following 1 : In French we see: (d) Je ne sais pas. 1 These examples are my own.
45 I (neg) know (neg) Or, even more marked in English, yet the standard in French: (e) Je ne sais rien. I (neg) know nothing. In African American English we see negation such as (Green 2007): (f) never have no problems. Existential It used to indicate that something exists, and it usually occurs in place of there or there is 2 (g) Sometimes it no chalk, no books, no teacher. times any chalk, any Question f ormation African American English differs in several ways in question formation, especially in the lack of auxiliary inversion. (h) I wanted to know could they do it for me? if they could 2 Interestingly, speakers from the New York area who only control AE can be heard using this form of it, from Yiddish, yet simply because I have only heard it from Jewish speakers does not necessarily mea n that only they use this variant.
46 Phonology Phonological patterns differ between Academic English and African American English, and while both are rule governed, the rules for each are somewhat different. s often the case that the final clusters st sk, sp, pt, kt (often spelled ct nd and ld ) shown below: African American English pronunciations are on the top, with Academic English written forms undern eath. (i) lis (j) des (k) was (l) accep (m) contac (n) spen (o) buil (Green, 2002, p. 679) Devoicing the consonant cluster is not a random, haphazard occurrence. It is rule governed, and can be described by a voicing generalization, which Green dialectal speakers of Academic Engli sh, and also even African American English speakers believe that when
47 sounds are omitted it is simply because the speaker dropped a sound, and they are unaware that this deletion is only in certain phonological environments. That they do not know the dele tion rule factors into their assessment that African American English is Consonants and consonant clusters are phonologically produced in different ways within African American English and Acade mic English, depending on the position of the consonants within the word. We examined this during Lesson Three: Phonology, dialectal rather than reading mistakes. F or example, in a word initial environment in Academic English, voiced th produces the sound // in Academic English, whereas /d/ may occur in African American English, producing the words dese and dat in African American English. In a word internal enviro nment the same voiced th // may occur as voiced /v/ in African American English, producing brova in African American English. Also, the voiceless th internally in Academic English can be produced as /f/ in the same environment in Afr ican American English, producing the word birfday. (Green, 2002). These differences are very salient to Academic English speakers, yet because they do not know that there are specific phonological patterns, they are apt to ascribe negative characteristic s to speakers who make these phonological substitutions. It is imperative that teachers know about these differences so that they are able to differentiate between a reading and comprehension mistake, and a dialectal, linguistic difference. While the com plexity of this system is greater than the analysis just presented, the above examples are given to substantiate that African American English speakers
48 are following specific syntactic and phonological rules, and not haphazardly omitting sounds. Green not patterns in African American English do not sound like patterns at all, and such generalizations play crucial ro les in developing strategies in helping speakers see the differences between the African American English and mainstream English Education Language Laws Although there are laws that are intended to serve linguistic minority students, most schools do not have policies that adequat ely serve the needs of African American English speaking students. T here are policies that protect language minority students and direct schools to create programs that take the ir language n eeds into account, as decr eed in Lau v. Nicholson in 1974, however minority students are still very under served linguistically. Ann Arbor Because parents were concerned that their children were not being adequately served in their public schools and wer e graduating without the necessary academic skills, in 1979 the parents of eleven (although Newell & Chambers, 1982 account for fifteen) African American children won a lawsuit against the school district of Ann Arbor. Judge Joiner decreed that it was not deficits in the African American English speaking students that were causing low academic performance, but rather the fact that the teachers were not taking African American English into account when teaching the students, thus contributing to the students
49 that did exist was in the form of unconscious negative attitudes formed by teachers towards 193). The teachers in the district were required to attend a twenty hour tr aining session identifying speakers of the dialect, ways of distinguishing mistakes in reading from differences in pronunciation, and strategies for helping children switch fr om Black 194). Nonetheless, Labov believes that mented (p. 194). This current research project included both work on attitudinal change as well as giving participants specific strategies to use with their students to help enable students to master Academic English and learn to code switch when appropri ate. This research therefore furthers work that has been previously executed, and attempts to fill in a gap noted by Labov that was unaddressed in previous attempts to help close this linguistic achievement gap. Smitherman (1983) notes that there was a ne gative reaction within the African including the Conference Report from the proceedings of The National Institute of Education (NIE), indicate that the public perception surr ounding the Ann Arbor decision was that teachers were either going to teach African American children how to speak
50 African American English, or that they were going to use African American English as the sole language of instruction for the students (Newel l & Chambers, 1982, although the conference took place in 1980). This is the same reaction many Americans had when the 1996 Oakland Controversy erupted, as will be discussed. E ven though African American children have been shown to have unique language African American English speakers (Baugh 1995, p. 91). This renders a situation in which there is documentation and there are c ourt rulings asserting that African American English speakers have unique linguistic needs, yet funding is absent to aid in their language acquisition and overall academic achievem ent. The Oakland Controversy In 1996 the media in the United States erupt ed with misinformation about a language program that the Oakland, California school district proposed. The media believed that the teachers in Oakland, California were going to start teaching their students, who were primarily African American, how to spe ak African American English, in African American English. In actuality, the Oakland School Board was aware that their African American students were speaking a variety other than Academic English and they wanted to was going to educate the teachers about African American English and give them strategies to use with their students to help them a cquire Academic English since the Board saw that the African American students were performing more poorly than their
51 White counterparts. This would be done mainly through the use of Contrastive Analysis, a method in which differences between the varietie s are explicitly pointed out and discussed. Linguists wrote to the major newspapers to explain what was actually being proposed, but as John Rickford (1999, pp. 270 271) candidly notes, many were not published: In the case at hand, the mainstream view was that Ebonics itself was street slang, and that Oakland teachers were going to teach in it, or allow students to talk or write in it instead of in English. It was in response to this misrepresentation of Ebonics and the Oakland resolutions that editorials, Op Ed pieces (texts submitted to newspaper by the public), letters to the editor, cartoons, and agitated calls to radio talk shows were directed, and attempts to get alternative viewpoints aired were often very difficult, especially in the most prestigiou s media. For instance, although the New York Times published several editorials and Op Ed pieces critical either of Ebonics or the Oakland resolutions, linguists' attempts to get them to present a different viewpoint were all unsuccessful. I know of at l east four Op Ed submissions which they summarily rejected (by Salikoko Mufwene, by Geoffrey Pullum, by Gene Searchinger, and myself), and there were undoubedtedly others. Similarly, other linguists (like Geneva Smitherman) had experiences similar to mine, in which leading television stations would do one and two hour int erviews with them on the Ebonics issue, but never us e any of it in their broadcasts. Although many people do not specifically remember the Oakland Controversy today, many people in the cou what the media said would be taught in the schools. When I describe to people who have not been exposed to the idea of African American English what the research is about, including the participants in this study, I am often asked if I am referring to what the media was saying, they did become familiar with the term and the idea that there is a different way of speakin g that is utilized by many African Americans. They
52 are, however, very rarely also informed that this variety is rule governed and not a lazy, slang manner of speaking. Strategies for Aiding in the Acquisition of Academic English One might think that beca use African American children are exposed to Academic English throughout the school day for numerous years that they would naturally acquire Academic English, however as Wolfram (1970) points out, children gravitate towards the speech of their peers and no t towards the speech of the schools (p. 14), and in order to acquire Academic English students must notice the differences between it and the variety they speak. They must also be willing to acquire this variety, and this idea of volition will be discusse d shortly. Baugh (1995) asserts that the linguistic situation in Hawaii, where Hawaiian Creole English is the L1 for many students, is the most similar case in the United States for comparison to African American English Many of these students enter the educational system speaking H awaiian C reole E nglish in the same manner as many African American students enter the educational system only speaking African American English. F or both groups there is the challenge of ensuring that the students acquire lit eracy and code switching skills in order to master Academic English and have access to all social amenities brought on by speaking what is gener ally the most valued code in our society. We see a somewhat similar situation with the usage of Standard French, Non standard French, and English in the schools and classrooms in the Franco Ontarian classrooms in Canada. The linguistic situation is Ontario a very complex one and I will not delve into the details here, but an overall issue is that students enter the educational
53 system with varying degrees of French fluency, including non standard French fluency, to the extent that students have limited access to a variety of social conte xts where the use of French is conventional, they cannot master the varieties of French necessary for them to be able to function as francophones. To the extent that the variety of French used at school differs form that used at home, they have difficulty performing well in classroom contexts. This difficulty is accentuated for many, if not most, of these students who speak nonstandard French 24, italics mine) The difficulty that the non standard French speaking children experience can be compare d to the difficulty of African American English speaking students in the United States. They speak a variety of the standard, yet it is not quite the standard, and because of their lack of command of the Standard variety they perform more poorly. The sit uation with African American English is somewhat different from that of other countries and other non standard varieties, though, as determining which variety to speak is very culturally laden, and many students do not ever acquire the standard, or Academi c English, in the United States. In contrast to other language situations in which language and culture are intertwined and students enter the educational system not knowing the standard variety, in the case of African American English many students exit the system without ever having acquired it. Currently in the United States the language of power is Academic English. Although all students may not consciously be aware of the power differences, many are likely aware that the way they speak is different from Academic English, and they may associate their variety with their family and their friends. While in order to acquire Academic English students must be aware
54 for the African American speaking students. Many students believe that acquiring the Negotiating Language and Identity In an ethnograph ic study Fordham (1999) conducted with stud ents at a high school she referred rican American student despairingly described th e way in which her mother spoke on the phone. The student said about her mother: u know not. (p. 284, brackets in the or iginal) This account clearly ide ntifies that the student expected s an African American English speaker. The mother seems to be aware of the able to code switch between African American English and Academic English when she was not speaking to other members of her family. Ogbu (1999) also speech forms in relation to a social situation. The student say s, The intervie wer then says and the student responds
55 We see here that both the student and the mother are aware of the social implications of using one variety over the other. If either of them were to use Academi c English in an environment in which African American English were expected they would sound out of place and potentially be rebuked by the interlocutors. On the oth er hand, if they were to use African American English in a situation in which Academic Eng lish were expected, we can imagine that there wou ld be consequences for that as well. Acting White E nglish with their peers and community w hile also using and mastering Academic English, a variety some see as necessary for social mobility. Fordham (1999) surveyed students American comm unity is aware that there ar e certain situations in which Academic English is expected, however to use Academic English i n a situation that calls for African American English such as when speaking with other African Americans in a more casual setting, can 1999). Fordham & Ogbu (1986) assert that conforming t acting W They assert that academic o not want to must rebuke the school requirements to avoid being seen as such. With nearly every marker indicating that African American students are faring worse academically than their White c ounterparts across major subjects (Green 2002;
56 Becker & Luthar 2002), it is disheartening that many African American students are not acquiring the language of power for fear of retaliation by racial co members. Noticing Hypothesis Schmidt (1990) has po teaching mono dialectal African American speaking students Academic English because they have likely bee n exposed to Academic English throughout their school years. Most teachers control Academic English, and even if students were not consistently exposed to Academic English from their teachers, they have taken English grammar classes since primary school. I have found that when I describe to laypeople that many students have heard Academic English for years in school yet they have not acquired it, I am often told that the students are lazy because they have not acquired it, and/or that they have had gramma r classes since they started first grade and that it is are not intelligent enough to acquire it because they would have already acquired it if they were capable of doing so. However, because it is likely that no teachers specifically pointed out the ch the varieties differed, thus adding to the reasons why they never fluently acquired Academic English. Although students must at some level notice the differences between the varieties in order to master each one, in the case of African American English in the United States there is a strong cultural factor at play which keeps many students from acquiring Academic English, and that is that they want to belong to a peer
57 group that values speaking African American English and specifically rebukes the use o f Academic English. Some prominent African Americans, such as Lisa Delpit, who has written much about the difficulties of acquiring Academic English and gaining entry into what she xplicitly taught the differences between the varieties, they would have been able to be successful academically and socially much faster and on a much easier path. Explicitly Teaching Differences Delpit (1995) promotes the idea that teachers explicitly tea ch differences relate to linguistic forms, communicative strategies, and presentation of self; that is, Fogel & Ehri (2000) and Delpit (1995) posit that without explicit instruction as to how to identify the similarities and differences between African American English and Academic English students will have a hard time differentiating between the two systems Green (2002) also suggest s that teachers make clear the differences between African American English and Academic English in order to assist students in recognizing the differences. Fogel & Ehri (2000) conducted an experiment in which students were put into one of three categories: those who received no ex plicit instruction regarding Academic English but who were exposed to it through a story, those who received explicit instruction, and those who received both explicit instruction and were also able to
58 practice using the two varieties. Results show that s tudents with no explicit instruction and those with explicit instruction but no practice fared similarly on post t est measures, while students who were both exposed and practiced performed significantly better than the other two groups. In light of these results, the lessons I taught the participants included specific ideas for teachers to use with their students to ensure that the students became aware of the differences between the varieties and that they were able to identify when they were using which variety. Fogel & Ehri (2000) exhibited by the students, writing several sentences containing varied examples of the nonstandard forms, explaining the SE [ Academic English ] rules and having students pr actice transforming the nonstandard sentences into SE [ Academic English ] while providing corrective feedback, and then having the students practice using the forms in their free writing accompanied by corrective feedback In discussing the acqu comparing are unlikely to occur naturally for learners when many aspects of the L1 and Although Academic English and African American English are different varieties, they share numerous similarities, especially considering that they are both forms of English. Because there are so many similarities, and because African American English is stigmatized and Ac ademic English is not, there are a variety of hurdles over which mono dialectal speakers of African American English must jump in order to successfully
59 must notice the differences between these two related language varieties if they are to successfully acquire the other. Contrastive Analysis There have been a variety of ways of instructing African American English speakers in the rules of Academic English, with Contra stive Analysis (such as described in detail in Rickford, 2005) at the forefront as one of the most respected and successful methods. Contrastive analysis is the process in which differences between varieties are explicitly identified and students are taug ht to see the similariti es and differences between them In this approach, teachers talk about linguistic differences with their students and help them to understand and pinpoint the differences between the varieties in order to master the equivalent Acad emic English forms. When students become meta cognitively aware of linguistic differences they are on a faster path towards becoming bi dialectal. An example of a contrastive analysis activity would be for a teacher to give the example John BIN hungry a nd elicit from them that John has been hungry for a long time. This can be done with any grammatical category that differs between African American English and Academic English. An important point that Green (2002) makes is that one does not have to be a fluent speaker of African American English in order to recognize that it is a fully formed system, and that simply recognizing elements of constructively assist their s tudents. Teachers, thus, can learn about African American style game, creating a
60 formal writing or speaki ng event, talking about register and domain and how our language shifts depending on our audience, and journaling. Pointing out similarities and differences between any forms of speech will bring to focus the idea that there are many ways of expressing t he same idea, and that African American English and Academic English both have valid and expressive means. American English in order to assist students. She uses African A merican English throughout her article, and in discussing the need for teachers to recognize and accept of reading teachers in Black speech just being able to dig it, no American English fluently in order to be able to assist students who speak it as an L1 to acquire Academic English. Likewise, in the current res earch I am not suggesting to the participants that they become fluent or bi dialectal in African American English if they do not already speak it. Instead, I am suggesting that they become familiar with its syntactic and phonological structure, as well as issues surrounding the use of a non standard variety, and use this knowledge to enable their students to acquire Academic English in a meaningful way. Several studi es have shown success using the contrastive analysis approach, including the Dekalb Bidia lectal Communication Program used in Georgia, which showed improvements in verbal scores among African American English speaking students (Harris Wright 1999, as cited in Siegel 1999 ), as well as the Academic English Mastery Program in the Los Angeles Uni fied School District. The 1998 1999 Academic
61 Unified School District, shows that for African American English speaking students, ucationally meaningful difference between experimental and control groups at the end of the program as measured by the Language Assessment Writing Test. AEMP program participants outperformed those amela, 2000 p. 3). This program shows that targeted instruction for speakers of dialects other than Academic English can be successful. My personal pilot data that I gathered in a high school in the Harlem section of New York City showed a decrease in th e usage of African American English features and an increase in the usage of Academic English forms in writing samples from the ninth grade students after I gave professional development sessions to the ninth grade English teachers. After the professional development sessions, in which I taught the teachers about the rule governed nature of African American English, focusing specifically on syntax and phonology as well as discussing cultural and social issues surrounding the acquisition of Academic English 5.25%, which the administration credited to my work with the teachers. The teachers took the information I taught them and incorporated it into their teaching practice with their students. Another way of teaching students methods of acquiring Academic English was explored by Pandy (2000). Pandy Academic English as an L2 to African American English speaking students. Her study produced positive result s in the sense that students
62 improve d their Academic English skills, however there is not a tremendous amount of social support in the literature for treating Academic English speakers as non native En glish speakers (Baugh, 1995) Remembering the social backlash in Oakland when teachers were going to incorporate African American English into the classroom yet not teach it to the students, we must be very careful to consider social implications regardin g how the practice is described and how programs are labeled. Valuing African American English that it is a valid and community accepted way of speaking. If teachers are u neducated about African American English, or if they are unwilling to accept it as a variety, there will be little progress (Harper et al., 1998). There are many ways of showing that one values a variety, and Ladson Billings (2000) agrees that what she te them to school (p. 210). Finding strategies to encourage reading comprehension is a skill with which many English and other content area teachers grapple. Bell & Clark (1998) show that providing African American children with story images depicting African American, as opposed to White children, improved recall of events when the characters were African American, and they interpret these findings by suggest children process information more efficiently when it incorporates their socio cultural Teachers can also openly talk about different varieties other than African American English and Academic English to show that there are many ways of speaking, such as the Southern Variety, different New York boroughs varieties, Appalachian
63 variety, Chicano variety, etc. While it would help if a teacher is somewhat knowledgeable about different varieties, if their perso nal variety is different than the an online video of someone speaking in a different variety and play this for their class and discuss the differences. Shirley B different linguistic practices and ideologies in language communities and how the cultural values and language practices that children in two different communities were exposed to readied (or in this case did not ready) them for school and academic success. Brice Heath worked alongside students, teachers, parents, and community members and found that the differing cultural and linguistic practices in the communities greatly affected the st that would be expected of them in school. The two towns she worked with were called more linguistically pre pared for the types of questions that would be asked of them and the types of activities in which they would participate in during class activities, both sets of students were not habituated to responding to teachers in the ways that the teachers expected, Students from Trackton were not exposed to the types of questions that teachers would ask them in school and they were unaccustomed to the time limits and scheduling of activities that teachers placed on them. They produced stories and narratives that did not conform to the strict patterns that the teachers expected, and because of this it was easy for teachers to believe that the students were not as smart.
64 question. I get blank stares to my question. Yet when I am making statements or telling stories which interest th statements reflect the fact that students from Trackton are not accustomed to being asked the types of questions that teachers ask of them, and because of this, it is easy for teachers to think tha t they might have a hearing problem or are otherwise academically deficient. Students from both communities were also unaccustomed to the types of indirect requests that teachers employed with students in attempts to elicit a certain behavior. Says Bric (p. 280). When the norms and expectations between groups of people, in this case studen ts and teachers, are different from each other, it is easy to ascribe negative traits or values to the other group. Linguistic research shows that negative traits are often ascribed to speakers of African American English, and the current study investigat es how these negative attitudes might be mitigated, which could ultimately lead to more norms of speaking and interaction into account. It is important for teachers to note that differences do not necessarily equal deficiencies. Literacy and African American English Across the board, Second Language Acquisition studies sho w that when students are taught to read in their L1, regardless of which language is used in their c lassrooms, they acquire literacy skills more readily than when they are taught in their
65 L2 first (e.g. Stewart, 1969; Goodman 1969; Wolfram et al. 1998). Siegel (1999) reviews several experiments, such as Ravel & Thomas (1985), Bull (1990), Osterberg (1 961), and Siegel (1992, 1997), among others, that indicate that using the L1 aides acquisition of literacy skills as well as other content area skills. Dialect Readers Simpkins & Simpkins (1981 as cited in Rickford & Rickford, 1995 ) show that for African American English speaking students, learning to read in African American English before Academic English accelerated acquisition and this program using dialect readers became quite controversial in at tempt to aid African American English speaking students bridge the gap between African American English and Academic English scores were lagging far behind their W hite counterparts and linguists were looking for ways to increase literacy skills (Rickford & Rickford 1995). Simpkins et al. (1977, as cited in Rickford & Rickford 1995) cr eated the Bridge reading series, which consisted of three stages. First, children would read stories written exclusively in African Ame rican English, next they would read a series written partly in African American English and partly in Academic English completely in Academic English Rickford & Rickford (1995) cite Simpkins & Simpki ns (1981) who discuss the 417 students who took part in the Bridge series and say that for grades 7 12, the average gain in grade equivalent scores for the group using Bridge was 6.2 months for four months of instruction compared to only an average of 1.6 months of instruction for students in their regular scheduled classroom reading activities. The group using Bridge exceeded the normative level (four months gain for four months of instruction), many of them for the first time in their academic careers. ( p. 238, as cited in Rickford & Rickford 1995, p. 113, italics in original)
66 The program was clearly successful, yet there was such a strong negative reaction from parents and teachers t hat the program was quickly disbanded The positive outcome of literac y in conjunction with a program that supports code switching is immediately evident. However, we see again the impressive impact a community can have regarding an educational program. The Bridge series takes the concept of teaching students in their L1 an d then transitioning into their L2, a practice often used with speakers of other languages. However, African American English is so stigmatized, and there is such fear in the African American community that African American English speaking students will be denied access to Academic English 3 that a program that had been shown to effectively enable students to acquire literacy skills at a faster rate and simultaneously code switch between the two was disbanded; the publishers quickly ceased publication of the readers. Rickford & Rickford (1995) argue that even though the Bridge series met with serious community objections when first introduced, it is still a viable method of enhancing literacy skills, but that it needs to be implemented with support from th e communities Dialect Differences Versus Reading Comprehension Errors Goodman (1969) outlined several ways in which African American English speaking children can be helped to acquire literacy in Academic English and suggest s that the most viable method is that children are taught to read based on the skills and language that they already possess, and that Likewise, Piestrup (1973) condu cted a large investigation of six different teaching styles 3 Additionally, the media was inaccurately reporting the purpose and methodology of the program, which led to the widespread misunderstanding.
67 with African American English speaking students, and found that th e teachers who used the student s Academic English had studen ts who were the most successful in terms of reading achievement. The common thread here, and one which will be explored further along in the discussion, is that students are more successful when teachers are separating the teaching of reading skills with the teaching of phonological differences between the different linguistic varieties. Wolfram (1970) asserts that if children can read sentences that are in Academic English in their own variety (African American English), they have successfully read the se class black child reads a standard sentence such as as Jane go to Mary house he is considered to have read it properly, since third person singular s and possessive s suffixial absence are part of the lower teacher who understands the syntax and phonology of African American English is in the position to accept reading rendered in African American English as a legitimate und erstanding of the text (p. 31). into account when teaching reading skills also reflects elements that Labov strongly insists upon, including that teachers not confuse dialect diff erences with reading mistakes and that teachers only correct mistakes when not teaching reading comprehension (Labov 1995) As is explored in Lesson Two: Phonology of the current research, Delpit (1995) makes a clear case that the practice of teachers f ocusing on
68 frequently often causes students to refrain from speaking when they are corrected in suc h a manner (Gopaul McNicol et al., 1998), and Harper et al. (1998) expressly state that if teachers do not take African American English into account when dealing with African American English speaking students, they will not be successful with their stude nts. The Necessity of Acquiring Academic English Many researchers believe that it is important for African American English speaking children to acquire Academic English. Charity et al. (2004) found that children who were more familiar with Academic En glish scored higher on reading assessments, which lends support to the method of famili arizing African American English speaking students with Academic English They suggest a concept they refer to as dialect awareness and say familiarity with SE [ Acade mic English ] may reflect a form of metalinguistic insight that facilitates reading acquisition more generally. That is given equal exposure to SE [Academic English ] and equivalent oral language proficiency, some children may be more attuned to linguistic variation in their environments, whereas others are less inclined to notice or appreciate that different ways of speaking are used in some contexts and not in others and by some speakers but not others. (p. 1353) They then hypothesize that children who are more meta linguistically aware might learn to read more easily because they are more aware of language and its differ ent manifestations (p. 1353). of African American children in the study di d appear to have more difficult y spelling dialect sensitive (e.g. inflections) as compared to dialect neutral (e.g. consonant and vowel patterns) orthographic patterns that their peers, and their errors were consistent with linguistic mismatches between sp oken African American English and standard
69 et al. awareness and says that her results are in line with this analysis. According to not given the opportunity to become proficient in responding in mainstream contexts, they may be permanently that African American children, as opposed to White children, m ay tend to tell stories in a more loosely organized way known as topic associating as opposed to a very topic centered manner expected by teachers and more aligned with stories told by White children (Michaels 1981), however Hyon & Sulzby (1994) have show n that while this is true, African American students also tell more topic centered stories under different circumstances. When teachers are expecting a certain type of story they may react unfavorably to stories that do not follow this specific pattern (M ichaels 1981 Brice Heath 2009 ). We must therefore ensure that children who do not speak Academic English as an L1 be taught about what is expected of them in an Academic English environment. The ultimate goal is for students to become bi dialectal and able to code switch between African American English and Academic English. The purpose is additive, not subtractive; it is not to remove African American English form their repertoire, yet to assist them in mastering Academic English for academic purposes career success, and social and economic mobility. Any program should be centered around valuing the variety of speech that students bring to school while also ensuring that they gain the variety that our society currently values most, Academic English, and entry into the
70 The Need for This Study This study is necessitated by many interconnected pieces of research. It is unique in that it combines sociolinguistic and psycholinguistic methodologies, as there is sociolinguistic research documenting the need and strategies for bridging the language towards different language varieties in a psycholinguistic manner. This study attacks the question of ass measure and several explicit measures, and synthesizes the findings across subfields. been discus sed, none of these measure implicit attitudes. It is important to understand student achievement. We may not understand the larger picture if we do not have a true sense of the prevailing ideologies if participants do not identify their true, deepest seated beliefs on a paper and pencil questionnaire or survey. Baugh (1995) acknowledges that some educators have attempted to aid African American English speaking stude done so based on policies that are uninformed by relevant linguistic research, thereby for the current work. A dialect interference in reading as well as the authentic types of errors that occur in learning to read, the teaching of standard English is usually done in a haphazard and (p. 12). It is therefore important that teachers are given skills and
71 manner in which teachers are correcting, and this current study provides teachers with such strategi es. Wolfram (1970) asserts that in order for a teacher to successfully aid a student in the acquisition of reading the teacher must know the rules of African American English and that otherwise they will not know if a child is reading in a different mann er from the text due to differences in African American English or actual reading miscues (p. 18). It is unsurprising, then, that Hollie (2001) notes that in the Linguistic Affirmation Program (LAP) in Los Angeles, the element of the program with which th e teachers most struggled was with knowing the phonological elements that differ between African American English and Academic English and applying this knowledge to assist learners during phonics instruction (p. 58). In this current research the lessons address the syntactic and phonological patterns of African American English, thus equipping the participants to better understand their students and to be able to identify differences between the varieties in a systematic way, both for themselves and for t heir students. Wolfram (1998) says that language professionals with an authentic understanding of the nature of language diversity have a responsibility and a challenge to educate other professionals, practitioners, and the general public about these [ling uistic] issues if we wish to avoid repeated controversies like the one that surfaced in widespread idea or social practice with important consequences that is invalidated by his own data is obligated to bring this error to the attention of the widest possible
72 172). His discussion is embedded within an analysis of the Ann Arbor Case, and this principle can be directly applied to the current research. Previous research indicates that teachers tend to have negative attitudes towards African American English, be lieving that it is a faulty way to speak. Furthermore, the research shows that explicitly teaching students the differences between African American English and Academic English improves academic performance. It stands to reason, then, that adhering to t he Principle of Error Correction would involve educating teachers, as I did in the current research. Rickford (1997) discusses how there is vast knowledge about African American English within the linguistic community, however this information is not diss eminated to the general public and communities with which linguists work. This current project aims to do just that: disseminate information to the educators who work with minority students. that traditional teacher education has not made a significant impact on improving teacher attitudes Billings (2000) explicitly says teach African American service teachers be taught about African American English. It makes sense, then, for a linguist to enter into the equation and instruct teachers about Afric an American English. In particular, one reason to focus attention specifically on African American English is that of all non standard varieties spoken in the U.S, African American English has the highest numbers of speakers (Wiley & Lukes, 1996, p. 525).
73 Although one might believe that by the time African American students enter universities all of their linguistic needs have been met, this is not the case. Taylor pres of awareness of African American English among educators is that many African American students are sent to speech pathologists because of language differences, when in fact it is the differences that are apparent and not a specific disorder (Cole & Taylor, 1990). Smitherman (1998) describes entering the university system and failing a speech test, as she had not become bi dialectal by that point in time. She, along wit h other African American English speakers, was sent to a speech pathologist who was unable to diagnose any of the students with a disorder, rendering her unable to assist the students. Ultimately, the speech pathologist taught the students to how to prono unce, in a mid Western accent, all of the words on the speech test. All of the students then passed (p. 139). a stage in which some of the American populace is willing to see beyond skin color to ement of the experiment in which the matched guise methodology is used. The research presented thus far indicates the urgent need for student achievement to be raised, as well as for teachers to be informed of differing linguistic varieties and how languag e affects students and academic success. Over the years
74 there have been attempts, often mandated by law, to assist African American English speaking students acquire Academic English, yet the majority have not been successful or led to long term change. The goal of this study is not to measure student growth, but negative attitudes adversely affect student achievement, and because we know that when teachers have more positive students perform better academically, I wanted to test whether teaching teachers lessons in a particular manner would improve their attitudes. If these lessons are successful, we can then hope that that would lead to improved student performance, thus positively affecting the achievement gap and leading to more academic, social, and economic benefits for the students as they continue their education. Although no student data was collected for this study, the scope of the lessons taught to participants did include carefully crafted activities for the participants to engage in with their students in order to improve student performance. sh, as a key component to raising achievement and opportunity and an area that has been greatly overlooked when asking questions about how to improve student achievement in minority or under performing groups. Although this research did not involve stude nts directly, previous research indicates how important teacher attitudes are in rel ation to student achievement, and I therefore wanted to investigate whether my methods of teaching would positively affect teacher attitudes, with the trickle down result b eing that students knowledge. As will be described in the following chapters, the methods of teaching did
75 s the idea and usage of African American English, and participants anecdotally stated that they were able to apply principals they learned during the sessions with their students to further
76 CHA PTER 3 SOCIOLINGUISTIC METHODOLOGY Selecting the Participant Population could be changed following a series of lessons that I created, thus I wanted to find a group of teac hers with whom I could work on a regular basis. I decided that graduate students of education would be an ideal population because they would either already be in the classroom or entering it soon, and I thought that the information I would present would be quite valuable for them and their students. I have worked in public schools and know how difficult it can be to gain access to teachers, as well as how often schedules change, and I therefore thought that a graduate classroom would be more feasible tha n attempting to work with a group teachers through an individual school. Securing a Research Site Securing a research site was uncomplicated. I contacted one college and one university in New York City regarding the potential partnership and the first ch airperson to reply had the most desirable situation. The university is located in Harlem and the participants were all graduate students of Education in the English department. The chairperson and I met and talked about the project and what would be pos sible. He and I shared similar teaching backgrounds, both of us having previously taught with Teach For America, and he was immediately receptive to discussing my ideas and desires for the experiment. I described what I envisioned for the research and he was very willing for me to work with his department. We discussed a variety of possible scenarios for executing the project and h e suggested that I contact three professors who we
77 focused on th e issues of literacy, bilingualism, and special education. He thought that my research would fit well with the literacy and bilingual elements of the course, and he welcomed my proposed topics and methods of data collection and evaluation. I contacted the three pr ofessors and spoke with each of them Two of the professors taught graduate sections and one of the professors taught an undergraduate section. All three professors were women, and the two professors who taught the graduate sections were African American (one described herself as African Heritage), and the professor who taught the undergraduate section was White. The two graduate professors very enthusiastically agreed for me to collaborate with them and to use class time to carry out all aspects of the experiment, while the undergraduate professor said that she could not give up any class time for the project. The experiment was thus run with the two graduate professors, one of whom had one section of participants and the other of who m had two. Class Schedules and Data Collection had two sections of students and Professor B had one section, and I met with each section once per week in the evenings. Classes were s cheduled to be two and a half hours long, although from time to time they lasted only two hours. On the days that I taught I was given about an hour, although because the classes consisted heavily of student presentations that sometimes exceeded their all otted time, there were a few days when I was given less than an hour. In total, I spent three days running the Pre Psycholinguistic Experiment (which will be described in Chapter 5), six days (thus six hours) teaching a total of six lessons, and three day s running the Post Psycholinguistic Experiment.
78 with African American English Both pr ofessors were bi dialectal in African American English and Academic English They described growing up speaking African American English but be c oming adept at code switching into Academic English for academic and social purposes. They were both aware of the fact that African American English differs in significant ways from Academic English. However, at the start of the semester, and therefore at the beginning of the experiment, they were un aware of the systematic rules governing it. They both had native speaker knowledge and intuition about African American English, which proved to be helpful in some instances where par ticipants were challengi ng the validity and grammar of African American English The professors, using their clout as both professor and native speaker, were able to underscore that what I was teaching was correct and legitimate. Both professors made it very clear to their stu dents that they valued and believed in the research that I was presenting. They did this through giving clear expectations that that the participants pay attention to my lessons and participate actively, and they verbally underscored the points that I mad e, saying how they were true and that they either experienced certain situations themselves or saw them play out in a classroom setting. I believe that the willingness of both professors to have me in their classroom, and that they indicated to the partic ipants how glad they were to have me there, helped pave the way towards the participants engaging with the lessons I taught. Participants T he participants were all graduate students in the English department. The majority currently taught in public school s, with a few teaching at charter schools, and
79 some were not yet teaching but intended to teach in the near future. Table 3 1 below presents the characteristics of the participants, and a description will follow. Table 3 rmation African American Asian Hispanic Mixed White Totals Race 9 (20%) 3 (7%) 3 (7%) 4 (9%) 25 (57%) 51 Female 6 (24%) 3 (12%) 1 (4%) 3 (12%) 12 (48%) 25 Male 3 (16%) 0 (0%) 2 (11%) 1 (5%) 13 (68%) 19 Age 20 29 8 (30%) 2 (7%) 1 (4%) 3 (11%) 13 (48%) 27 Age 30 39 0 (0%) 1 (13%) 2 (25%) 1 (13%) 4 (50%) 8 Age 40 49 1 (25%) 0 (0%) 0 (0%) 0 (0%) 3 (75%) 4 Age 50 59 0 (0%) 0 (0%) 0 (0%) 0 (0%) 4 (100%) 4 From New York 8 (31%) 2 (8%) 2 (8%) 2 (8%) 12 (46%) 26 Not f rom New York 1 (6%) 1 (6%) 1 (6%) 2 (11%) 13 (72%) 18 Cu rrently t eaching 4 (16%) 2 (8%) 1 (4%) 2 (8%) 16 (64%) 25 Not currently t eaching 1 (14%) 1 (14%) 2 (29%) 1 (14%) 2 (29%) 7 Years teaching: l ow 5 (22%) 2 (9%) 0 (0%) 2 (9%) 14 (61%) 23 Years teaching: m edium Unknown Unknown 1 Unknown Unknown -Years teaching: h igh Unknown Unknown Unknown 2 Unknown -
80 The participants who were currently teaching had student populations that were mainly Hispanic or a Hispanic/African American mixture. Parti cipants ranged in age from 22 57, many et hnicities were represented, and the majority of the participants were of Caucasian, A frican American, Mixed, Hispanic, or of Asian descent. The other categories included only one participant per category, and in light of not identifying those two p articipants, they were not included in this table. Not all of the potential participants 4 from the three sections participated There were seven in total who chose not to participa te, and while I did not invade their privacy and decision by asking why, it seemed that they were wary of the subject matter and the formality of an experiment. There were a few participants who participated in either the Pre Assessment or the Post Assessment but not both. Researcher Introduction On the first d ay of class of the semester the professors asked me to introduce myself. I briefly described my academic background and that I would be working with them on language issues within education I did not use the te rms Academic English or African American En glish since I would shortly be administering the Pre Assessment Questionnaire and the Psycholinguistic Experiment within the next few weeks, and I did not want to influence the results in any way I described that I was working with the classes in conjunc tion with my dissertation and that I was giving them the opportunity to participate in an experiment that sought to explore the relationships between language and education. Participants were told that if they consented to the experiment that they 4 Participant numbers do not add up consistently due to some participants fil ling out the demographics questionnaire and some not, as well as some participating in either the pre or post experiment/surveys but not the other. Additionally, there were some participants who did not fit into any of the categories in the table, and I d id not include them in the table as indentifying their race/nationality would potentially identify them personally. Percentages do not all equal 100 due to rounding.
81 would h ear four short stories and select adjectives that best represented the speakers, and that after that I would ask for basic demographic information. It was not until the first class session that I described that my research was focused specifically on Acad emic and African American English Participants were told that they were not required to participate in the experiment by answering the Pre and Post Attitude Questionnaire, Demographics Questionnaire, or Psycholinguistic Experiment, however the professors explained that the lessons I taught would be part of the class material and that all class members were responsible for the information I presented and the related class discussions. Several of the class members had questions that were generally related to the invasiveness of the Psycholinguistic Experiment and whether their data was personally identifiable. I explained in more detail what the participants would do during the Psycholinguistic Experiment, namely to sit in front of a computer wearing head phones with their hands placed on a plastic response box and press buttons on the box in response to adjectives they saw on the screen. One participant asked whether I would be able to give her her personal results after the experiment, and I responded th at while theoretically it was technically possible to compare participants I would be looking at the data in the aggregate and that I would not be able to give participants individual feedback, and I would not identify them in any way when describing the d ata. I emphasized that the data and feedback would be anonymous and that their names and identifying information would not be made available to anyone beside myself, and that I needed to have it only in order to be able to pull their personal data out of the experiment if they changed their minds at any point. I explained that they would shortly
82 be receiving an identification number used solely for the experiment, and that they would never write their name on any assessment. Once all of the questions were answered I distributed IRB forms and instructed class members who would like to participate to read them, ask any questions, sign two copies, and return one to me. After collecting the IRB forms I assigned parti cipants their did not want the participants to think that I had access to or was using their university identification number, so instead I referred to their personal identification number as and only two participants later asked what ART stood for. I said that it represented the this when we were alone in the experimentation room, so the entire class did not hear these conversations. I considered whether the p articipants would keep their personal ART numbers or trade with a classm ate before the Post Experiment if for any reason t hey disagreed with the material I was teaching, if they were disgruntled for any reason, or if they just wanted to play a prank on me or the experiment. The potential problem was avoided because by the end of the semester I knew the participants names a nd when the y presented their ART number in the experimentation room I unobtrusively checked their number against the form that contained both the ART numbers and the participants names. No participants had switched numbers. Three participants had mispla ced the form with their ART number that they were given at the beginning of the semester, however they believed they knew their number. Because I had previously recorded
83 their ART number with their name I was able to easily locate their number and verify that they remembered correctly. Pre Assessment Before any content discussion s tarted, participants took a Pre A ssessment that gauged their prior knowledge of African American English. A copy of this assessment is located in Appendix A, and the results o f the assessment are described in detail in Chapter 4. Participants spent about ten minutes completing the Assessment. The Pre Assessment questionnaire was administered during Lesson One, after assigning the ART numbers. I n order to avoid biasing parti to the q uestionnaire with respect to their prior knowledge of African American English and the idea of non standard varieties measure prior knowledge of the topics we would be di description was intentionally vague, and no participants asked for clarification before beginning the assessment or while completing it. Participants wrote their ART numbers at the top of the page where indicated. The assessment q uestions were designed to ascertain whether the participants were aware of African American English as a system, whether they were familiar with news coverage of the Oakland Controversy in 1996, whether they knew of syntactic differences between African Am erican English and Academic English, and whether they knew of any strategies to help students code switch between African American English and Academic English. The assessment con tained the following questions : 1.
84 2A. Are you familiar with the Oakland, California School Board Controversy of 1996? 2B. If so, please describe what you remember. 3A. What does the following mean in African American English? John be running. 3B. What does the following mean in African American English? Alexis been done ate. 4. Do you know of any ways to help children who speak African American English master Academic (Standard) English? If so, please describe. 5. Is Academic English better than African American English? Please describe. An identical Post Assessment was given during the final class session, s ave for an additional question asking if the participants had implemented any of the suggested linguistic activities in their classrooms, which was phrased as follows: 6. Have you implemented any of the suggested linguistic activities? If so, which ones, how, and what were the results? A copy of the Post Assessment is located in Appendix F, and a discussion of the results is located in Chapter 4. Demographics Questionnaire The Demographics Questionnaire was administered directly after the Pre Psycholinguistic Assessment (which will be described in Chapter 5), and a copy is located in Appendix B. It asked race and race of parents, household income growing up, L1 and potential L2s, as well as several questions about the teaching status of the participants and their students. There was also a large b lank space left open for participants to describe any particular language issues that they had noticed with their students or any language circumstances on which they would like to comment.
85 Attitude Questionnaire The participants also completed a paper and pencil Attitude Questionnaire directly after completing the Pre Psycholinguistic Experiment, and again at the end of the semester after completing the Post Psycholinguistic Experiment. After they stepped out of the experimentation room they sat privately filled out the qu estionnaire, and put it into a manila envelope along with all of the other design was created with the intent of giving the participants a sense of anonymity aside from their identification number Th e Attitude Q uestionnaire was taken directly from Taylor (1973), though modified for the vocabulary of race. Any mentions of Black or Negro were substituted with African American. five questions, twelve th irteen Participants responded to questions on a five point likert scale. The full qu estionnaire is in Appendix C, and a discussion of how the questionnaire was coded and measured is discussed in Chapter 4. The Attitude Questionnaire was given because I wanted to have not only the psycholinguistic results, but also more standa rd paper and pencil results. This particular questionnaire was selected for its content relevance, as well as because even though it is dated in its language, it is still seen as the standard for the type of information it gathers. Not only does it phras e approximately half of the questions positively and approximately half of them negatively, it also gives participants a scale on which to base their judgments. In the Psycholinguistic Experiment the participants are forced to make a binary decision, whic h I correctly anticipated would disturb some participants, and with this questionnaire they are able to express more refined
86 judgments and attitudes due to the likert scale. While some participants in their exit surveys complained about the forced binary judgments they made with the Psycholinguistic Experiment, there were not any complaints about the likert scale of the Attitude Questionnaire. Sociolinguistic Lessons The scope and sequence of the lessons was systematically and carefully planned. Again, t here were six teaching days and each lesson lasted for approximately one hour. There was therefore a total of six hours of instruction and thus approximately six weeks (there were some breaks during the semester) between when the P re and Post E xperiment s were run Note that I only had contact with the participants and not written work samples It is against New York state law for a non teacher to have access to stud such as their names, thus I never examined student work myself, and I 5 part of the design of the experiment. While the goa l of the lessons was to inform participants about the legitimacy of African American English, the ways in which it affects their students and linguistic repertoire, y et this angle was not directly measured in any way, other than self 5 Participants did bring samples of student work during Lesson Five, however I did not se e the work or see any student names.
87 report by the participants. They did indeed report that their students became more aware of different styles of speaking and that they became more able to detect whether they had used an element of African American English or Academic English, but again, this was not measured in a systematic way. The specific goal of this research was to see whether teaching participants about African American English as a variety would affect their attit udes towards it. Lesson and Experimentation Schedule I delivered six lessons to the participants, however I interacted with them more often than this because of the psycholinguistic testing. Below is a list of the weeks, including the weeks that I conduct ed the psycholinguistic testing. Week One : Introduction, ART Numbers, Pre Assessment, Background Questionnaire Weeks Two, Three, Four : Psycholinguistic Experiment, Demographics Questionnaire, End of Survey Response, Attitude Questionnaire Week Five: Lesson One: Introduction to African American English Week Six: Lesson Two: Syntax Week Seven: Lesson Three: Syntax and Phonology Week Eight: Lesson Four: Phonology Week Nine: Lesson Five: Application of Strategies Learned Week Ten: Lesson Six: Review of le ssons, Post Assessment Questionnaire Weeks Eleven, Twelve, Thirteen: Psycholinguistic Experiment, End of Survey Response, Attitude Questionnaire Lesson One I have found from previous experience that many people initially react negatively to the idea that A frican American English is a rule governed linguistic system and not a
88 lazy, ungrammatical way of speaking. People are often somewhat slow to accept that what I am talking about or teaching is factual. Lesson One was therefore presented as exposure to Af rican American English for participants and to underscore that the material presented is academically sound and that the subsequent presentations would be based on facts. I heavily emphasized that everything I was presenting was based on well founded rese arch and that it was not my personal opinion that I was espousing. While my normal teaching style incorporates a lot of audience participation and discussion, I very specifically avoided asking the participants for their opinions about African American E nglish or Academic English during Lesson One. I did this because I believed that they likely did not have enough factual information to make an educated statement about the varieties. I did not want a negative tone to enter into the discussion if particip want any participants to make a statement that they would later regret once they had more information, which might lead them to feel embarrassed and less likely to participate in t he future. As previously described, I first gave the Pre Assessment and Background Questionnaire before teaching any content. Using a Power Point presentation I created and a general lecturing format, I gave a brief history of African American English, di scussed the achievement gap between African American and White students based on statistics and well documented research, talked about how many students are graduating (or not graduating) without control of Academic English, and I spoke about the intersect discussed research indicating that negative attitudes towards non standard language
89 varieties adversely affects student achievemen t, as well as r esearch that shows that teachers gen erally have a negative attitude towards non standard varieties as was discussed in Chapter 2. In the Power Point I used direct quotes and well as summaries of research and I included numerous citations to support my assertion that the information I was p resenting was well grounded and accepted within the academic community. I concluded and tied in the work I would be doing with the participants by discussing that if altered t hen student achievement scores should increase. An area I touched on briefly, and that I would have liked to explore in more depth had there been more time, is why Academic English became the norm in the United States. Referencing Bonfiglio (2002), I discussed the reasons why what is now considered the standa rd actually became the standard, and how this is related to the migration out of the cities caused by xenophobia during the times of great immigration into the United States. During this first lesson, in order to get participants thinking and talking about language use, I asked them to consider the different ways in which they personally use language. We spoke about the differences in how they speak amongst their colleagues, friends, family, bosses, etc, and I pointed out that the way I was speaking with t hem was am speaking with my friends, which is much faster and uses more casual vocabulary, and we discussed how that differed markedly from how I had been speaking to th em in our more formal interactions. I also modeled an exaggerated rural southern accent and asked them to compare what they thought of my intelligence based on how I was
90 speaking up to that point versus if I had been speaking in the rural southern variety the entire time. The majority of the participants agreed that I seemed more professional and smarter when speaking in my regular manner. I then asked participants to pair up and discuss what instantaneous judgments they make about people based on their u se of language, and also how this intersects with racial or intelligence judgments. My goal was for participants to see that we automatically judge people based on how they sound and the language they use, and that these judgments are not necessarily refl ective of how smart the person is and of who they are. I wanted them to see that their students may speak a non standard variety, but that this does not mean that they are not just as intelligent as a student who commands Academic English. As an example for the participants, I said that they probably would not hear This example was to hi ghlight the reality that I have found that people often show their sentiments towards and about race through talking about language instead of race White, such as Africa n American, and the commentary will without a doubt be negative. what was supposed to be a compliment about educated African Americans, is a very racialized statement. implication that the other members of the race do not and are not educated Likewise,
91 Fishman (1977) Language and identity are inextricably linked, and language and racism are equally entwined. It is nearly impossible, if not completely impossible, to discuss language in a deep and meaningful way without also touching on the topic of race. We also spoke about the fact that language is continually in flux, yet that only certain changes and additions are mar ked as negative. The example I gave is that the noun invitation has colloquially been shortened to invite and while some people find this shortening annoying or offensive, it is my impression that they represent a smaller minority to those who dislike th e shortening of the noun conversation or perhaps the verb converse to the verb conversate which is often heard in the African American shortening of invitation to invite Saturday?), yet when I have heard White speakers comment on the usage of conversate both changes to the English language, however one is laden with assumptions about the intelligence of the speaker and with racial overtones ( conversate ), whereas the other is not ( invite ). Lesson Two I have fo und over the years that when someone is unaware that African American English is a rule governed system, giving them solid syntactic examples is the quickest, most effective way to convince them that African American English is indeed rule governed, and it commences the process of generating trust and a willingness to
92 believe among participants. It opens their eyes to the possibility that what they have always heard and thought was a lazy, uneducated way of speaking might be part of a larger system of whic h they were previously unaware. Lesson Two therefore focused on syntax. I placed a h eavy emphasis on the rule governed syntactic nature of African American English and gave many e xamples that illustrated that African American English is indeed governed by grammatical patterns. provide a profound example that generally promotes others to question their assumptions yet it is not too jarrin g for there to be much resistance Once inte rlocutors understand this concept they are more open to hearing other ways in which African American English differs from Academic English During this lesson I gave participants a handout that listed examples that were presented in order for them to have a guide to follow along during the discussion as well as to give them explicit examples to use when listening to students and when examining their work for elements of African American English Grammatical features that we examined included auxiliary inv ersion, habitual be, remote past BIN, existential it, multiple negation, and several others. After presenting numerous syntactic examples of how African American English and Academic English differ, I played audio clips of actual people in a natural envir onment using examples that were similar to the ones we had just discussed and seen on the Power Point. I then previously held thoughts, and how, if at all, this had changed their thinking. We also
93 discussed how they could envision themselves applying this information with their students. sentences from African American English into Academic English, and vice versa. They were encouraged to use the handout from the beginning of class to help them, and they were allowed to work in pairs or small groups. This activity was difficult for them, and I think it underscored that there truly are major differe nces between the two varieties. Several participants commented that since this was difficult for them it must be difficult for their students to translate from African American English to Academic English for school purposes. I agreed, and I think that t heir having made this connection reinforced well as make a concerted effort to address language differences in order to ensure that their students acquire Academic En glish. At the end of the lesson I asked all of the participants to listen throughout the coming week for examples of African American English that they would have not otherwise noticed, and I opened up the next lesson by talking about what they heard and what it made them think about. Lesson Three Lesson Three focused on phonology because this is another area that can easily be shown to be very rule governed and it gives participants the chance to find rules governing the sound system. It is generally fu n, engaging, and underscores the point that use the term phonology and I defined it for them to give credence to the fac t that this is
94 definition of a new word when what I wanted them to focus on was the content of what we were discussing. During this lesson I placed a h eavy emphasis on the rule governed phonological nature of African American English and I provided e xamples showing phonological rules and t heir applications to words and categories of words For example, I showed them the following list of words (from Green, 2002) and challenged them to see a pattern within them: List Accept Wasp Build Contact Desk Spend Nobody in any of the class sections saw the pattern, so I gave them pieces of information bi t by bit. I started by pointing out that each word ended in two consonants. I then taught them about voiced versus voiceless sounds, and guided them to see that each of these words ends in two consonants with the same voicing value. Finally, I explained that when a word ends in a consonant cluster with the same voicing value, the last sound can be dropped. This example worked very well, and I believe that it was in part because the rule is so specific and there are several parameters which must be true namely that there must be two final consonants and that the voicing must be the same in order for the final sound to be devoiced. The complexity of the rule gives the participants the ability to see
95 that African American English is truly rule governed, as they had a difficult time seeing the rule on their own. That they were unable to figure out the rule on their own likely also helped them see that it is possible that even though they thought they understood everything that their students were saying, there were potentially nuances that they had been missing. For participants who already commanded African American English, they intuitively knew that the final sounds of these words could be devoiced, yet they did not know the formal rule. Learning that there was a formal rule governing when sounds can be devoiced lent weight to the notion that the way they speak is not simply haphazard and lazy, but that it is guided by rules and that it is a legitimate linguistic variety. As I did with the syntax exam ples, I also played sound clips of naturally occurring speech in which speakers dropped sounds as per the phonological rules we had discussed, plus some other examples that followed different rules that we had not reviewed. Along the same lines of the invi te/conversate example in Lesson Two, where I pronunciation, I gave an example using ask/ax and orangutan/orangu tang I asked participants to indicate by raising the ir hand whether they had ever heard someone pronounce ask as ax or if they had heard someone commenting on the fact that many African Americans pronounce ask as ax In every class section, all participants raised their hands. I then said that I was goin g to describe an animal, and that I wanted them to all say the animal that I had described in unison. This was so that everyone would be participating and that nobody would be singled out for their particular pronunciation. I described an orangutan and some participants answered with the correct animal while others guessed other types of similar animals. I listened to their pronunciation, and
96 many of them said orangu tang orangu tan I pointed out that I had heard several of them say orangu tang orangu tan Many of the participants were unaware that the word did not in fact end in /g/, which led to my explanation of how language often changes towards easier pronunciations. I described ho w the pronunciation of ask is somewhat difficult, whereas ax is a much smoother combination for the mouth to physically make. I then had the participants say each word a few times and feel what their mouths were doing. The same difficulty is true with oran gu tan where it is easier to copy the first /g/ and pronounce it again at the end of the word instead of a more unnatural /n/. At this point I asked the participants what my point was in creating this example. In each class section someone was able to ar ticulate that the ask/ax change is very marked, and linked to race, and that the orangutan/orangu tang example is the same process -the speaker making the pronunciation easier -yet it is not marked, racially or otherwise. This led to a rich discussion abou t race and language. him master reading skills. This text highlights in particular the substi tution of /v/ for // in brother and the past tense ed marker that the student was omitting such as in washed reading mistakes and dialect differences, and to be aware o f when to correct a student and when to ignore dialectal differences and focus on reading skills and making meaning of a text. I suggested that when working with a student with the goal of improving reading skills, that the teacher not correct dialect dif ferences. Instead, dialect
97 differences should be addressed and talked about openly on an ongoing basis, and corrections should be made in a formal writing assignment (not a journal entry assignment) and during a formal presentation, or whenever the subjec t of Academic or she speaks or asks a question she will begin to shut down and that this would cause much more harm than good. We again engaged in a c lass discussi on expl oring previously held thoughts and how, or if this changed their thinking, and how they could see this information applying to their interaction with their students. Although during Lesson Two participants were given a handout with syntactic examples, I did not give a handout with phonological examples during Lesson Three, and in retrospect this would p robably have been equally useful. If I were to give the lesson aga in I would provide a handout with the examples that they saw on the Power Point Lesson Four Lesson Four focused on synthesizing Lessons Two and Three, and then segued int o ways in which participants could use this information to inform their teaching. The lesson continued the explanation of syntax and phonology, and introduced the idea of discussion centered around how this new linguistic knowledge affects the participants and how they see it affecting thei r lives and the lives of their students I gave the participants some e xamples of how they could bridge the gap between African American English and Academic English with their students, and encouraged them t o think of ideas on their own. We discussed ways in which they could show their students that they value the way that they speak, but also show them that acquiring Academic
98 English is valuable and that it is worthwhile to learn to code switch between the two varieties. Bulletin board I encouraged participants to create a very visible bulletin board for students to reference. The class would create it together, and on one side they would write phrases and sentences in one variety (either African Americ an English or Academic English), and on the other side they would write the equivalent in the other variety. During Lesson Four I modeled how this would be done using the whiteboard, and as a group we identified phrases and sentences that would be good ex amples. For example, we Formal speaking event I also suggested t o participants that they have their students engage in a formal speaking event. The logic behind this assignment is that it is a specific reason to model and practice Academic English, and it is an appropriate situation in which to correct grammar and any other features that are not part of Academic English. Participants got into groups based on subject area, and they created several speaking topics that they could conceive of using with their students. The plan that I suggested would be for their studen ts to research a topic or use a topic that was currently being talked about in the classroom, and write a speech. There would then be several revisions of the writing to ensure that it was in Academic English, and then students would have time to practice reading their speech aloud to practice pronunciation features with a partner, who would listen for features of African American English and Academic English.
99 Jeopardy I also showed participants how they could play a jeopardy style translation game with th eir students. I brought in a demonstration jeopardy board that I had created and modeled how the game would be played. Each card would have a phrase, sentence, phonological pattern, or content student s would either answer the content question or they would translate the phrase or sentence into the other variety (some would be written in Academic English and some would be in African American English). During this to the next class session. We talked about what types of writing and grammatical errors their students made, and for the first time some of the participants were seeing how some of t he errors their students were making stemmed directly from the application of African American English grammatical rules instead of the application of Academic English grammatical rules in their writing. I suggested that participants focus on one element to correct at once, and not every single flaw that they found. For example, if a student is using habitual be, deleting third auxiliary inversion, I suggested that the participant chose on e area to focus on, and that it be the easiest area. In the hypothetical example I just provided, I would encourage a participant to focus on the third person plural marker. The teacher would explain the grammatical rule to the student, help the student find a few examples of that in her writing, and then encourage the student to continue editing the paper and find the
100 remaining instances. This revision process would go on until the paper was completely written in Academic English. At the end of this l esson I asked participants to pay close attention to their personal attitudes and beliefs surrounding different language varieties. Because most of the participants ride the New York City subway, they are exposed to many different types of speech on a dai ly basis. I asked them to notice these different ways of speaking and to truly try to figure out what their gut reactions and attitudes were to the speakers. I suggested that if they could become aware of their own thoughts regarding different types of s peech, it would be easier to address their feelings towards their students. At this point I reminded them of the research that shows that if a teacher holds a negative attit more poorly academically. Lesson Five Lesson Five encouraged participants to recall and synthesize the information they learned over the previous four sessions, to think critically about their personal attitudes towards African American English and Academic English, and it gave them a hands on, practical application of the material for both the benefit of their students and to help the participants absorb and retain the information. I wa African American English and Academic English and how they had changed, if at all, from the beginning of the semester, but I wanted them to have a while to think deeply about thi s before we began talking. I thus asked participants to take about ten minutes to write down how they currently felt about different language varieties, how those
101 opinions had changed since the beginning of the semester, and how they thought this did or d id not affect their teaching and their students. After about five minutes a lot of the participants seemed to have finished writing, yet I thought that if I gave them longer they might become a bit more introspective and have more to write. Many of them paused at around the five minute mark and started looking around the classroom, however the majority went back to writing and when I asked them to stop at the ten minute mark it was apparent that many of them could have gone on much longer. This produced a very rich discussion, and on the whole the participants said that they realized that they had had very negative attitudes towards African American English attitudes. Several participants also said that they thought their attitudes were changing. work, and during Lesson Five they examined this work for elements of African American English. very few to none, however for the most part the participants were able to see that phonological patt erns that we had discussed during the previous lessons. They then as effective as teaching the students the differences between the varieties and showing them how to co de switch. To discuss the intersections between race and language a bit further, I asked participants if other teachers in their schools made negative comments about their ve
102 comments, and I then asked what they did about it. This caused a silence in the room, as I expected it would. I asked the participants to get into pairs or small groups and talk about the possibility of saying something to a teacher who is equating ne gative judgments with a student who is using African American English, and I asked them to think of a few phrases that they could say, if they felt comfortable, to the teacher who was making the negative remarks. After about ten minutes we re convened as a group, and all of the groups had thought of a couple of phrases that they could say. Many participants said that they did not think they would be comfortable saying anything, however there were a few who said that they could see themselves using some o f the strategies either their group or another group had created, and most, if not all, of the participants were writing down the possible phrases that they could use to talk During the last segment of Lesson Five I divided the participants into small groups, gave them all poster board, markers, tape, and index cards, and they began creating their own Jeopardy game, as previously described, which we played during Lesson Six. Lesson Six At the beginning of Lesson Six I gave a brief overview of the main topics we had covered over the past five weeks, and I opened up the floor for discussion about how participants could see this information affecting their teaching practice. Although purpose in this research was to find ways to positively affect student achievement and close the achievement and opportunity gap, and the angle that this experiment took was
103 -in this case the -attitudes. Participants said that they found the lessons to be extremely helpful, and several from each section said that it was the most powerful and useful part of the semester. These sentiments were strongly echoed in the Post End of Response open ended questionnaire, which will be discussed in Chapter 6. We then played the Jeopardy games that the participants created during Lesson Five. There we re approximately four or five Jeopardy boards in each class section, and we played two or three, depending on how much time there was. Interestingly, and this is not something that I had anticipated, some of the answers to the questions that the participa nts created were incorrect. This tended to happen in instances when they wrote a phrase or sentence in African American English and wanted the audience to translate it into Academic English, yet the syntax was not quite right in African American English. I used this opportunity to suggest that they base their questions on actual student work so that they are sure that the language is correct if they are not native speakers of African American English. During the last fifteen minutes of the class the par ticipants took the Post Assessment, which will be described in Chapter 4, and I then thanked them for their participation and described the logistics for the Post Psycholinguistic Experiment, which would begin the following week.
104 CHA PTER 4 SOCIOLINGUISTIC RESU LTS Pre and Post Assessment Questionnaire As previously discussed in Chapter 3, the Pre A ssessment was administered at the beginn ing of the first class session after I introduced myself to the class and i n order to avoid biasing partic tionnaire the topic was a way to measure prior knowledge of the to pics we would discuss in For ease of reference, the questions are below: 1. What comes to mind when you hear the ter 2A. Are you familiar with the Oakland, California School Board Controversy of 1996? 2B. If so, please describe what you remember. 3A. What does the following mean in African American English? John be running. 3B. What does the following mean in African American English? Alexis been done ate. 4. Do you know of any ways to help children who speak African American English master Academic (Standard) English? If so, please describe. 5. Is Academic English better t han African American English? Please describe. I coded responses to the question. Favorable responses ali gned with the idea that African American English is a true linguistic system and a valid variety, in accordance with t he views held in the academic linguistic community regarding African American English. These responses indicated that the participants di d not degrade the variety itself or speakers of the
105 the academic did not indicate belief in or knowledge of African American English as a linguistic system. They often degraded the variety and/or its speakers. knowledge about African American English as a linguistic system, however in the interest of having a reasonable number of codes, this was selected as most appropriate. response. These questions were not opinion or belief based, but rather questions that had a factual b asis. The t ranslation questions ( Question 3A and 3B ) in which participants were asked to translate between African American English and Academic English wrong about an answer. Example s of favorable and unfavorable respo nses are as follows. Question One Favorable responses included: the isolation of African American slaves and their way of Unfavorable responses include d :
106 with 26% of all responses containing it.) use of plurals, and differing abilities to distinguish between first and Bias in coding was mitigated by the consideration of whether each response agreed or disagreed with the literature describing African American English as a rule governed system and a variety that is effective at commu nicating meaning. The literature disagrees with the notions that African American English is a lazy manner of speaking, indicates lack of or lower intelligence, has a faulty grammar system (it is recognized that the grammar is different than that of Acade mic English, yet that there is nothing inherently wrong with it), does not follow rules, or that it only serves as a form of slang, among many other indicators. The question of whether the responses portrayed the variety and its speakers in a positive or negative light was considered, and this Only 19% of all responses to the questions on the Pre Assessment were answered favorably/accurately. Professor participants answered 20% favorably/accurately and Professor participants answered 19% favorably/accurately, indicating that across the classes participants had the same exposure, or lack thereof, to the idea of non standard varieties and African American English itself.
107 Table 4 1 gives a pictu re of how favorably or accurately participants responded to each question individually, collapsed across professor. Table 4 1. Percentages of Pre Assessment favorable/accurate responses Question 1 2 A 2B 3 A 3B 4 5 Response rate 42% 5% 5% 11% 8% 18% 45% Questions One and Five had the highest number of favorable/accurate responses as compared to the other questions. Question One asked what came to mind when participants heard the term African American English or Ebonics. Certainly this question solicited an opinion since it asked what they specifically thought, but the intent was to ascertain whether they thought of it as a slan g, lazy form or speaking, or if they knew that it was a distinct, rule governed variety. Participants were the least kno wledgeable regarding Questions Two A and Two B and Three A and Three B. In general, participants had not heard about the 1996 Oakland Controversy and were unclear abo ut the meaning of the sentences thought remote past element. For Question Four several participants were a ble to identify useful strategies for helping students who speak African American English acquire Academic English. Favorable examples included: Unfa vorable responses for Question Five included:
108 Many of the participants responded favorably to Question Five in that they of a contradiction in the responses, however, wh en comparing some of the responses to Question Four that are directly followed by Question Fiv e For example, one participant answ ered the following to Question Four Standard English. If they have no choice they will get use to speaking Standard This is followed by Q uestion Five to which the to see that someone who believes that neither variety is better than another simultaneously believes that we should only allow students to speak Academic English and not allow space for both in the classroom. One of the goals of the lessons was to give participants methods to use with their students in order to help the students acquire code switching skills, yet this does not include for bidding students to speak African American English. Participants who were currently teachers answered the Pre Assessment questions favorably/accurately 20% of the time, while participants who were not yet teachers answered only 11% of the questions favorab ly/accurately. This di fference could be for a variety of reasons. One may be that current teachers have attended more classes from an Education department and were therefore exposed to the material. Another possible reason is that in the design of the assessment, participants
109 were not asked if they had learned about African American English in any of their previous classes. Further, it is possible that participants who are teachers have been exposed to African American English in the classroom and have learned some of the nuances of the variety from this exposure. This final possibility might be less likely, indicates a n habitual occurrence that would best be expr essed in Academic English as Additionally, it is well documented that simply by virtue of exposure one does not necessarily acquire a specific variety. Many participants who were teachers were exposed to African American English from their students, yet the participants did not acquire African American English. Likewise, many participant), yet they did not fully master Academic English either. This balance between linguistic exposure, identity, and adoption is exemplified in of the core social or linguistic group. Like their peers, they spea k African American integrated into the social life and culture of the community semicategorical rules of BEV are weakened to variable rules by the Lames; rules that are in strong use in BEV are reduced to a low level by the Lames. Whenever there is a contrast between SE and BEV, the language of the Lames is shifted dramatically
110 There are a variety of reasons why the linguistic repertoire of Lames differs from that of the members. By definition, Lames are more isolated members of the anyone can acquire the deep familiarity with local doings and the sure command of local slang that are needed to participate in verna cular culture. To be lame means to be culture they therefore do not have the opportunity to learn and participate in the vernacular, and likewise, by not knowing t he vernacular, they are marked as outsiders and unable to gain entry into the inner circle. The Lames are exposed to African American English since they go to school and live around members who possess the vernacular. They have not fully mastered it, tho ugh. They are not exposed to it to the degree that a member is exposed by virtue of the fact that they do not participate in the social activities of the members. They have not adopted the vernacular, but this is not because they are incapable of learnin g it. They may or may not have reasons for not closely associating with members, whether they are not interested in being friends with the members, or the members are not interested in being friends with them. Likewise, i n terms of linguistic exposure and adoption for the participants and their students, participants have been exposed to African American English from their students, and students have been exposed to Academic English from their teachers (participants). As with the Lames, who are exposed to the vernacular at points throughout their day yet do not fully internalize the variety, the participants and their fact that being exposed to a variety does n ot entail that the variety will become part of
111 students have been exposed to Academic English throughout their schooling they are personally, and as a racial group, at fault for not acquiring it and producing it when expected. Certainly there are numerous factors affecting whether a student will or will not acquire a school variety, ye t according to the participants many of their students have not fully acquired Academ ic English. Simply assuming that students will naturally acquire it by virtue of attending classes with Academic English speaking teachers or peers is not acknowledging the full extent of the nuances involved with the intersections between language and id entity. Interestingly, participants who indicated that they only taught Hispanic students answered favorably 23% of the time, yet participants who taught both African American and Hispanic students only answered favorably 14% of the time. As mentioned in Chapter 2 Blake & Cutler (2003) showed that teachers who did not teach African American students were more accepting of the idea of using African American English as a medium of instruction than those teachers who actually taught African American students (Blake & Cutler, 2003). The current data is consistent with their results in that participants in this study also reported more favo rable views towards a group that they did not teach. were Dominican, a s was made evident through class discussions, and participants reported that the Hispanic and African American students lived in very close quarters to each other. For example, in the high school that is directly across from the research site which is in an area in which many of the participants taught, 60% of the students were Hispanic and 40%
112 were African American. After I interacted with students at this school for other purposes, there was impressionistic evidence that both ethnicities w ere using features of African American English. This does not explain, however, the reason why participants teaching African American students would score less favorably than those only teaching Hispanic students. In terms of the gender of the participa nt, the female participants scored five percentage points more favorably/accurately than the male participants though the reason for this is unknown Table 4 2. Favorable/accurate responses by gender of participant Female Male Pre Test 21% 16% Figure 4 1 below shows l of the questions by the race of the participant. There were so few participants who described themselves as neither African American nor White that they are not included in this figure Figure 4 1. Pre Assessment favorable/correct responses by race AA, Q1, 0.4 AA, Q2a, 0 AA, Q2b, 0 AA, Q3a, 0.6 AA, Q3b, 0.2 AA, Q4, 0.2 AA, Q5, 0.4 White, Q1, 0.45 White, Q2a, 0.05 White, Q2b, 0.09 White, Q3a, 0 White, Q3b, 0.05 White, Q4, 0.23 White, Q5, 0.5 % Correct Question Number AA White
113 Perhaps the most striking difference between races in th is chart is Question 3a, which American participants answere d accurately whereas not one White participant was accurate. Also, 20% of the African Ameri can students answered Question 3b White participants answered accurately. It is i mportant to note that in order for an answer to count as accurate for Question 3b nted as accurate, both for the P re and P ost A ssessment. In retrospect, asking a question that did not require such a specific response or including more statements to create a larger number of the participants gave answers that in context might have counted as accurate as they seemed to understand the nuan ced meaning, however they were not explicit in their responses and I was very conservative in counting a response as correct Answers were only counted as accurate if they included all elements of the statement. Post Assessment Results The percentages o f favorable answers between the Pre Assessment and Post Assessment are statistically significant for every question Table 4 3. Total favorable/accurate responses Condition Percentage Pre 19% Post 70% Change 51%
114 Participants across the three classes gre w by 51%, thereby dramatically increasing their knowledge about African American English and the existence of non standard varieties. Both Professor the lessons and they increase d by roughly the same amount, with participants answering 68% favorably/accurately and answering 72% favorably /accurate ly Table 4 A B Pre 19% 20% Post 68% 72% Change 49% 52% While the same material w as taught in all three classes, there was a different dynamic in all of the classes and the discussions inherently varied In spite of the differences in discussions that occur naturally when working with separate classes, favorable/accurate responses increased similarly among classes, with both classes gaining at a significant level when the responses were submitted to a paired sample t test : Professor A t(1,20)=2.09, p<.0001 ; Professor B t(1,11)=2.20, p<. 0001 Table 4 5 reports the favorable/accurate responses broken down by question, along with the degree of the change. Table 4 5. Pre and Post Assessment favorable/accurate responses Q1 Q2 A Q2B Q3 A Q3B Q4 Q5 Pre 42% 5% 5% 11% 8% 18% 45% Post 95% 60% 50% 75% 28% 90% 90% Change 53% 55% 45% 64% 20% 72% 45%
115 Table 4 5 shows that the overall shift in positive responses is reflected to some degree in every question. This observation is supported by statistical analysis: the pre and post results for each question were submitted to a pair ed sample t test. For Question On e which asked participants about their opinions towards African American English, 42% answered favorably on the Pre Assessment, compared to 95% on the Post Assessment for a 53% change: t( 1,31)=2.04, p<.0001. Question Two A asked participants if they wer e familiar with the Oakland Controversy, whereas Question Two B asked them to describe what they remembered. For the Pre Assessment if particip ants indicated for Question Two A that knew about the Controversy, th en they also answered Question Two B accur ately, and this accounted for 5% of the participants. Interestingly, however, for the Post Assessment 60% answered that they were familiar with the Con troversy (Question Two A ), yet only 50% of the participants accurately descr ibed the Controversy (Quest ion Two B). Questions Two A and Two B indicate that they have achi eved a state of awareness about the subject, yet that not all participants accurately remembered the details of the Controversy. The 55% increase in a ccurate responses for Question Two A i s significant at t(1,31)=2.04, p<.0001, and the percentage change for a ccurate responses for Question Two B was 45%, significant at: t(1,31)=2.04, p<.0001. Questions Three A and Three B ascertained how accurately participants could interpret the meaning of two sentences written in African American English. The change in accurate responses for both questions was significant, with a 64% change for Question 3 A : t(1,31)=2.04, p<.0001, and a 20% change for Question 3 B : t(1, 31)=2.04, p<.04.
116 Question Four asked particip ants if they knew any methods that would assist students who s poke African American English master Academic English, and favorable responses increased by 72%: t( 1,31)=2.04, p<.0001. Question Five asked if Academic English is better than Afr ican American English, and there was a 45% favorable growth (meaning that at the Post Assessment they believed that Academic English was not better than African American English) which is significant at: t(1,31)=2.04, p<.001. status al so proved noteworthy. Note in Table 4 6 below that the favorable/accurate responses of participants who were currently teaching grew by 29%, whereas the participants who wer e not yet teaching grew by 60% Table 4 6. Favorable/accurate responses by tea cher status Teacher Non Teacher Pre 20% 11% Post 49% 71% Change 29% 60% Consider the following hypothesis: Many of the current teachers seemed resistant to the idea that their students spoke in a way that they did not understand; they thought that they completely understood their students and that their students were not using syntactic or phonological structures that they did not know. This perception was gained said. On the other hand, the non teachers were not yet in a position to determine whether or not they understood their students, as they had none. Perhaps those participants who had no or little prior experience with students were more willing and able to accep t that there is a way of speaking that they did not yet fully understand. F or
117 participants who already had students and truly believe d tha t they could understand them, it might b e harder to believe that they did not in fact always understand them and that there were differences to which they we re no t attuned. Perhaps there was an un willingness for participants to surrender to the idea of not fully understanding their students and to become completely open to learning about the differences in the language varieties. Favorable/Accurate Answers by Race of Students Although the participants who were teachers and who taught only Hispanic students began the semester with more favorable/accurate responses than those who taught both Hispanic and African American students, they grew almost exactly the same amount: 46% for teachers of Hispanic and African American students, and 47% for teachers of Hispanic only students Table 4 7. Favorable/accurate responses by race of students Hispanic Hispanic AA Pre 23% 14% Post 70% 60% Change 47% 46% This indicates that for the participants who were teachers, their growth was My original hypothesis, before knowing that the African American and Hispanic students used many of the same linguistic features, was that participants who taught African American students might attend more to the presentations because they might see it as more applicable to their personal teaching experience and therefore grow more As previously noted, h owever, in the example of the high school across from the university, H ispanic students were also using features
118 of African American English. Therefore it might not be surprising if all of the participants saw the information as beneficial and worth retai ning. Female V ersus Male Participants Female and male participants grew roughly the same amount as each other, with the males growing 8 percentage points more than the females. Table 4 8. Favorable/accurate responses by gender of participant Female Male Pre 21% 16% Post 69% 71% Change 47% 55% Post Assessment Responses by Race of the Participant Figure 4 2. Post Assessment favorable/correct responses by race The Post Assessment responses as viewed by race of the participant are not nearly as dramatic in difference as in the Pre Assessment Below is the chart by race and by question. There is an additional question (Question Six) which asks participants AA, Q1, 1 AA, Q2a, 0.83 AA, Q2b, 0.83 AA, Q3a, 0.83 AA, Q3b, 0.17 AA, Q4, 1 AA, Q5, 0.83 AA, Q6, 0.33 White, Q1, 0.86 White, Q2a, 0.48 White, Q2b, 0.62 White, Q3a, 0.76 White, Q3b, 0.24 White, Q4, 0.81 White, Q5, 0.9 White, Q6, 0.37 % Correct Question Number AA White
119 if they have implemented any of the suggested activities with their students. Responses indeed implemented an activity or if they had had some class discussions about language use of the type we had during the lessons. Notice that save for Question Three B, Question Five and Question Six, the African American participants answered more questions favorably/accurately than the White participants. This is in contrast to the racial divide on the Pr e Assessment which follows for ease of reference: Figure 4 3. Pre Assessment favorable/correct responses by race In Figure 4 3 the White participants answered all questions except fo r Question Three A and Question Three B more favorably/accurately than the African American participants. There is therefore a shift that occurred during the teaching sessions. The African American participants started with a lower percentage of favorable/accurate answers yet they ended the c ourse with the majority of the favorable/accurate answers, surpassed by the White participants only on Questions Three B, Five, and Six There could be many reasons for this. One possibility is that they might identify with the information more closely than the White partic ipants because it is about some AA, Q1, 0.4 AA, Q2a, 0 AA, Q2b, 0 AA, Q3a, 0.6 AA, Q3b, 0.2 AA, Q4, 0.2 AA, Q5, 0.4 White, Q1, 0.45 White, Q2a, 0.05 White, Q2b, 0.09 White, Q3a, 0 White, Q3b, 0.05 White, Q4, 0.23 White, Q5, 0.5 % Correct Question Number AA White
120 of their native language varieties and might therefore be more engaged with the material. There is a caveat to this proposition, however. While during class discussions many of the White participants held strongly to the notion that they understood ever ything that their students said and that there were no linguistic nuances of which they were unaware, at the beginning of the semester the African American participants were more likely to assert that th ey did not use features of African American English and that their students did not either. As the course progressed, however, the African American participants were more and more able to provide personal examples of phrases or sound features that followe d the rules described in class and they were able to give native speaker intuition to questions asked in class. Some African American participants also began to assert that they were bi dialectal with African American English and Academic English. It is possible that the African American students began the Pre Assessment with more of a negative attitude towards the idea of African American English, especially considering that many of them mentioned that for most of their lives they had been told (and also English. Throughout the class sessions they began to become more comfortable with the idea that African American English is indeed rule governed and that it deserves recognition as a valuable linguistic variety. If this is the case, and it did seem to be based on comments participants made during and after classes, then the African American students may have performed better on the Post Assessment than the White students because they became more invested in the information and they wanted to learn about their linguistic history and culture. Another possibility could be that the African American participants were more knowledgeable from the start, yet did not
121 display this knowledge for various reasons. Alternati vely, it is possible that the African American participants had more implicit knowledge of the linguistic forms before the teaching sessions began, and the sessions simply enabled them to articulate them more clearly. Effectiveness of Teaching Participan Switching Strategies to Use with Students When discussing the experiment with the professors and what the sessions would contribute to the course at the beginning of the semester, I was very invested in making sure that the participants left with not only a more in depth understanding about language diversity and how one variety is not inherently better than another, but also that they had specific strategies that they could immediately implement with their students in order to positively affect st udent achievement. In terms of measuring how well participants would be able to implement the suggested activities (based on the si gnificant growth with Question Four which asks participants to describe ways in which they could assist their students in ac quiring Academic English and code switching techniques), the data shows that this element was indeed successful. Most participants were able to describe in detail at least three activities to engage in with t heir students. The impact the training session s had on the participants is clear. On the Post Assessment participants were asked if they had implemented any of the activities explored in class. While only approximately 35% indicated that they had implemented an activity many said that they planned to, and several said that they had begun examining issues of linguistic diversity with their students. It is understandable that many participants e xpressed that they had not yet ha d time t o implement the activities since they answered this question only two weeks after having been
122 introduced to the activities. However, a statement made by John Baugh (199 5) is very `a propos He said, "s ome urban educators have remarked that language development is a low academic priority for poor minority students becau se of the myriad of other social problems they face in daily life. The paradox at hand lies in the fact that a good education is critical to breaking perpetual cycles of poverty, and linguistic skills are a strong indicator of the likeliho od of eventual ac ademic success (p.88). This highlights the need for linguistic intervention, arguably such as the type pres ented during the six lessons During the last lesson I again emphasized the educational disparity for minority students and the role that the tea Nonetheless, teachers indicated that they did not have time to implement the activities It would be of interest to ascertain how participants might have responded to this que stion if instead of six session s there were more, such as twenty or if the question were asked a month or so later, after the participants had had more time to digest the information and make room for it in their practice. Below are some examples of parti Six asking whether they had implemented any of the strategies: with them that it was a strange word. We appreciated its uniqueness b ut recognized its bizarre form. Of note is t hat on her demographics questionnaire this participant indicated that she taught Hispanic students so it is clear that it is not only participants with African American students who are applying this information, even though during the lessons we were ref How her students were As discussed in Lesson Three, African
123 American English has a final consonant If they were indeed dropping the final /t/, then this harkens back to the notion that there is American. That the participant explored the issue with her students is positive, as thi s was a strategy encouraged during the class sessions. Another participant who also taught Hispanic students said, I am in the process of creating a bulletin board in my room which addresses the most common uses of African American English in my classroo m. In revising their essays this week, my students looked for instances where African American English was used and underlined them. In pairs, students re wrote these sentences on loose leaf. During IR [homeroom] on Friday, students will transfer their work onto sentence strips for display on the classroom bulletin board. So far, students seem to enjoy and are ret aining the information taught. Another participant who taught both African American and Hispanic students replied I used a match up game for my science class using African American English defini Responses to Question Five which asked whether Academic English is better than African American English, on the whole received answers indicating that it was not be tter. Following are several common responses: neither variety better linguistically, but Academic English is generally is the language of Based on the anecdotal evidence above, in combination with the statistically significant increase of favor able responses between the Pre Assessment given during Lesson One and the Post Assessment given during Lesson Six it seems clear that the
124 purpose of this experiment, which was to positively African American English a nd give them specific strategies to enable their students to acquire Academic English and code switch, was successful Attitude Questionnaire The participants completed the Attitude Questionnaire directly following the psycholinguistic element of the experiment both after the Pre and Post Psycholinguistic Experiment. I will refer to it as the Pre Attitude Questionnaire and the Post Attitude Questionnaire, however it was t he exact same questionnaire Note that never attitude After the y stepped out of the experimentation room they sat privately, filled out the Questionnaire, and put it into a was created with the intent of giving the participants a sense of anonymity aside fro m their identification number in that they were no t giving me, the researcher, the Questionnaire, but rather putting it in an envelope themselves. Of course the question of whether participants were responding to the questionnaire in a positive manner in attempts to please me, the researcher, cannot be known. This Hawthorne Effect plagues any experiment in which the participants know the researcher will read their responses, and certainly when the researcher is physically in the room or near the participants. Fortunately, the psycholinguistic element o f this experiment, which will be discussed in Chapter 5, included methodology to help avoid or mitigate the Hawthorne Effect. Having the participants complete the Attitude Questionnaire privately and put it in the manila folders was an attempt to help mit igate this possible effect. Ultimately, whether or not participants were affected by their
125 interactions with me as the researcher, they completed the Pre and Post Attitude Questionnaire in the same environment both times, which supports the notion that t he results are a valid measurement. As previously discussed, the Attitude Q uestionnaire was taken directly from twenty five questions, twelve rticipan ts responded to questions on a five point likert scale. The full questionnaire is in Appendix C I coded the responses to the Attitude Questionnaire as follows: Because twelve of the questions were phrased positively and thirteen of the questio ns were phrased negatively, the responses were coded on a five point scale of most positive to least value of five points, neutral responses were assigned three points, and the least favorable responses were assigned one point. Paired sample t tests were calculated based on raw numbers for each question, and averages were calculated and submitted to t te sts for comparing groups of responses such as totals by race, gender, professor, etc. Only participants who completed both the Pre Attitude Questionnaire and the Post Attitude Questionnaire were included in the results. Table 4 9 displays the means of the responses to the Attitude Q uestionnaire, broken down by individual question. Notice that the favorable responses increased for
126 every question, with the lowest in creasing by 4.6% and the highest by 56.6%. The six lessons previously discussed were not cre ated with the specific intent of addressing each of the questions on the Attitude Questionnaire C ertainly many of the concepts on the Questionnaire are in line with material that I felt was important to address in terms of affecting attitudes and introdu cing participants to strategies to assist their students in acquiring Academic English. At no time, however, were the questions in the Questionnaire systematically incorporated into the creation of the lessons or the guidance of the class discussions. Table 4 9. Means of Pre and Post Attitude Questionnaire responses Pre Average Post Average Increase % Increase Q1 2.98 3.84 0.87 29.10% Q2 3.2 4.16 0.96 29.90% Q3 3.18 4.11 0.93 29.40% Q4 3.47 3.96 0.49 14.10% Q5 2.2 2.53 0.33 15.20% Q6 2.89 3.31 0.42 14.60% Q7 3.33 3.93 0.6 18.00% Q8 2.62 3.56 0.93 35.60% Q9 3.58 4 0.42 11.80% Q10 2.6 3.02 0.42 16.20% Q11 3.84 4.29 0.44 11.60% Q12 2.76 3.07 0.31 11.30% Q13 2.84 3.4 0.56 19.50% Q14 3.18 3.71 0.53 16.80% Q15 2.51 3.93 1.42 56.60% Q16 3.71 4.42 0.71 19.20% Q17 2.04 2.49 0.44 21.70% Q18 2.44 2.67 0.22 9.10% Q19 2.64 3.44 0.8 30.30% Q20 3.87 4.2 0.33 8.60% Q21 3.27 3.76 0.49 15.00% Q22 3.62 4.04 0.42 11.70% Q23 3.47 4.04 0.58 16.70% Q24 2.89 3.02 0.13 4.60% Q25 3.73 4.02 0.29 7.80%
127 This lack of intentionally growth, speaks to the overall scope and nature of the since the change was not only in direct relation to the specific points addressed i n class. The Q as a nonstandard form of English. It is unknown whether during the Pre Attitude Question naire participants thought of African American English and Nonstandard English as the same constructs, yet by the Post Attitude Questionnaire participants had been told that African American English is one specific nonstandard variety. In his analys is, T aylor (1973) divides the twenty five questions into four main content categories, which are as follows: 1. The structure and inherent usefulness of nonstandard and Black English dialects 2. Consequences of using and accepting nonstandard and Black English in th e educational setting 3. Philosophies concerning the use and acceptance of non standard and Black English dialects 4. Cognitive and intellectual abilities of speakers of nonstandard and Black English The results for the current experiment will be addressed w ithin the content categories created by Taylor. Taylor identifies how many questions belong to each category, however he does not itemize which questions belong to which category. His breakdown of the number of questions per category is as follows: Conte nt category (1): 8 questions Content category (2): 8 questions Content category (3): 8 questions Content category (4): 1 question
128 Although Taylor did not itemize which questions belonged to which content category, I did, and based on the numbers it is cle ar that he and I are not categorizing every question identically since in my analysis the numbe rs align identically only with Content Category Three My itemization of content categories is as follows : Content category One : 7 questions [Numbers: 2, 5, 9, 12, 14, 15, 23] Content category Two : 10 questions [Numbers: 1, 3, 6, 13, 16, 17, 18, 21, 24, 25] Content category Three : 8 questions [Numbers: 4, 7, 8, 10, 11, 19, 20, 22] Content category Four : 0 questions Content Category 1: The Structure and Inherent Usefulness of Nonstandard and Black English D ialects The paired sample t test comparing whether participants believed that African American English was a misuse of Standard English or not [Q2] indicates that by the Post Assessment they did not believe that it is a misuse of Standard English, and this result is significant t(1,36)=2.03, p<.0001 This is a topic that was discussed during every lesson. Participants also did not believe that African American English was an inferior language system [Q9] t(1,36 )=2.03, p<.04, and while this is significant, it is not as highly significant as some other findings. The idea that African American English is not an inferior language system was also discussed during every session, however the results might indicate tha t they held more strongly to the notion of Academic English instead o because the semantic load of the terminology is reinforces what they believe to be true about the variety. I told them, however, t hat it
129 was their personal choice as to what terminology to use, though I underscored why I would be using the terminology that I had chosen. English English reinforcing the terminology that I believed was most appropriate. At the beginning of the semester during class discussions there were several participants who used the term English cade began u right away there was one participant who continued It is possible that other participants in this class session were affected by this usage, however this would only affect one third of the participant group. American English sounds versus the grammatical system under which African American English operates. Qu estion 5, which asked participants whether they believed that African American English sound ed as good as Standard English was not significant t(1,36)=2.03, p>.05, and while Question 12, which asked participants if they believed that African American Engli sh sounded sloppy is significant t(1,36)=2.03, p<.05, indicating that they did not believe that it sounded sloppy, the results are less significant than many others. It seems that participants were more willing to change their attitudes towards the idea o f African American English as a valid grammatical system as opposed to a system that sounds acceptable. Although there was one class session exploring the phonology of African American English, greater emphasis was placed on the rule governed nature of th e system in terms of syntax. This may be why there is a
130 difference in positive response changes. Alternatively, perhaps participants were more willing to accept syntactic differences instead of phonological differences. There might be a bigger hurdle to overcome in terms of phonology as many participants have heard African American English for years, yet have never (or rarely) been exposed to it in a written medium. Interestingly, Lisa Delpit, an African American scholar who believes in valuing African American English while still ensuring that students have access to Academic English, does not believe that African American English sounds as good as Academic English. She gives the following anecdote about her Academic English speaking daughter who had transferred to a charter school with a predominately African American population: When I heard these words spoken by my eleven year old daughter it seemed as though a hundred conflicting scripts raced through my mind at (Delpit, 2008, pp. 33 34) On one hand Delpit believes that African American English should be valued as a valid linguistic variety, and on the other hand see sees the value in acquiring and using Academic English for social mobility and power purposes. That even she is struck by the sound of African American Engl ish from her daughter indicates how insidious the The final two questions in this content category addressed whether participants believed that African American English is a clear, thoughtful, and expressive language [Q14], and whether nonstandard English is as effective for communication as Standard
131 Engl these questions, respectively: [Q14] t(1,36)=2.03, p<.03; [Q23] t(1, 35)=2.03, p<.03. Conte nt Category 2: Consequences of U sin g and Accepting N onsta ndard and Black Eng lish in the Educational S etting Participants did not believe that the scholastic level of a school would fall 6 if teachers allowed African American English to be spoken [Q1] t(1,36)=2.03, p<.0002, and likewise they did not believe that acceptance of nonst andard dialects of English by teachers would lead to a lowering of standards in schools [Q21] t(1,35)=2.03, p<.03. During the lessons I presented the case that students speak a variety that is different fro m the academic, school variety. I said that the variety that the students speak is integral to the ir personal identity of self, and that is should therefore not be taken away or erased. I suggested bring to school while simultaneously ad dressing the ir need to acquire Academic English discussions, however it is clear from the positive change that participants extrapolated the information to suggest that sta ndards would not be lowered if African American English were accepted. Over the course of the sessions participants grew to believe that teachers should allow African American students to use African American English in the classroom [Q6] t(1,36)=2.03, p <.03. This point was discussed in terms of allowing students to express themselves comfortably and most importantly that participants should not correct students while they are making an academic point, but rather when they are specifically 6
132 focusin g on l anguage systems or grammatical points. Interestingly, they also believed that speakers of African American English would be more motivated to achieve academically if the use of African American English were encouraged [Q13] t(1,36)=2.03, p<.03. Class dis cussions touched upon the idea that achievement would increase if students were made to feel like they, and their language, were accepted explicitly encouraging it. Pa rticipants found that attempts to eliminate African American English from schools would result in a situation that could be psychologically damaging to African ever employed in class, the idea of the interconnectedness of language and identity was a recurring theme. They also believed that when teachers reject the native language of a students, they do him/her great harm [Q16] t(1, 36)=2.03, p<.0001. This conce pt was addressed specifically and was integrated frequently into class discussions in terms of harm to academic success. It is directly tied to research as discussed in Chapter 2 L1 perform more poorly, and this point was emphasized throughout th e lessons with the their Question 17 asked participants if they thought a teacher should correct a of growth needs to be cla rified. T he significant growth indicates that participants grew in of nonstandard
133 English, and that by the Post Attitude Q uestionnaire they believed that teachers should This question is ambiguous in terms of how and when a teacher should correct. A grounding principa l of the study was to impart that teachers should give stude nts access to Academic English but that they should do so in a way that values the variety the students already possess, and that it should be a linguistically additive as opposed to a subtractive process. When scoring this question, my underlying assumption was that correcting was negative because there is a certain type of correction wherein the and p articipants described correcting their students in this manner. It is unknown, however, how participants interpreted this question. During the discussions one recurrent theme was the necessity of highlighting differences between the two varieties and ensu ring that students acquired Academic English. To a certain degree this requires correction, yet it is not the type of correction we might envision where a in a way that devalues the variety the student employed. It is hard to because the type of correction a tea cher is said to give is unknown By the end of the course participants did not believe that in a predominantly African American school, African A merican English as well as Standard English should participants had to originally believe that teaching African American English was negative, and that by the Post Asses sment teaching African American English was positive. During the class instruction participants were told that the goal was to enable
134 students to code switch, but that this was not to be accomplished by teaching in African American English or that African American English should be taught as a specific variety. In fact, the point was made that during the 1996 Oakland Controversy it was a misconception that teachers would teach students how to speak African American English because in reality the students already possessed the se linguistic skills. It notion that they should be teaching students in African American English, as this is not widely supported in the lite rature. Additionally, a topic that was not specifically addressed and that the literature would not support was whether one of the goals of the American school system should be the standardization of the Eng lish language [Q24], and participants did not gr ow to believe that this should be the case t(1,34)=2.03, p>.05 The final question in this content category asked participants if one successful method for improving the learning capacity of speakers of African American English would be to replace their dialect with Standard English. Participants grew significantly to believe that this would not be a successful method [Q25] t(1,34)=2.03, p<.03. Discussions revolved around the benefit students would receive if they acquired Academic English, yet this was presented as goal of adding Academic English into their repertoire and not of subtracting their competency with African American English. Content Category 3: Philosophies Concerning the Use and A cceptance of Non Standard and Black English D ialects The q uestions in Content Category Three are divided evenly into two categories: those that portray African American English in a positive light, and those that portray it in a negative light. Of the positively charged questions, p articipants responded extremel y positively to the question asking if African American English must be
135 accepted if pride is to develop among African American people [Q8] t(1,36)=2.03, validating a wa y of speaking that was previously thought of negatively bled into the idea of a sense of pride for the speakers. Participants also believed that widespread acceptance of African American English is imperative [Q19] t(1,35)=2.03, p<.0003, and that Nonstand ard English should be accepted socially [Q22] t(1,35)=2.03, p<.03. These points were not made directly, and instead discussions touched on acceptance within the schools. A topic that was never discussed in class was whether African American grew to believe that it is [Q10] t(1,36)=2.03, p<.04. Of the negatively charged questions in Content Category Three Question 7 changed the most significantly, indicating that participants did not believe that African American English should be discouraged t(1, 36)=2.03, p<.0001. It is interesting to compare the responses from Question 6 and Question 7. Question 6 is Content Cat egory Two and asks participants whether teachers should allow African American students to use African American English in the classroom. Both questions were significant ([Q6] t(1,36)=2.03, p<.03), but Question 7 changed more than Question 6. For Question 6 participants were asserting that it is acceptable for students to use African American English in the classroom, whereas in Question 7 they were indicating that use of African American English should not be discouraged. It is interesting that the change in Question 7 was more significant, and this could possibly suggest that the participan ts grew to believe that discouraging the use of African American English in general is not necessary, yet they are not quite as comfortable with the idea of using
136 African American English in the school classroom. It would be interesting to see if the resp onses to Question 6 would change after more sessions or after more time spent exploring the issue with their students in the classroom. Participants dis continued usage of a nonstandard dialect of English would accomplish n [Q4] t(1,36)=2.03, p<.03, and that the sooner we eliminate nonstandard dialects of English, the better [Q20] t(1,35)=2.03, p<.03. Class discussions revolved around the principle of acceptance of African American English as a that sparked the belief that nonstandard varieties should not be eliminated. Finally, participants did not believe at the end of the lessons that African American English should be considered a bad influence on American culture and civilization [Q11] t(1, 36)=2.03, p<.03. Growth by p rofessor nses grew significantly, independent ly cla ss they attended. Professor participants grew: t(1,22)=2.04, p<.0001, whereas Professor participants grew: t(1,13)=2.17, p<.03, indicating more positive growth for Professor participants. h owever Professor A did have two class sections so there were more participants in this analysis. However, for the purpose of the study, it is positive that all classes grew at significant levels, indicating that even though material was presented somewhat differently in each class due to the nature of different participants and slightly different discussions, the information I wanted to convey was consistent across class sections.
137 Growth by g ender Female and male participants grew very similarily with fe male participants growing significantly at: t(1,20)=2.04, p<.0001 and male participants growing significantly at: t(1,15)=2.14, p<.0002. Growth by r ace While in this analysis there were 21 White participants, there were only 5 African American participan ts, and the growth for the African American participant pool is not comparable because of the small sample size 7 Of the 5 participants in the sample, the growth was t(1,5)=2.57, p<.0002. Again, however, this sample size is too small to draw well grounde d conclusions. Growth for the White group, however, is the most significant of all growth reported in this chapter. White participants grew t(1,21)=2.08, p<.0001, where the actual p value is 4.39E 18, or p<.00000000000000000439. Due to the small sample size of African American participants, an ad hoc comparison was run thus enabling t he comparison of growth for non White participants. This group of participants grew significantly at t(1,12)=2.18, p<.004. While this group grew at significant levels, the growth is less dramatic than that of the White participants. Why the White partic ipants grew more is unknown. One possibility might be that the information was newer to them, thus they had more knowledge to gain. This is not necessarily the case, though, because while many of the African American and non 7 These numbers are smaller than those in Table 3 1 because the paired sample t t ests that were run for the Attitude Questionnaire only included those participants who completed both the Pre and Post Attitude Questionnaire.
138 White participants were speak ers of African American English, on the whole they believed that it was an inferior, lazy system, and their negative attitudes could be regarded as equally uninformed as the White participants attitudes. On the other hand, it might be that since the non White participants had been told for so many years that the way they spoke was incorrect, lazy, or sloppy, they might need much more time to accept the fact that the academic community regards it as a legitimate, acceptable variety with distinct rules and patterns. Overall, the Attitude Questionnaire indicates great positive attitudinal change among the participants between the beginning and end of the lessons. I hypothesized that participants would grow positively for most questions, and I was very pleased that favorable responses increased for every single question, with the majority of the questions increasing at a statistically significant rate. End of Experiment and Class Sessions Survey After completing the Pre and Post Psycholinguistic Experiment, pa rticipants e Attitude Questionnaire, and/or the class lessons (after the Post Psycholinguistic Experiment) in an open ended, free flowing way. They completed these forms directly following the Psycholinguistic Experiment and put them in a manila envelope just as was described for the Attitude Questionnaire. Once they exited the experimentation room the order of forms was Demographics Questionnaire (only after the Pre Psycholinguistic Experiment), Attitude Questionnaire, and End of Task Response forms. After the Pre End of
139 Task Response Psycholinguistic Experiment that form was labeled of Experiment/Class Sessions Resp o you have any Responses from the first session will be referred to as the Pre Response, and responses to the final session will be referred to as the Post Response. The comments that specifically addressed the Psycholinguistic Experiment will be addressed following those results in C hapter 6 I had a variety of hypotheses regarding how participants might initially r eact to the experiment and class sessions. In my experience with speaking about the topic of African American English with friends, family, and other acquaintances, I have found often negative or disbelieving After fu rther explaining the topic and answering questions, however, the conversation s tend to become more positive. Nonetheless, there have been situations in which interlocutors have become somewhat angry and hostile about the idea of African American being a v eritable linguistic variety and in these conversations the other person tends to have been an English major in college or someone who considers her/himself an expert at language Because of these interactions, I was unsure of how participants wou ld react to the experiment, especially considering that many of them were English teachers. I was very aware that the Attitude Questionnaire and Psycholinguistic Experiment might seem racist to participants since they would not know the context of what was to com e in the lessons and that this could affect the beginning of the class sessions if participants
140 a lready thought that I was unaware of the interactions between social class, language, and other factors affecting how one perceives another speaker. Pre Experiment Responses The responses to the Pre and Post Response varied greatly between each other One particularly positive response to the Pre Response was as follows: Makes me think more about language and the presumptions I make based on what I hear. I should assume less, as I am co nstantly misjudged. Thank you. On the Pre Response another thus indicating confu sion over terminology. Two participants stated that the statement underscores the idea th at participants were unfamiliar with African American English as a valid linguistic construct before the class sessions. Of great note, however, is this same pa Res ponse. On the Post gain, racist. Non discussed with participants, the term o means a perfect label, but I chose to use it as it is currently the most popular term in linguistic writing. That the participant changed from standard though, indicates a great shift in thinking and a deeper understan ding of the construct.
141 Post Experiment Responses Not all participants responded to the Post Response, but most did, and responses focused mainly on the class sessions and Psycholinguistic Experiment. Participants had overwhelmingly favorable reactions t o the lessons Comments generally indicated that participants found the lessons interesting, informative, and useful. Most of the responses included sentiments such as the following: am so grateful to have learned how to bridg e African American English and Academic English class to a speaking African American English. Now I understand what students are trying to say, and how to address them switching to Standard English in a constructive a great experiment, an eye opener for me. The system should give it a Development] for all teachers so at least they can teach students about their heritage and where t The following response was written by an African American participant and is a reflection of his experience with the course: As an African American I felt that this class was a refreshing experience to know that how my peers, family, and I manipulate the English language is well studied and has meaning to it. I believe that classes like these will help many African Americans to code switch easily and be able to communicate in both varieties. I feel tha t the experiment was a good way to do some introspection as to how I perceive how speakers of different language and different varieties speak the English language. I hope that this becomes a requirement for inner city teacher esp. English teachers.
142 Some speak Academic English, and below is one example of this type of comment. Note, however, that class discussions specifically addressed the need for students to acquire Academic En glish for academic pursuit and social mobility, and what the participan t was regularly addressed and is in line with the lessons: I thought the experiment was helpful and I learned a lot about the structur e of African American English I still believe that in an academic setting, the and enforced, however this does not mean that African American English is not accepted. To prepare for the future success of our students, we need to give our students the highest academic standard, what educators would expect to see at each step of the education process and th I examined r esponses to both the Pre and Post Response in several ways to identify any trends by race. For both r esponses I considered whether African American and non White participants were more likely to not answer and only minimal diffe rences were noted. For the Post Response I was interested in whether more African Americ an and non White participants responded more positively or negatively to the sessions, and no difference was noted. Finally, the responses were broken down by identifying whether the African American and non White participants were more likely to insist t hat Academic English be taught, and this was not the case. Overall it seems that reported experiences did not vary by race in any discernable ways. Sociolinguistic Conclusion s ged in signi ficant ways over the course of the six lessons. During the first lesson participants were unfamiliar with African American English as a legitimate linguistic varie ty and instead thought that it was a slang, lazy way of speaking. On the whole t hey were unable to
143 translate between African American English and Academic English, and they were not aware of specific strategies to help mono dialectal speakers of African American English acquire Academic English. Responses were generally negative to questions measuring how positively participants perceived African American English as a construct, as well as its worth in the educational system and in society as a whole. At the end of the lessons, however, in terms They stated that the lessons were useful and that they valued the ways in which the lessons had already informed their teaching practice and the ways that their practice would be informed in the future once they implemented the suggested strategies Several participants suggested that the type of lessons I taught be given to all educators as professional development and this se ntiment was expressed to me as well as the university class professor s and the chairperson of the department at the university.
144 CHAPTER 5 PSYCHOLINGUISTIC METHODOLOGY Experiment: Reaction Time Task The Psycholinguistic Experiment measure d reaction times and provided insight into their implicit processing through their responses toward s African American English and Academic English. In this task, participants listened to four stories, two of which were spoken in African American English and two of which were spoken in Academic English. Participants were then were presented with a series of opposing adjective pairs and were asked to select the adjective that best described the speaker of the story. The experiment used a matched guise methodology in which across the four stories there were only two speakers: one African American woman and milliseconds, and longer response times are understood to be an indication that speake rs had more difficulty making that adjective selection, while faster response times indicate that participants had an easier time selecting that adjective. Thus, the faster a participant selected an adjective, the easier is was for the participant to comm it to her or his choice. attitudes towards different speech varieties. I will define the implicit attitudes t hat I am attempting to elicit in three ways: (1) those that the p articipants are potentially aware of but might not want to display on a self report measure, (2) those tha t they are unaware of but that still exist and might or might not show up on a self report measure, (3) or those that they may try to control even und er the quick decision task they perform but that will surface when I examine reaction times.
145 Hawthorne Effect and Affiliative Social Tuning Hypothesis Participants might feel a certain degree of so cial pressure from the experiment and from having an ex pe rimenter near them and they may therefore have attempt ed to answer in a more socially acceptable way as with the Hawthorne Effect or the Affiliative Social Tuning Hypothesis, as will be discussed. Nonetheless, I was expecting their reaction times to be slower in conditions when they would personally be biased. As mentioned previously in Chapter 2 research shows that on explicit measures White participants often respond in more socially acceptable ways when in the presence of a Black experimenter (Lower y et al. 2001), and in situations in which they feel that the person with whom they are interacting holds the same social opinions that they do they tend to respond more towards the way they think their partner feels (Higgin s & Rholes, 1978). I am White yet how they perceive d my social attitudes is unknown. I hypothesize d, though, that since we were in a university environment and since I was conducting a study on attitudes, th ey would have a sense that I was sensitive towards racial attitudes and they may have attempt ed to alter their response s in favor of what they perceive d me to believe. The affiliative social tuning hypothesis, as defined by their beliefs to the ostensible beliefs of another social actor when they desire to get along with this ther or not the participants were something I knew or e xplicitly measured in this s observer is certainly a possibility. As will be described, I was in the room while the participants completed both the Pre and Post Test portions of
146 responses were time locked and the change between the timing of their responses on the Pre and Post Test was measured. Since I was in the experimentation room both times, I believe the growth that will be reported in Chapter 6 to be a true measure of the change in their attitudes. Additionally, the participants did not know that their responses were being measured by time as well as by adjective selection, which means that while they could have adjuste d their responses in a direction to try to please me in terms of adjective selection, they showed their implicit attitudes through the speed of their responses. Timing Responses As discussed in Chapter 2, the timing element of the experiment was based o n principles from Gawronski et al. (1998). When faced with a decision, Gawronski et al. (1998) found that easier decisions require less cognitive processing and are mirrored by faster button response times, whereas more difficult decisions that require mo re processing time are reflected in slower button response times. The current research took advantage of these findings, and participants responded on a response box that timed their responses to the millisecond, which allows insights into which decisions were easier to make and which were more difficult. Method Participants Forty nine (twenty eight female and twenty one male) participants voluntarily participated in the experiment including the two professors, whose data were omitted from the results
147 Design The experiment consisted of four stories told by two female speakers and eleven adjective pairs that participants assigned to the speakers of each of the stories. Both speakers were women; one was African American and the other was White. Each of them told two stories: one in African American English and one in Academic English. The four stories were different from each other, and all participants heard all stories. Once the experiment began the participants heard one story and then saw a pair of opposing adjectives on the screen. The participants selected which adjective best described the speaker they just heard, for a total of eleven sets of adjective selections. The participants then heard the following story and assigned the adjectives to t hat speaker, and so on, for four stories. Materials Stories When designing this experiment my original intent was to use naturally collected speech samples. This proved very difficult, however, as it was not possible to control for as many factors as I deemed necessary. I could not find a public location where naturally occurring s peech could be recorded without too much background noise, and speech samples from television, radio, and YouTube that were actually spoken in African American English contain ed either inappropriate content or were otherwise unusable. I therefore chose to create speech samples for the purpose of the experiment. I wrote f our stories that were then spoken aloud by carefully selected speakers and the speaker selection process will be discussed momentarily Two of the stories were in African American English and tw o were in Academic English, and I solicited
148 feedback on the stories from native speakers of each variety. Before writing the stories I made a list of the African Ame rican English features that I wanted to include, such as final consonant cluster devoicing, lexical features characteristic of African American English, dropping of the third person singular marker, etc. I then wrote two stories in African American Englis h, with each story containing the same number of syntactic, phonological, or lexical features of African American English. In contrast, the Academic English stories had the same number of instances where African American English features could occur but di d not. The four stories were within seconds of each other in terms of length, and the sentence structure was similar. Each story indicated at the beginning that the speaker was a student, and the content of the stories was neutral in nature. The stories were delivered using the matched guise methodology. In this methodology one speaker speaks in two di fferent ways or guises, and participants are ostensibly unaware that they have heard two different samples of speech from the same person. This a attitudes to wards different language varieties that they may otherwise not express to others. Table 5.1 below indicates the configuration of the four speech samples, or stories. In order to aid the reader in remembering wh ich story was spoken by which AAE) was from the African American speaker speaking African American English, Story Two (AA AE) is from the same African American speaker speaking Academic English, Story
149 Three (W AAE) is from the White speaker speaking in African Americ an English, and Story Four (W AE) is from the same White speaker speaking in Academic English. Table 5 1. Story, speaker, and variety Story Speaker Variety One African American African American English Two African American Academic English Three White African American English Four White Academic English Each of the speakers came to my apartment in New York, separately, to record the stories. This location was chosen because it was quiet and there were not external factors creating ambient noise. Before the speakers arrived I instructed them to practice reading the stories aloud so that they would sound as nat ural as possible while still adhering to the script since it was carefully constructed. The speech samples were recorded using a Rola nd Edirol R 09 24 bit/48kHz recorde r, and the speech files were The written stories are below: Story One (AA AAE) African American Speaker: African American English subway place him. I saw some other people staring hard at him too and finally I realized that son play football. I got home and told my sistah and she was everywhere this star and that star and I be thinking you just wanna see a celebrity! You Story Two (AA AE) African American Speaker: Academic English
150 I really like watching tv but usually I only have time for the news -I have to really vivid. I dreamt that I was a news reporter but that I was also one of the people in the story! I saw a fire across the street and called 911. should go to journalism school to see if you really li ke it. You spend so Story Thre e (W AAE) White Speaker: African American English she ax me do My birfday was on Friday and she brought me a card and a book on how the Brooklyn Bridge was b minor. My bruvah study History too and he did lik e it a lot. That would be Story Four (W AE) White Speaker: Academic English My favorite hobby is definitely photography. I took it as an elective a few years ago and I really loved it! I like taking all kinds of pictures but my favorite is going to famous buildings like the Flatiron or the Chrysler building. I often go on weekdays around mid afternoon when the streets are less crowded. My sister asked me if I could take pictures of her and her famil y and so one afternoon we went all across the city. When we got there et some of my photos into a local art would be really nice if I could get some of my pictures into the show and then maybe start a career.
151 Speakers Speakers were located onlin e using Craigslist. Craigslist is an online forum where people can post items for sale, community events, and services wanted, among many other categories and both posting and replying to posts is free. One can post an ad for a specific item or service and people viewing the website can respond by sending an email through the website. I posted advertisements for this experi category with the hopes of finding speakers who were more linguistically aware and who might be more adept at s witching between the two varieties when asked. I was looking for an African American woman who could speak both African American English and Academic English, and a White woman who could also speak both varieties. Two different advertisements were poste d: one for an African American woman between the ages of twenty to thirty who was fluent in both African American English and Academic English and one for a White woman of the same age group who was also fluen t in the two varieties The ad vertisement indi cated that they would meet with me in person and read two short stories that they had previously rehearsed in order for the stories to sound as natural as possible, and that they would be paid $50 for their time. Once someone responded to the ad vertisemen t I sent an email asking the interested speaker to answer the following questions: How old are you? What is your highest level of education? Where are your parents from? Where are you from? What was your average family income growing up? These questions were asked in order to choose an African American woman and a White woman who were as similar to each other as possible in order to help control
152 for linguistic and socioeconomic differences. Finding speakers was somewhat difficult One o f the reasons is that I asked the potential speakers to perform a very difficult task. If they responded to the questions I then sent them an email to schedule a time to speak on the phone. On the phone I told them what the experiment was about and what their role would be. When they spoke on the phone they used Academic English so it was apparent that they controlled this variety. I then asked them to perform a very difficult task, which was to speak for a moment, extemporaneously, in African American English It is n ot easy to perform this task, especially considering that I only control Academic English and could therefore not hold the other end of the conversation. Nonetheless, I wanted to hear if they understood what African American English was a nd if they could produce it on command since they would be asked to speak in it for the speech samples. When the African American women attempted to speak in African American English, however, the content was uniformly about drugs and violence. They were not using features of African American English and instead were relying on speaking loudly and on content that they thought would fit the situation. One female speaker was found wh o was able to speak for about thirty seconds while using features of African American English with neutral content and she is the African American speaker that I selected In retrospect I think it would have been just as effective to email the candidates a script already written in African American English and have them r ead it over the phone for the interview since the actual task was not requiring them to produce African American English on their own. My intent, however, was to find a native speaker and not someone who could simply read a script, and ultimately the meth od I used was successful.
153 Finding a White speaker proved significantly more difficult than finding an African American speaker. There were very few responses to the ad vertisement s posted on Craigslist and it seemed that there were not many White women w ho controlled both varieties. Additionally Craigslist has a feature whereby anyone on the n advertisement, which means that it gets deleted, and this accounted for the other major difficulty The ad vertisement s for a White speaker who was fluent in African American English and Academic English were repeatedly flagged, and therefore deleted, on a regular basis. For several days I reposted the ad vertisement written in various ways in hopes of avoiding flagging ( several times an hour in many cases ) trying to get one to stay on the site long enough for someone to respond to it. The fact that the ad vertisement kept getting deleted will be addressed in Chapter 7. Below is detailed information about each of the speakers I selected and why I believed them to be similar enough to be appropriate for the experiment. The African Age: 26 Highest level of education attained: BA in Communication and Art Nigerian Where she grew up: Queens Average household income growing up: $100,000+ Age: 33 Highest level of education attained: Masters Where she grew up: Brooklyn Average household income gr owing up: $70,000 Although the two speakers grew up in different areas of New York City, I judged their accents to be very similar. While the African
154 from Nigeria, the speaker herself was born and raised in Queens New Yor k and showed no trace of a foreign accent. I judged these two speakers to be similar enough to use for the experiment because they were both from New York City, they were each able to produce African American English and Academic English, neither of them had especially strong borough accents, they were within a few years of each other in age, and they both came from upper middle to upper class families socioeconomically. Adjectives The experiment used an adjective discrimination task to assess implicit att itudes towards African American English and Academic English. Participants responded to eleven sets of adj ective pairs that were opposite in nature. The adjectives are listed below in Table 5.2. Table 5 2. Positive, negative, and neutral adjectives Posi tive Negative Neutral In t elligent Unintelligent Talkative Hard working Lazy Not talkative Rich Poor Upper class Lower class Friendly Unfriendly Uneducated Educated Low achiever High a chiever Irresponsible Responsible Loud Quiet Unsuccessful Successful The neutral set of adjectives was Talkative and Not Talkative, and these were not including in the analyses. They were included in the experiment because I had intended them to be part of the positive/negative set, however after further thought I decided that there was not a clear positive or negative attribute to either of them. The adjective pairs were taken from Gawronski et al. (2008) and Rudman et al. (2001), with the exception of Loud/Quiet, Low achiever/High achiever, and Talkative/Not
155 talkative, which were created expressly for this experiment. The length of most of the pai rs intentionally contained approximately the same number of letters as each other, and aside from the neutral pair (talkative/not were both either one or two words. There were two forms of the adjective lists: one with the positive adjectives on the left side and the negative adjectives on the right side and one with the positive adjectives on the right side and the negative adjectives on the left side This was to avoid right left bias, whereby the participants wo uld select the adjective more quickly when it was presented on the side of their dominant hand. There were therefore eight lists of stories: 1a, 1b, 2a, 2b, 3a, 3b, 4a, and 4b. Participants were randomly assigned the story they heard based on the order in which they came into the experimentation room. The list was ordered 1a, 2a, 3a, 4a, 1b, 2b, 3b, 4b, etc. The first participant heard 1a, the secon d heard 2a, and so on. Procedure Logistics In order to run the Pre and Post Assessments for all of the participants as well as teach six lessons within the semester, it was important that the experiment be run efficiently, with little time wasted bet ween when participants arrived for the test. To facilitate efficient movement between the classroom and the experimentation office and so that participants would know when it was their turn to leave the classroom without their having to decide who would leave next and disrupt the class, I created a numbering system that I randomly distributed to ten participants at the beginning of each of the testing days. The system consisted of ten small rectangular pieces of red paper with a number ranging between on e and ten on one side, and the
156 experimentation room number on the other side. Each day I would hand out the ten cards and this would indicate the order in which the participants would go to the experimentation room. All psycholinguistic testing was perf ormed in the same quiet office (which I will removed suite with three other offices and a small common space between the offices. In the common space was a bookshelf and enough space for several chairs and a desk. T here were never any professors in their offices whi le the testing took place, which meant that when participants were filling out quest ionnaires after their testi ng session they were in a quiet area removed from the hallway and classrooms, and were thus afforded nearly complete privacy. At the beginning of the testing sessions the participants with numbers one and two left with me to walk to the testing office. The participant with the number one would enter the room first while the participant with the number two would sit in a chair outside the room Once the first participant was finished with the Pre Psycholinguistic Experime nt she would exit the room and the next participant (participant number two) would enter. Once the first participant exited the office she would sit the chair and fill out the D emographic Questionnaire (only after the Pre Psycholinguistic Experiment), the Attitude Q uestionnaire, and optionally the End of Task Response sheet in this order, as I instructed 8 At the top of each form the participant would write the ART number that he was given on the first day of class. There were three clearly marked la rge e nvelopes 8 It is impossible for me to know if the participants actually filled out the paperwork in the order I requested however I have no reason to believe that they did not.
157 sitting on the bookshelf in the common area and when the first participan t was finished he would put his responses in the appropriate envelope. As previously mentioned, the design of having the participants put their responses in a manila envelop e and not giving them directly to me was to help create a sense of anonymity, which I hoped would lead to more honest responses. Also, that the participants completed these questionnaires after the Psycholinguistic Experiment was a conscious design. I di d not want to prime the participants for language, race, socioeconomic status, or any other factors before they participated in the experiment. Next the participant (who was number one) would go back to class and let the participant with the number three know that it was time to go to the experimentation room By the time this third participant arrived the second participant was done with the Pre Test and he would begin filling out the forms while the third participant entered the office. This system w orked very smoothly and on average, eight participants were run during each class session. Two participants were taken at the beginning of each class session in order to create a flow such that there was always one participant in the room and one particip ant filling out the forms outside the offi ce. Apparatus The experiment was run using PsyScope X on a MacBook A response box that measures button presses with millisecond timing accuracy was on the desk between the participant and the computer. Partic ipants wore h eadpho nes that covered the entire ear, and the volume was set at the same level for each of the participants. The Experiment Participants sat at a desk in the experimentation room in front and a button response box I orally des cribed what the participant would do during the
158 experiment, and these instructions were followed by written instructions on the computer along with a practice trial with the response box. They placed the headphones on their ears and put their hands comfor tably on the response box. Any remaining questions were clarified, and on occasion a participant would ask a question during the instructional screens. Participants saw several of these instructional screens and were given an example screen during which they made an adjective selection choice by pressing the associated button on the button box. This gave them di rect practice as to how to select an adjective and use the button box during a real trial. Once the participant was ready to continue onto the ex periment they pressed a ny button. They then heard a story and at appeared. Once they were ready they pressed a button and thus began the adjective selection task. First a fixation cross appeared in the center of the screen to direct their attention away from either side and immediately after that, two adjectives appeared. One adjective was on the left side of the screen and one was on the right. If they thought the adjective on the left of the screen best descr ibed the speaker then they pressed the left most button on the response box, and if they thought the adjective on the right side of the screen best described the speaker then they pressed the right most button. After each pair of adjectives disappeared an other fixation cross appeared on the screen and immediately thereafter the next set of adjectives appeared. Once the participant s selected between the eleven adjective pairs they heard the second story and proceeded in the same manner until they had comp leted the task for all four stories. After the last selection task participants saw a screen thanking them for their time and
159 instructing them to announce that they had completed the task. I was sitting at a desk with my back facing the participant, and I then exited each one out of the program once they were finished. After the participants completed the experiment they exited the room and filled out the questionnaires that were described in the Logistics section above. The results from these questionna ires were described in Chapter 4, and the results from the experiment just described will now be presented in Chapter 6. I did not ever ask the participants if they were aware that they had only heard two speakers, and this would have been a good question to ask on the Post End of Survey Response form.
160 CHA PTER 6 PSYCHOLINGUISITC RESULTS The adjective discrimination ta sk was created to reveal biases that would be revealed by both adjective selection and the speed of response times. Within this discussion of psycholinguistic results, I will refer to the results as either results for the Pre Test or the Post Test. For t he Pre Test, overall, I thought that participants would most prefer Story Four (W AE) and least prefer Story One (AA AAE). I was not sure whether participants would prefer Story Three (W AAE) or Story Four (W AE) more, since this is a question of whether they prefer Academic English from a non White speaker, or if they prefer African American English from a White speaker, and I was was unsure as to whether participants wou ld be able to identify the race of each speaker, though Thomas & Reaser (2004) found that participants are often able to differentiate between African American and White speakers when they hear speech samples from speakers of the two different backgrounds. My underlying assumption Story One (AA AAE) was expected to elicit the most negative r espons es, with the other two stories in the middle. For the Post Test, I still expected participants to prefer Story Four (W AE) the most and Story One (AA AAE) the least, but I hoped that they would dislike Story One (AA AAE) less than they disliked it during the Pre Test. I was again unsure as to whether they would prefer Story Two (AA AE) or Story Three (W AAE), as it is a question of whether they prefer Academic English from an African American speaker or African American English from a White speaker.
161 A number of questions guided the analysis of the psycholinguistic results, which are divided into Pre Test results, Pos t Test results, and Demographics results Brief analysis of the questions guiding the regressions will appear within each section, and a more global analysis of the results tying the Sociolinguistic and Psycholinguistic results together can be found in Chapter 7. Table 6.1 again shows the Story combinations. Table 6 1. Story, speaker, and variety Story Speaker Variety One African American African American English Two African American Academic English Three White African American English Four White Academic English Pre Test Results Regression 1: Overall Results for the Pre Test A multivariate mixed effects regression was employed in order to find out how much time participants spent making their adjective selection in each of the story conditions and to get a ranking of speed across all four study conditions. The dependent variab le was the Pre Test time (measured in milliseconds) for each story adjective choice and the independent variables were story condition, adjective choice, and LSMEANS t tests and Tukey Kramer method for cont rolling for multiple comparisons. Table 6 2. Overall Pre Test speed of results Story Milliseconds One 2110 Two 1980 Three 1878 Four 1732
162 Table 6 2 above shows the average Pre Test response time by story in milliseconds, after controlling for the effect of adjective and type of adjective chosen. While we see in the timing of responses that Story One (AA AAE) was the slowest and in a linear fashion Story Two (AA AE) is the next slowest, Story Three (W AAE) is the next to fastest, and Story Four (W A E) is the fastest, we cannot say that this the exact wise comparisons are significant. Although looking only at the response time it may appear that the stories can be ordered, pair wise compariso ns show that only the interactions between Stories Two versus Four (t=2.87, adjusted p=.021), Stories One versus Three (t= 2.65, adjusted p=.040), and Stories One versus Four (t= 4.08, adjusted p=.0003) are significant. Regression 2a: Pre Test Positive Adjective C hosen A similar multivariate mixed effects regression was run to investigate the relationships between the stories when only the positive adjective was chosen. The dependent variable was the reaction time, and the independent variables were stor y condition and adjective choice, and the analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. According to r with results from the regression looking at the Pre Test positive adjective selection. The average Pre Test response times for selecting a positive adjective are from slowest to fastest: Table 6 3. Pre Test positive adjective chosen Story Speaker Speed T statistic P value 95% CI One AA AAE 2828 20.55 <.0001 2558, 3097 Two AA AE 1974 17.3 <.0001 1751, 2198 Three W AAE 1951 16.94 <.0001 1725, 2177 Four W AE 1640 15.61 <.0001 1434, 1846
163 The ranking of these stories is not as clear cut as it may look, however, since Stories Two (AA AE) and Three (W AAE) are not significantly different from each other (t=0.21, adjusted p=1.00). If we were to rank the stories according to preference, the results would be as follows, in order from slowest to fastest: Tab le 6 4. Pre Test story order Slowest Fastest Story One (AA AAE) Story Two (AA AE) Story Four (W AE) Story Three (W AAE) Looking at pair wise comparisons of the response times, all pairs are significantly different from each other except for the interaction between Stories Two (AA AE) and Three (W AAE). Regression 2b: Pre Test Negative Adjective C hosen A multivariate mixed effects model was created to investigate response times when participants selected a negative adjective in the Pre Test. In the model, response time was the dependent variable, and the independent variables were story condition and ad jective choice. The analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. For participants who chose a negative adjective, the story response times from slowest to fastest are as follows: Table 6 5. Speed of Pre Test negative adjective chosen Story Speaker Speed T statistic P value 95% CI Four W AE 2432 13.66 p<.0001 2084, 2781 Two AA AE 2374 16.32 p<.0001 2089, 2659 Three W AAE 2209 14.95 p<.0001 1919, 2498 One AA AAE 1976 14.96 p<.0001 1717, 2235 The slowest responses are for Story Four (W AE), then Story Two (AA AE), followed by Story Three (W AAE), and the fastest response is for Story One (AA AAE). Pair wise comparisons show that the interactions between Stories One (AA AAE) and Tw o (AA
164 AE) are significant (t=3.55, p=.0023) and between Stories One (AA AAE) and Four (W AE) are significant (t=2.98, p=.016), however the rest are not. To be slow at selecting a negative adjective means that it took the participants more time to decide t hat a negative adjective best described the speaker they just heard. Likewise, selecting a negative adjective quickly means that the participants had less difficulty with ascribing a negative trait to that speaker. After examining the results by overall preference and preference based on whether a positive or negative adjective was chosen, several other tests were run to investigate particular interactions that I thought might be of interest. Regression 3: Story Two (AA AE) Versus Story Four (W AE) A m ultivariate mixed effects model was created to answer the question of whether participants were slower to answer more positively for Story Two (AA AE) than for Story Four (W AE). Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. It was run on all story adjective observations in which participants chose the positive adjective in the Pre Test. Pre Test response time for each story was used as the dependent variable and the independent variable s were Stories One (AA AAE), Two (AA AE), and Three (W AAE) (with Story Four (W AE) as the reference category), adjective choice, adjective valence, and three interaction terms (which are products of Stories One (AA AAE), Two (AA AE) and Three (W AAE)), an d whether or not a positive adjective was chosen. I was interested in testing the significance of this relationship because both Story Two (AA AE) and Story Four (W AE) employ Academic English, however I was interested in whether participants would show a preference for one over the other. This question seeks to
165 ascertain whether participants are sensitive to an African American versus a White speaker using Academic English. For Story Two (AA AE), participants chose a positive adjective (marginally) sign ificantly more slowly than they did after hearing Story Four (W AE). As compared to hearing Story Four (W AE), participants responded to Story Two (AA AE) on average 339 milliseconds slower than they did to hearing Story Four (W AE) (t=1.67, p=0.095). R egression 4a: Stories Three and Four Versus Stories One and Two: Number of Positive Adjectives A mixed effects multivariate regression was run to determine whether Stories Three (W AAE) or Four (W AE) were associated with a higher number of positive adjec tives chosen than Stories One (AA AAE) or Two (AA AE). Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. Responses to Stories Three (W AAE) or Four (W AE) were associated with on average 2.3 6 (with 95% CI: (1.67, 3.04)) more positive adjectives chosen than responses to Stories One (AA AAE) or Two (AA AE) (p < 0.0001). Regression 4b: Stories Three and Four Versus Stories One and Two: Speed of Positive Adjectives This multivariate mixed effe cts regression examined if the ef fect of Story (Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE) ) on Pre Test response time was modified by whether the adjective was positive or negative. Pre Test respo nse time was the dependent variable, an d the independent variables were Story (Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE) ) the adjective choice, adjective valence, and an interaction term between adjective valence and story (Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE)). Analysis was
166 performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. There was a non significant effect of the adjective valence for the Pre Test response time s for Story One (AA AAE) or Two (AA AE) (t=1.66, p =0.096). T he effect of Story (Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE) ) on the Pre Test v aried according to whether the participant chose a p ositive or negative adjective (t= 4.16, p<0.0001). Participants respon ded on average 153 milliseconds slower to Stories Three (W AAE) or Four (W AE) (compared to Stories One (AA AAE) or Two (AA AE) ) when choosin g a negative adjective (t=1.65, p=0.10). On the other hand, participants responded on average 496 milliseconds faster to Stories Three (W AAE) or Four (W AE) (compared to Stories One (AA AAE) or Two (AA AE) ) when choosing a positive adjective (t= 5.98, p<.0001) Regression 5: Speed of Response When Choosing a Negative Adjective with All Stories Compared to Story Four (W AE) In this analysis I wanted to know if participants responded faster when choosing a negative adjective after hearing Stories One (AA AAE), Two (AA AE), and Three (W AAE), as compared to having heard Story Four (W AE). A multivariate mixed effects model was run, with response time as the depended variable and story condition, adjective valence, and adjective choice as the independent variables. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. Compared to Story Four (W AE), participants were significantly faster in choosing a negative adjective after hearing Story One (AA AAE), and they were faster, though not significantly, after hearing Stories Two (AA AE) and Three (W AAE), after controlli ng
167 for adjective chosen. Compared to having heard Story Four (W AE), participants who chose a negative adjective responded on average 456 milliseconds faster after hearing Story One (AA AAE) (p=0.0029), 59 milliseconds faster after hearing Story Two (AA A E) (p=0.72), and 224 milliseconds faster after hearing Story Three (W AAE) (p=0.17). Regression 6: Speed of Choosing a Negative Adjective with All Stories Compared to Story One (AA AAE) In this interaction I wanted to know whether participants responded more slowly when choosing a negative adjective after hearing Stories Two (AA AE), Three (W AAE), or Four (W AE) than they did after hearing Story One (AA AAE). As in the above interactions, a multivariate mixed effects model was run to answer this questio n. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. Participants were significantly slower in choosing a negative adjective after hearing Stories Two (AA AE), Three (W AAE), and Four (W A E), as compared to Story One (AA AAE), after controlling for adjective chosen. Compared to having heard Story One (AA AAE), participants who chose a negative adjective responded on average 397 milliseconds slower after hearing Story Two (AA AE) (p=0.0004) 232 milliseconds slower after hearing Story Three (W AAE) (p=0.038), and 456 milliseconds slower after hearing Story Four (W AE) (p=0.0029). Participants thus did respond more slowly when choosing a negative adjective after hearing Stories Two (AA AE), Three (AA AE), or Four (W AE), as compared to Story One (AA AAE). Overview of Pre Test Results Table 6 6 below shows the number of positive and negative adjectives selected, as well as the response times for the Pre Test. As discussed above, the signifi cant
168 interactions for the overall results are Stories Two (AA AE) versus Four (W AE), Stories One (AA AAE) versus Three (W AAE), and Stories One (AA AAE) versus Four (W AE). For the positive adjective selections all interactions are significant except for the interaction between Stories Two (AA AE) and Three (W AAE), and for the negative adjective selection the significant interactions are between Stories One (AA AAE) and Two (AA AE) and Stories One (AA AAE) and Four (W AE) Table 6 6. Overview of Pre Te st responses Story One (AA AAE) Story Two (AA AE) Story Three (W AAE) Story Four (W AE) Positive a djectives Number 160 278 279 380 Average t ime (ms) 2689 1853 1839 1603 Negative a djectives Number 80 61 81 41 Average t ime (ms) 2205 2159 2195 2066 Post Test Results As described in Chapter 5, the Post Psycholinguistic Experiment was given after the six teaching lessons. Regression 7: Overall Percentages of Positive Results A multivariate mixed effects regression model was created in order t o determine whether participants chose a positive adjective more often in the Post Test than they did in the Pre Test. The results show that they choose a positive adjective in the Post Test 67.44% of the time as compared to 59.31% of the time in the Pre Test, and some of the differences are significant, as will be discussed momentarily. Table 6 7 below displays the percentages indicating how often a participant chose a positive adjective for the Pre and Post Test.
169 Additionally, paired t tests were used to test if the difference between the mean Post Test number of positive adjectives is significantly different from the mean Pre Test number of positive adjectives chosen for each story. Four separate t tests were performed, and they are reported as follo ws in Table 6 8: Table 6 7. Percentage of positive adjective selected for Pre and Post Tests Story Pre Test Post Test One 34% 47% Two 60% 64% Three 61% 75% Four 83% 84% Table 6 8. Differences between the mean numbers of positive adjectives Story Speaker Mean d ifference T statistic P value 95% CI One AA AAE 1.04 2.4 0.02 .17, 1.91 Two AA AE 0.14 0.28 0.78 .87, 1.16 Three W AAE 1.06 2.06 0.045 .027, 2.10 Four W AE 0.22 0.43 0.67 1.28, .83 These results indicate that for Stories One (AA AAE) and Three (W AAE), the mean number of positive adjectives chosen across all subject adjective conditions in the Pre Test was significantly different from the mean number of positive adjectives chosen acr oss all subject adjective conditions in the Post Test within each condition. Regression 8: Overall Story Order Results for the Post Test To answer the more global question of which stories the participants preferred on the Post Test following the six le ssons, a multivariate mixed effects model was created. variables were the story condition, adjective selected, adjective valence, and Pre Test response times. Analysi s was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons.
170 The average Post Test response time by story is as follows in Table 6 9, from slowest to fastest in milliseconds: Table 6 9. Average Post Test res ponse times Story Speaker Speed 95% CI One AA AAE 1946 1765, 2128 Two AA AE 1918 1741, 2095 Four W AE 1799 1618, 1979 Three W AAE 1718 1541, 1895 The story order indicates that Story One (AA AAE) is the slowest, followed by Story Two (AA AE), followed by Story Four (W AE), and Story Three (W AAE) is responded to the fastest. After completing pair wise comparisons for each of the story interactions to determine significance, only the interaction between Stories One (AA AAE) and Three (W AAE) is significant. This means that the only true difference among prefer t he White speaker using African American English over the African American speaker using African American English. Because the rest of the interactions are not significant, the stories cannot be ordered in terms of most to least preferred by the participan ts. Regression 9a: Post Test Positive Adjective C hosen In this analysis I wanted to understand the relationships between positive adjectives selected and response times. In the multivariate mixed effects regression, the dependent variable was the Post Test response time, and the independent variables were the story condition, adjective chosen, and the Pre Test response times. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons.
171 Below are the response times in milliseconds from slowest to fastest for the positive adjectives: Table 6 10. Post Test positive adjective response times Story Speaker Speed 95% CI One AA AAE 2369 2147, 2590 Two AA AE 1786 1591, 1982 Three W AAE 1687 1505, 1868 Four W AE 1656 1467, 1845 Here we see the same order of preference as for Regression 8: Overall Story Order Results for the Post Test, which is from slowest to fastest Story One (AA AAE), Story Two (AA AE), Story Four (W AE), and Story Three (W AAE). According to pair wise compari sons, only the differences in response time between Story One (AA AAE) and the other three stories are significant; that is, Story One (AA AAE) versus Story Two (AA AE) (t=5.21, p <0.0001), Story One (AA AAE) versus Story Three (W AAE) (t=6.21, p <0.0001), and Story One (AA AAE) versus Story Four (W AE) are significant at (t=6.61, p<.0001). Regression 9b: Post Test Negative Adjective C hosen In this analysis the selection of a negative adjective in the Post Test was examined. In the multivariate mixed eff ects model, the dependent variable was the Post Test response time and the independent variables were the story condition, adjective chosen, and the Pre Test response time. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for contr olling for multiple comparisons. Pair wise comparisons show that there are no significant differences between any of the stories for response time, with all p values greater than .57. Although the response time order is presented below, there is no disce rnable difference between
172 indicated in Table 6 11, are Story Four (W AE), Story Two (AA AE), Story Three (W AAE), and Story One (AA AAE) is the fastest. Table 6 11. Post Test negative adjective response times Story Speaker Speed 95% CI Four W AE 2132 1751, 2513 Two AA AE 2097 1787, 2406 Three W AAE 2038 1727, 2349 One AA AAE 1933 1659, 2206 Regression 10a: Overall speed of adjectives chosen for Story One A mixed effects multivariate regression model was created to examine whether in the Post Test the positive responses to Story One were faster than they were in the Pre Test. Response time was the dependent variable and the independent variables were inter vention (Pre Test or Post Test) and adjective choice. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were not found to significantly affect response times for responses fr om Story One (t= 1.17, p=0.24). Participants were on average 104.33 milliseconds faster in choosing any adjective in the Post Test than they were for the Pre Test. Regression 10b: Story One: Comparison of speed of positive adjectives chosen between Pre and Post Test A mixed effects multivariate regression model was created to examine whether in the Post Test the positive responses to Story One were faster than they were in the Pre Test. Response time was the dependent variable and the independent vari ables were intervention (Pre Test or Post Test) and adjective choice. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were found to significantly affect response t imes for responses from Story One (AA AAE) ( t= 2.69, p=0.0076). Participants were on
173 average 517.32 milliseconds faster in choosing a positive adjective in the Post Test than they were for the Pre Test Regression 10c: Story One: Comparison of speed of negative adjectives chosen between Pre and Post Test A mixed effects multivariate regression model was created to examine whether in the Post Test the positive responses to Story One (AA AAE) were faster than they were in the Pre Test. Response time was the depe ndent variable and the independent variables were intervention (Pre Test or Post Test) and adjective choice. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were not found to significantly affect response times for negative resp onses from Story One (AA AAE) (t= 0.39, p=0.70). Participants were on average 35.56 milliseconds faster in choosing a n egative adjective in the Post Test than they were in the Pre Test. Regression 11: Story Two (AA AE) Versus Story Four (W AE): Positive Adjective Selection This interaction was created to answer the question of whether participants were slower to answer in the positive for Story Two (AA AE) than for Story Four (W AE). A multivaria te mixed effects regression was used to address this question, and Post Test response time for each story was used as the dependent variable, with the independent variables in the regression being Stories One (AA AAE), Two (AA AE), and Three (W AAE), versu s Story Four (W AE), adjective and Pre Test response time This regression was run on all Story adjective observations in which participants chose the positive adjective in the Post Test. The results show that there was not a significant interaction be tween Story Two (AA AE) and Story Four (W AE). Participants chose positive adjectives after hearing
174 Story Two (AA AE) on average 100 milliseconds slower than after hearing Story Four (W AE) (p =0.29), so while they were somewhat slower to select positive adjectives after hearing Story Two (AA AE) versus Story Four (W AE), this result is not significant and there is therefore no discernable difference between the two. Regression 12a: Stories Three and Four Versus One and Two: Number of Positive Responses A mixed effects multivariate regression model was used to see if Stories Three (W AAE) or Four (W AE) were associated with a higher number of positive adjectives chosen in the Post Test than Stories One (AA AAE) or Two (AA AE), controlling for the number of positive adjectives chosen in the Pre Test condition. Responses to Stories Three (W AAE) or Four (W AE) were associated with on average 1.28 (with 95% CI: (0.72, 1.85)) more positive adjectives chosen than responses to Stories One (AA AAE) or Two (AA AE ) (p < 0.0001). Regression 12b: Stories Three (W AAE) and Four (W AE) Versus One (AA AAE) and Two (AA AE): The Relationship Between Positive and Negative Responses In this regression I wanted to know what the relationship was between positive and negati ve responses for Stories Three (W AAE) and Four (W AE) versus Stories One (AA AAE) and Two (AA AE). In the regression the Post Test response times were used as the dependent variable and it was examined whether the effect of s tory (Three (W AAE) or Four ( W AE) versus One (AA AAE) or Two (AA AE) ) on Post Test response time was modified by whether the adjective was positive or negative. Therefore, the independent variables in the regression were Story (Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE)), positive adjective chosen, interaction term created using story and positive adjective chosen, adjective choice, and Pre Test time.
175 There was a significant effect of the interaction term on the outcome for the Post Test response time. The ef fect of Story ( Three (W AAE) or Four (W AE) versus One (AA AAE) or Two (AA AE)) on the Post Test varied according to whether the participant chose a positive or negative adjective. Partic ipants responded on average 9 milliseconds slower to Stories Three ( W AAE) or Four (W AE) (compared to Stories One (AA AAE) or Two (AA AE) ) when choosing a negative adjective (p=0.93), which is non significant, however particip ants responded on average 300 milliseconds faster to Stories Three (W AAE) or Four (W AE) (compared to Stories One (AA AAE) or Two (AA AE)) when choosing a positive adjective (p=0.016), which is significant. There was therefore a significant interaction of story condition and adjective valence on the Post Test response time. Regression 13a: Stories Two (AA AE), Three (W AAE), and Four (W AE) Versus Story One (AA AAE): Positive Responses A multivariate mixed effects regression was used to address the question of whether participants were slower to answer in the positive for Story One (AA AAE ) than for the other stories. Post Test response time for each Story was used as the dependent variable and the independent variables in the regression were Stories Two (AA AE), Three (W AAE), and Four (W AE), versus Story One (AA AAE), adjective chosen, and Pre Test time for that same adjective Story condition Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The regression was run on all story adjective observations in which participants chose the positive adjective.
176 For Stories Two (AA AE), Three (W AAE), and Four (W AE) participants who chose a positive adjective in the Post Test conditions responded significantly faster than they did after hearing Story One (AA AAE) Compared to h aving heard Story One (AA AAE), p articipants responded to Story Two (AA AE) on average 582.29 milliseconds faster (p <0.0001), 713 milliseconds faster to Story Three (W AAE) (p <0.0001), and 682 milliseconds faster to Story Four (W AE) (p <0.0001). Overa ll, then, we see that participants responded more positively to Stories Two (AA AE), Three (W AAE), and Four (W AE) than to Story One (AA AAE) for positive responses. Regression 13b: Stories Two (AA AE), Three (W AAE), and Four (W AE) Versus Story One (AA AAE): Negative Responses The preceding results were then re run and were restricte d to participants who chose a negative adjective in the Post Test instead of a positive adjective using the same methodology as described above For Stories Two (AA AE) Three (W AAE) and Four (W AE) participants who chose a negative adjec tive in the Post Test responded significantly more slowly than they did after hearing Story One (AA AAE) Therefore, for all stories, participants were faster to choose a negative adje ctive after hearing Story One (AA AAE) than after hearing any other story. Compared to having head Story One (AA AAE), p articipants responde d to Story Two (AA AE) on average 546 milliseconds slower (p <0.0001), 379 milliseconds slower to Story Three (W AAE) (p =0.0139), and 6 48 milliseconds slower to Story Four (W AE) (p=0.0004). Again, participants were most negative towards Story One (AA AAE).
177 Regression 14a: Stories One (AA AAE) and Two (AA AE): Speed of Positive Responses in the Post Test Versus Pre Test A mixed effects multivariate regression model was created to examine whether in the Post Test the positive responses to Stories One (AA AAE) and Two (AA AE) were faster than they were in the Pre Test. Response time was the dependent variable and the independent variables were story condition (One (AA AAE) or Two (AA AE)), Pre Test or Post Test, and adjective choice. Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were found to significantly affect response times for responses from Stories One (AA AAE) and Two (AA AE) (p=0.029). Participants were on average 201 milliseconds faster in choosing a positive adjective in the Post Test than they were for the Pre Test. When participants responded positively to Stories One (AA AAE) and Two (AA AE), they were faster in the Post Test. Regression 14b: Stories One (AA AAE) and Two (AA AE): Speed of Negative Responses in the Post Test Versus Pre Test The same mixed effects mu ltivariate regression model was used as in the previous question that assessed positive responses, yet in this case it was used to investigate negative responses. Response time was the dependent variable and the independent variables were story condition (One (AA AAE) or Two (AA AE)), Pre Test or Post Test, and adjective choice. As opposed to the previous question in which participants chose a positive adjective faster in the Post Test than in the Pre Test for Stories One (AA AAE) and Two (AA AE), in this case, in terms of the negative responses, the six lessons were not found to significantly affect response ti mes for responses from Stories One (AA AAE) and Two (AA AE) (p=0.64). Participants thus did not select negative adjectives faster in the Post Test
178 Regression 15 a: Stories Three and Four: Speed of positive resp onses in the Post Test versus Pre T est A mixed effects multivariate regression model was created to examine whether in the Post Test the positive responses to Stories Three (W AAE) and Four (W AE) were faster than they were in the Pre Test. Response time was the dependent variable and the independent variables were story condition (Three (W AAE) or Four (W AE)), Pre Test or Post Test, and adjective choice. Analysis was performed using the LSM EANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were not found to significantly affect response times for responses from Stories Three (W AAE) and Four (W AE) (t= 1.52, p=0.13). Participants were on average 91.18 milliseconds faster in choosing a positive adjective in the Post Test than they were for the Pre Test. Regression 15 b: Stories Three and Four: Speed of negative res ponses in the Post Test versus Pre T est A mixed effects multivariate regression mode l was created to examine whether in the Post Test the negative responses to Stories Three (W AAE) and Four (W AE) were faster than they were in the Pre Test. Response time was the dependent variable and the independent variables were s tory condition ( Thre e (W AAE) or Four (W AE) ), Pre Test or Po st Test, and adjective choice Analysis was performed using the LSMEANS t tests and Tukey Kramer method for controlling for multiple comparisons. The six lessons were not found to significantly affect response time s for responses from Stories Three (W AAE) and Four (W AE) (t=0.59, p=0.56). Participants were on average 74.41 milliseconds slower in choosing a negative adjective in the Post Test than they were for the Pre Test.
179 Overview of the Post Test responses Ta ble 6 12 below shows the number of positive and negative adjectives selected, as well as the response times for the Post Test. Table 6 12. Overview of the Post Test responses Story One (AA AAE) Story Two (AA AE) Story Three (W AAE) Story Four (W AE) Positive a djectives Number 210 285 330 369 Average t ime (ms) 2218 1663 1587 1609 Negative a djectives Number 240 158 110 71 Average t ime (ms) 1748 2359 2161 2410 As discussed, the only significant interaction in the overall results is between Stories One (AA AAE) and Three (W AAE). For the positive adjective selections, only the interactions with Story One (AA AAE) are significant, thus the interactions between St ories One (AA AAE) and Two (AA AE), One (AA AAE) and Three (W AAE), and One (AA AAE) and Four (W AE) are significant. For the negative adjective selections, there were no significant interactions between any of the stories. Percentages of positive and n egative adjectives in the Pre and Post Test Table 6 13 provides an overview of the percentages of positive and negative adjectives selected in both the Pre and Post Test. Table 6 13. Percentages of positive and negative adjectives in the Pre and Post Test Frequency Percent Pre Test Positive 1115 59 Negative 765 41 Post Test Positive 1214 67 Negative 586 33 In order to determine if the differences between the Pre and Post Test results were results are reported in Table 6 14.
180 Table 6 Chi s quare Degrees of f reedom P v alue 47.85 1 p<.0001 These results indicate that the percentages of positive results do differ significantly between the Pre and Post Test. Discussion and Post comparisons between Stories One (AA AAE) and Two (AA AE) are not significant, we see that participants are not differentiating between the speech of the African American speaker. Also, because Stories Three (W AAE) and Four (W AE) are not significant, we see that they are not differentiating between the speech of the White speaker. T here is also not a clear difference between Stories Two (AA AE) and Three (W AAE), meaning that the participants do not have a marked preference for Academic English from an African American speaker or African American English from a White speaker. They d o, however, strongly prefer the Academic English from the White speaker over the Academic English from the African American speaker, the African American English from the White speaker over the African American English from the African American speaker, an d finally the Academic English from the White speaker over the African American English from the African American speaker. They thus prefer the White speaker over the African American speaker in every combination in the Pre Test. In comparison, in the Pos t Test, we see that participants selected positive adjectives more often for every story for the Post Test as compared to the Pre Test. For St ories One (AA AAE) and Three (W AAE), the mean number of
181 positive adjectives chosen across all subject adjective conditions in the Pre Test was significantly different from the mean number of positive adjectives chosen across all subject adjective conditio ns in the Post Test within each condition. Considering that the participants either are or soon will be teachers, these results are promising in terms of achieving the desired outcome of having teachers respond more positively to students who speak a non standard variety. selecting a positive adjective, participants did not have a clear preference between an African American speaker using Academic English and a White speaker using Afr ican American English, but they did clearly prefer the White speaker using Academic English over the African American speaker using African American English. In comparison to the Post Test Positive Adjective Chosen after the six lessons, the participants still showed the most conflict in responding positively to Story One (AA AAE), yet there were no discernable differences between their preferences between the other stories. This is a positive result in term s of the positive towards African American English. Although they still disliked the African American English from the African American speaker, they did not strongly disli ke the African American English from the White speaker, or the Academic English from the African American speaker. preferred the Academic English from the African American speak er over the African American English from that same speaker. Also, like in Regression 2a, they strongly
182 preferred the White speaker using Academic English over the African American speaker using African American English. It is noteworthy that the partici pants preferred the Academic English over the African American English from the same speaker. This may possibly indicate that they were unaware that they had just heard the same speaker twice, but it clearly indicates that they prefer Academic English ove r African American English. Test negative adjective Regression 2b, participants strongly preferred the Academic English form the African American speaker over the African American English from the African American Speaker, as well as strongly preferring the White speaker speaking Academic English over the African American speaker speaking in African American English I consider this lack of significant differences in the Post Test for negative adjective choices a positive finding. It means that participants did not have such an immediate negative reaction to hearing African American English. A major research questi on in the current work was whether teaching about the systematized and rule governed nature of African current results indicate that the lessons may indeed have improved th eir attitudes. One (AA AAE): Comparison of speed of p ositive adjectives chosen between Pre and Post significantly affect the response times when investigating the positive adjectives, and
183 they were faster at choosing a positive adjective after the lessons than they were before the lessons. This is a very positive finding, as it indicates that after learning about the syntax, phonology, and cultural issues surrounding the use of African American English and Academic English, the participants selected positive adjecti ves significantly more One (AA AAE): Comparison of speed of negative adjectives chosen between Pre and Post times to negative adjectives. This also is a positive finding, as it indicates that participants did not become more negative towards African American English as a result of the lessons. AA AE) versus Story Four (W AE) in the Pre results show that participants did notice a difference between the Academic English of the White speaker and the African American speaker, and they needed more time to ascribe a positive adjective to the speaker after hearing Story One (AA AAE) than after hearing Story Four (W AE) versus Story Four (W AE): Positive Adjective Selection in the Po st significant interaction between Story Two (AA AE) and Story Four (W AE). This is a positive result in terms of the research goals, as it means that participants are not distinguishing between the Academic English of th e White speaker and that of the African American speaker. This is in comparison to Regression Three of the Pre Test in which participants did distinguish between the Academic English of the African American and the White speakers. A question that I have considered is whether African
184 According to these results they will not be, as long as they are being judged by someone who is knowledgeable about the linguistic v arieties. These findings could indicate that linguistically educated listeners care less about the race of a speaker and more about the grammar and phonology of the speech itself. In terms of educating all students about the grammar and phonology of Academic English, these results are promising vis `a vis social mobility and social equity for speakers of African American English. Racial disparity and inequity is a very real issue in the United States, especially in the job hiring process, so if an in terviewee who speaks African American English as an L1 is able to acquire Academic English, her career and financial prospects may well benefit from the acquisition of the additional, more formal variety. This would be especially true if the listening aud ience is familiar with the concept of African American English being a rule governed linguistic variety. AAE) and Two (AA AE): Speed of positive responses in the Post Test versus Pre positively to Stories One (AA AAE) and Two (AA AE), they were faster in the Post Test, indicating that their attitudes changed in the direction of being more positive towards African American English or an African American speaker after participating in t he six AAE) and Two (AA AE): Speed of Negative Responses in the Post Test Versus Pre select negative adjectives faster in the Post Test, and this is a positive finding in terms of the research goals.
185 AAE) and Four (W AE) versus Stories One (AA AAE) and Two (AA White speaker over the African American speaker, regardless of what variety the White AAE) and Four (W AE) versus Stories One (AA AAE) and Two (AA responded significantly faster when selecting a positive adject ive to Stories Three (W AAE) and Four (W AE), indicating that they preferred the White speaker regardless of what variety she was employing. All Stories Compared to Story Four (W AE) in the Pre more negatively to the African American speaker using African American English than Choosing a Negative Adjective With al l Stories Compared to Story One (AA AAE) in the Pre African American speaker using African American English than to any other story interaction. y Order Results for the Post participants strongly preferred the White speaker using African American English over the African American speaker using that same variety. It is interesting that the interaction that the participants m ost reacted to is the African American English usage by the White and African American speaker. I would have expected there to be a lingering preference for the White speaker using Academic English over the other three possibilities, but this was not the case. Perhaps this is because so much of the class
186 time was spent focusing on African American English and not on Academic English. That they preferred the White speaker over the African American speaker may indicate that they can tell the difference betw een the races even when both are speaking in the same African American English variety. As compared to the Pre Test, though, the participants did not distinguish strongly between the Academic English from the African American speaker and the Academic Engli sh from the White speaker, so this indicates that they heard Academic English simply as that -Academic English -AAE) and Four (W AE) versus One (AA AAE) and Two (AA AAE) and Four (W AE) are associated with more positive adjectives selected. This they showed a preference for a White speaker even in the case in which she was AAE) and Four (W AE) versus One (AA AAE) and Two (AA AE): The Relationship Between Positive and Neg for Stories Three (W AAE) and Four (W AE), and we see that it was easier, even for the African American English story (Story Three (W AAE)), for participants to ascribe positive chara cteristics to African American English after attending the six lessons than before attending the lessons. AE), Three (W AAE), and Four (W AE) versus Story One (AA AAE): Positive Responses in the Post show that for Stories Two (AA AE), Three (W AAE), and Four (W AE), participants who chose
187 a positive adjective in the Post Test did so significantly faster than they did after hearing Story One (AA AAE). This indicates a lingering resistance to accepting the speech of the African American woman speaking African American English. Regression 13b, AE), Three (W AAE), and Four (W AE) versus Story One (AA AAE): Negative responses in the Post AE), Three (W AAE), and Four (W AE), participants who chose a negative adjective in the Post Test did so significantly slower than they did when responding to Story One (AA AAE). We see therefore that overall for these two conditions, participants had the most diff iculty ascribing positive traits to the speaker in Story One (AA AAE). AAE) and Four (W AE): Speed of positive responses in the Post Test versus Pre significantly affect participan AAE) and Four (W AE); Speed of negative responses in the Post Test versus Pre the Pre and Post Test when selecting a negative adjective. Additionally, when considering the results from the percentages of positive and negative adjectives selected in the Pre Test compared to the Post Test, we see clearly that the participants responses became more positive. Demographics Results were also broken down by multiple demographic categories in order to determine whether there were differences within the participant groups before and after the lessons. All results were obtained by running multivariate mixed effects regression models. Demographic categories investigated were:
188 Gender Race Age From: East Coast versus West Coast From: New York versus Elsewhere Teacher versus Non Teacher Years Teaching Race of Students Pre Test Results Multivariate mixed effects regression models were run on all Pre Test results for the demographic categories above in order to ascertain whether the participants had approximately the same baseline responses. Aside from whether the participants were a tea cher or a non teacher (which was only marginally significant), all comparisons were non significant, indicating that there were not differences between the groups. This is an ideal place to begin, as we can assume that any differences after the six lesson s are actually changes resulting from having attended the lessons. The significance levels of the Pre Test results for the demographic categories were as follows: Professor (p=.95), Gender (p=.87), Race (p=.80), Race of Students (Hispanic mixed: p=.35), Years Teaching (p=.57), Teacher versus Non Teacher (results will be discussed), from New York versus Elsewhere (p=.49), East Coast versus West Coast (p=.32), and Age Group (p=.29). Thus, aside from whether the participants was a teacher or not, all other comparisons were shown to not be different from each other. Post Test Results Professor not associated with faster responses when controlling for adjective (positive or negative), Story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre
189 Test time for that same story and adjective condition. The estimate of the effect of 2.22 (standard error=155.55), with a p not associated with a greater likelihood of choosing a positive adjective in the Post Test adjective choice task (odds ratio= 1.23, 95% CI for odds ratio: (0.71, 2.14), p=0.45) when controlling for Pre Test adjective (positive or negative), story condition, or adjective choice. Gender The gender of the participant is not associated with faster responses when controlling for adjective valence, story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre Test time for that same story and adjective condition. The estimate of the effect of being female (vs. male) is 161.98 (standard error=152.22) with the p value=0.29. Being female was also not associated with a greater likelihood of choosing a positive adjective in the Post Test adjective choice task (estimated OR= 0.71, 95% CI: (0.43, 1.17), p=0.18) when controlling for Pre Test adjective valence, story condition, and adjective ch oice. Race not associated with faster responses in the Post Test when controlling for adjective valence, Story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective chosen and Pre variable in the multivariate mixed effects model and no p values were lower than .40.
190 Age groups Age groups were broken down by decade, thus 20 29, 30 39, 40 49, and 50 59. The age group of the participant was not associated with faster responses in the Post Test when controlling for adjective valence, Story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre Test time for that same story and adjective condition. The estimate of the effect of a one decade increase in age group is parameter estimate= 1.27, standard error= 7.36, p=0.86. East Coast versus West Coast Whether the partici pants were from the East Coast or the West Coast was not associated with faster responses when controlling for adjective valence, Story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre Test time for that same Story and adjective condition. The estimate of the effect of East Coast (compared to West Coast/other) is parameter estimate= 57.57, standard error= 194.45, p=0.77. New York versus elsewhere Being born in New York was al so not associated with faster responses when controlling for adjective valence, Story condition ( Two (AA AAE), Three (W AAE), or Four (W AE) with Story One (AA AAE) as the reference), adjective choice, and P re Test time for that same Story and adjective c ondition. The e stimate of the effect of being born in New York on response time s (compar ed to being born anywhere else) was parameter estimate= 189.11, standard error= 152.48, p=0.22.
191 Teacher versus non teacher The comparison of Teacher versus Non Teacher had a marginally significant association with response time in the Post Test time. Being a teacher is associated with somewhat longer response times when controlling for adjective valence, story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre Test time for that same story and adjective condition. The estimate of effect of being a teacher (compared to non teacher) is parameter estimate= 411.77, standard error= 212.84, p=0.053. These results show that participants who were currently teaching had a marginally slower response time. Length of t eaching There were three categories created for length of teaching: Low, Medium, and was 6 over 10 years. Based on these categories, length of teaching is not statistically significantly associated with response time in the Post Test time. However, a one unit increase in category (thus from low to mediu m, or medium to high) is significantly associated with a decreased likelihood of choosing a positive adjective in the Post Test when controlling for Pre Test adjective choice, story condition, and adjective valence: OR=0.95 95% CI: (0.90, 0.99), p=0.021. This means that participants who have been teaching for longer are more likely to ascribe negative traits to the speakers than are participants who have been teaching for a shorter period of time. Race of s tudents down by several categories, such as African American, Hispanic, and White, there were not enough students to fill these categories. Almost all participants listed having ei ther
192 Hispanic students or Hispanic plus another race, which was sometimes listed as Hispanic The results show that being Hispanic (and not Hispanic plus another race) was not associated with the adjective participants chose when controlling for Pre Test positive responses or adjective valence, story condition, or adjective choice: OR=0.65 95% CI: (0.29, 1.47), p=0.30. Having Hispanic students (compared to other races of students in the class) is also not significantly associated with a shorter response time in the Post Test. Students who are categorized as Hispanic in any way (which I will call are significantly associated with shorter response time in the post intervention time when compared to the other races of students. Participants with Hispan ic mixed students are associated with shorter response times when controlling for adjective valence, story condition (Two (AA AAE), Three (W AAE), or Four (W AE), with Story One (AA AAE) as the reference), adjective choice, and Pre Test time for that same story and adjective condition. The e stimate of the effect of Hispanic mixed (versus ) is parameter estimate= 910.23, standard error= 283.44, p=0.0040. Demographics: Pre and Post Test Comparisons Separate multivariate mixed effects regression mo dels were employed to address the question of what demographic variables predict the change in attitudes after participating in the six lessons. The difference in response time (in milliseconds) between Pre and Post Test response times was calculated on a story adjective level and used as the dependent variable in all regression models.
193 Table 6 15. Pre and Post Test comparisons P articipant and student d emographics Parameter e stimate (95% CI) T value P value s c lass 81.65 ( 358.37, 195.06) 0.58 0.56 Gender= f emale 31.82 ( 307.25, 243.62) 0.23 0.82 Race of p articipant (overall) F test=0.32 (7, 1793) 0.95 African American (versus White) 41.57 ( 399.76, 316.62) 0.23 0.82 Mixed r ace (versus White) 268.16 ( 223.90, 760.21) 1.07 0.29 Hispanic (versus White) 137.70 ( 694.84, 419.45) 0.48 0.63 Age 2.04 ( 15.18, 11.10) 0.3 0.76 Where from: east c oast v ersus west c oast 201.61 ( 140.06, 543.27) 1.16 0.25 Where from: New York versus e lsewhere 239.59 ( 28.17, 507.34) 1.75 0.079 Teacher versus n on teacher 89.37 ( 477.32, 298.57) 0.45 0.65 Length of t eaching 25 ( 68.97, 19.01) 1.11 0.27 Students are Hispanic (versus all other students) 123.55 ( 300.74, 547.84) 0.57 0.57 Students are Hispanic m ixed 592.69 ( 1173.09, 12.29) 2.00 0.045 These models examined the effect of each demographic predictor on change in response time, after holding constant the effects of the other predictors in the model: (1) Story Two (AA AE) versus Story One (AA AAE), (2) Story Three (W AAE) versus Story One (A A AAE), (3) Story Four (W AE) versus Story One (AA AAE), and (4) adjective choice. This regression was run on all story adjective observations. As shown in Table 6 15 above, none of the demographic variables were significantly related to the change in res ponse time between the Pre and Post Test,
194 after controlling for story and adjective conditions (p > 0.07), except for the participants associated on average with a 1200 milli second change in response time. Additionally, one group had a marginally significant effect on the change in response time, which is that respondents from New York (versus anywhere else) were associated on average with a 239 millisecond change in response time. Demographics Discussion For the most part, the demographic characteristics of the participants did not student was enrolled in, I consider it a positive finding that even though I taught two differences in the conversations we had in the three classes, the results still showed no differences between them. Regarding gender, I did not have a clear hypothesis about whether females or males would change more, though I did think it possible that females might grow more positively. In terms of significance, however, males and females performed the same on the Post Test. When planning thi s experiment and even throughout the experiment, I did not have a clear hypothesis about whether White, African American, Hispanic, or Mixed participants would attain more positive attitudes. I had several reasons for thinking that different groups would o r would not respond positively to the experience. I thought that it was possible that the White participants would change their attitudes more positively because they did not speak African American English and therefore had not grown up being told or beli eving that the way they spoke was lazy or sub standard. Because of this lack of negative buildup over the years, I thought it might be possible that it would
195 be easier for them to accept the information I was teaching as legitimate, and incorporate the id eals into their thinking. I also, however, thought that the White participants might not move forward as quickly because they are a majority group whose variety is more highly valued and esteemed, and I thought that it might be easy for them to discount the information and simply choose to not believe it because socially, since they are in the more powerful Keeping in mind that language and identity are inextri cably linked, reasons for thinking that non White participants might become more positive include that they would learn that tenets that they had believed their entire lives about their language were untrue. I reasoned that they might enjoy and feel empow ered by knowing that their variety is not any less valid than Academic English, and therefore they would embrace the information. However, because of the fact that many (if not all) of the African American participants grew up hearing that their language English is more valued, they might have had more hurdles to jump through than the White participants and therefore their attitudes would change mor e slowly. This would mean that they, and I grouped the other non White participants into this category, would show less positive attitudinal change than the White participants. However, because there were no significant differences between races, we see that participants grew together and not as separate groups. This is a positive result in terms of repeating this experiment with various groups of races of participants.
196 In terms of age, going into the teaching sessions, my hypotheses were that the you nger teachers might be more receptive to the lessons than the older teachers because I thought that they might be less set in their ways. Although education classes are not nearly thorough enough on the topics of linguistic varieties or the connections be younger teachers might be somewhat more exposed to those ideas. Throughout the teaching sessions, however, I did not notice this to be the case simply through conversations I had with the participants. In retrospect, a question that I wish I had asked on the Background Questionnaire was whether they had ever been exposed to this information in previous classes or from any other source. Considering where participants were f rom, with the East Coast versus West Coast analysis I thought that it was possible that there would be an effect of the showed otherwise. For the New York participants as with the East Coast versus West Coast comparison, I considered the possibility that because the speakers the participants heard were from New York, there was a possibility that New Yorkers would respond differently than non New Yorkers simply because they were familiar with the accent. This was not the case. Again, these results are positive in terms of replicating the experiment with participants who are from different regions of the country. Regarding the Teacher versus Non Teacher comparison, the increased response time from teachers could be because they have learned through teaching to take time before judging a student, however the exact reason for why they responded somewhat slower is unknown. In terms of length of teaching, the participa nts who had been
197 teaching for longer were more likely to ascribe negative traits to the speakers than were participants who had been teaching for a shorter period of time. This could be because more seasoned teachers have more ingrained thoughts about the language, or it could be because they have a more in depth understanding of the importance of acquiring Academic English for academic and future career purposes. We are unable to know the reason for this difference, and certainly there could be a variety of other reasons to explain this difference. In terms of the race of the students, I was especially interested in whether the hypothesized that participa nts whose students were African American or any other group that spoke African American English or another non standard variety might pay closer attention to the lessons and might see the value in learning and applying what I was teaching more than those w ho taught only students who already commanded Academic English. Whether or not my reasoning is what caused the growth, participants whose students were of Hispanic mixed attitudes changed in a positive way (and this includes African American students). P articipants whose students were only classified as Hispanic did not improve significantly, though they did have shorter response times in the Post Test, whereas participants whose students were a mix of Hispanic and other races or ethnicities did improve s ignificantly. This difference may be because the participants with Hispanic only students did not think that their students spoke African American English and they therefore did not pay as much attention or apply the strategies with their students because they did not believe their students to be
198 using linguistic features of African American English. This could perhaps be in part because of the nomenclature of African American English itself. Psycholinguistic Conclusions and Post Testing measures, and results from the Post Test show that the participants grew in the direction of being more positive towards African American English. In the Pre Test, wh ich was run before participants were exposed to the idea of African American English or Academic English in this setting, participants preferred the White speaker over the African American speaker in all combinations. They preferred the Academic English f rom the White speaker over the Academic English from the African American speaker, the African American English from the White speaker over the African American English from the African American speaker, and the Academic English from the White speaker over the African American English from the African American speaker. In the Post Test story order preference were between the White speaker using African American English and the African American speaker using African American English. The participants significantly preferred the White speaker using African American English over the African American speaker using African American English. This means that the participants were less negative towards the African American speaker when she used Academic English, as the participants did not show a preference for the White speaker over the African American speaker when they used Academic English. ction of positive adjectives, their preferences changed between the Pre and Post Testing. In the Pre Test they responded slowest to Story One (AA AAE), there was no significant difference between Stories Two and
199 Three, and they responded fastest to Story Four (W AE). By the Post Test they still responded slowest to Story One (AA AAE), however there were no significant differences between the speed of their responses to the other stories. This means that by the end of the teaching sessions they still dis preferred Story One (AA AAE), yet it was easier for them to make the decision to select a positive adjective for the other stories. It was also easier to select a positive adjective for Story One (AA AAE) in the Post Test than in the Pre Test, and this i s a very positive finding. When selecting negative adjectives, responses varied greatly between the Pre and Post Test. For the Pre Test participants significantly preferred the Academic English from the African American speaker over the African American English from the same speaker, and they also significantly preferred the Academic English from the White speaker over the African American English from the African American speaker. By the Post Test, however, there were no significant differences in parti responses, which indicates that they did not have strong negative preferences for any of the speakers. Looking at the interactions between Stories Two (AA AE) and Four (W AE) for the Pre and Post erred the Academic English from the White speaker in the Pre Test, yet by the post test there was not a significant difference between the two stories. This means that participants did not have a preference for Academic English from either the White or Af rican American speaker by the end of the lessons. In terms of broader implications this is a very positive finding. It indicates that people who have been educated about African American English do not show a strong preference for one race of speaker whe n Academic English is being
200 used. Considering that when interviewing for a job an employer is often expecting to hear Academic English, this is positive in that it indicates that if an African American person is able to produce Academic English, there mig ht be less discrimination based on language. It indicates that it is worthwhile for a mono dialectal speaker to acquire Academic English because people who have been educated about the variety (and certainly these assertions are based on a small sample) d o not discriminate based on what race of person is speaking the variety. For the Pre and Post Test, participants continued to respond faster to Stories Three (W AAE) and Four (W AE) when selecting a positive adjective. Additionally, for the Post Test par ticipants continued to dislike Story One (AA AAE) the most. For positive responses in the Post Test participants responded significantly faster to Stories Two (AA AE), Three (W AAE), and Four (W AE) than they did to Story One (AA AAE). Additionally, when participants chose a negative adjective in the Post Test they responded significantly slower to Stories Two (AA AE), Three (W AAE), and Four (W AE) than they did to Story One (AA AAE). When participants responded positively to Stories One (AA AAE) and Tw o (AA AE) in the Post Test, they were significantly faster in doing so than they were in the Pre Test. In terms of negative adjective selection, response times were not found to be significantly different from the Pre to the Post Test. Regarding demogr aphics, participants started with very few differences between each other, which means that in general, changes that are found in the Post Test can be ascribed to the lessons they were taught. Comparing the Pre and Post Test results there were no signifi cant differences found between the groups aside from a close
201 participants in the same way overall. Participants who taught students who were Hispanic plus another race, su ch as African American or Asian, did have significantly showed significant results. Before executing this research I thought that participants with African American stu dents would potentially show the most change because they would see the lessons as very relevant for themselves, however there were too few African American students to calculate this comparison by itself. Overall, I consider these findings very positive in terms of the research goals.
202 CHAPTER 7 CONCLUSIONS AND IMPLICATIONS Conclusions This research proved to be successful from both the Sociolinguistic and Psycholinguistic perspectives. In this concluding chapter I summarize the results that have been presented as well as discuss future implications. Summary of Sociolinguistic Results and Post Assessments grew in a positive direction at a statistically significant level for every question. Additionally, participants grew significantly for every question that I would have wanted them to, except for one. The question for which they did not grow significantly was whether they thought that African American English so unds as good as Academic English. That they did not adopt the belief that African American English sounds as good as Academic English could mean that spoken African American English will continue to be judged negatively even by people who have been expose d to the fact that it is a valid linguistic system. During the lessons I never expressed the idea that participants to know that African American English is a valid linguistic variety and that their students would benefit if they learned strategies to assist them in code switching. On the other hand, during the lessons the language was generally presented in writing, and it could be the case that the participants sim ply needed more auditory exposure to African American English before beginning to think that it sounds as good as Academic English.
203 The two questions for which they did not grow significantly were whether African American English should be taught in the schools and whether a goal of the American school system should be standardizing the English language, and I would not have wanted the participants to have adopted these beliefs over the period of our class sessions. African American English does not need to be taught in the schools because in many areas students already command it, and the purpose of the lessons was to ultimately enable students to code switch into Academic English. Additionally, whether schools should standardize English is not a topic we addressed and I did not express any preference about this idea. Another indicator of positive change is that by the end of the lessons participants were familiar with African American English as a rule governed system and they were to a much greater d egree able to translate phrases written in African American English into Academic English. This indicates that they might be more able to identify examples of betwe en African American English and Academic English in an effective way. class they were in, and all races grew significantly. White participants grew the most, and this is one of the more robust findings of the study, with White participants growing at t(1,21)=2.08, p=4.39E 18, or p<.00000000000000000439 on the Attitude Questionnaire for the measure of overall growth. In terms of the open ended Pre and Post End of Experiment re sponses, participants grew from believing that the idea of African American English is simply a governed system.
204 On the Post End of Experiment Response participants indicated that t should be part of mandatory professional development for teachers in New York City. The experiment was a success in their eyes and they appreciated it and felt like they benefited from the lessons, as some examples below indicate. throughout the duration of the lessons. At the beginning of the semester, especially after comple ting the Pre Psycholinguistic Experiment, the participants were somewhat wary of me. I had the impression that they were wondering what my motivation was in discussing the topic, and I thought that some participants wondered if I was racist and trying to disguise my racism through talking about language. The majority of the participants did not believe in the idea of African American English at the beginning of the semester, and for at least the first two class discussions I encountered a good deal of res istance to the idea that it is a rule governed, patterned variety. Starting around the third session participants seemed to begin to trust me more, and they began agreeing with the idea that they had heard this type of speech from their students and from simply traveling around the city. They were able to provide examples of African American English from conversations they either participated in or overheard, and their attitudes towards me shifted dramatically. As an example of their distrust for my mot ivations, on the open ended response sheet on which they were provided the opportunity to express any thoughts they had about the experiment, some indicated that they were concerned with my motivations
205 and insight into the larger implications of the experi ment. For example, two participants said: graduate student in NYC w ould have such limited perspectives on human The latter comment was one of the comments that struck me the most, as it made it clear that the participants were unaware of my motives and my knowledge about the nuances and intersections between lang uage and race. By the end of the sessions, this participant indicated on the response form that s/he still believed that the experiment was limited with the binary choices, however s/he did say: have been nice to spend I interpret this as a positive change of attitude, as the participant indicated that the class sessions were useful and that s/he would have enjoyed spending more time on the subject. By the time they completed the Post Test End of Experiment Response, almost every participant had something positive to say about the class sessions. Many of the participants declared that they thought that the lessons I delivered should be given to all teachers as required profess ional development. For example, among others: Development] for all teachers so at least they can teach students about their heritage and where this dialect came from and not to the strategies and activities discussed in my own classroom. All educators
206 English and this is something I can take and use in my career. There should be A linguistic course of this nature would prove to be beneficial to teachers, and the students The above comments, along with my impressionistic feelings ab out how the participants were responding to me (plus other comments that are too numerous to mention here), indicate that their attitudes towards me became very positive throughout the duration of the lessons. Summary of Psycholinguistic Results The purpos attitudes that they may not have realized that they had or that that might not want to display for social purposes. There was much significant change between the Pre and Post testing, whic h indicates that the six lessons I taught were successful. Overall, participants were slowest to respond positively and fastest to respond negatively to Story One (AA AAE) in the Pre Test. This pattern continued on the Post Test, however other measures i ndicating how comfortable participants were with the varieties moved in On the Pre Test participants preferred the White speaker over the African American speaker for every possible combination, though b y the Post Test the only significant story interaction was that participants significantly preferred the White African American English. I consider the fact that the rest of the interactions are not
207 significant to be a positive result because it indicates that the participants are not reacting as negatively to hearing an African American speaker or African American English as they did during the Pre Test. Also on the Pre Test the participants Post sh. Exploring the Sociolinguistic and Psycholinguistic Results Together While the Sociolinguistic and Psycholinguistic measures gathered somewhat different data, an underlying theme was whether participants felt positively or negatively towards the idea of African American English and acknowledging it as a valid variety. The results of all the measures together indicate that at the beginning of the semester the participants did not believe that African American English was an actual linguistic variety an d they did not like the way it sounded. On the Sociolinguistic measures they on almost every measure that it was not as valid as Academic English. Likewise, on the psycholinguistic measure, they showed through the speed of their responses and their adjective selection that they did not believe African American English to sound as pleasing or to be as acceptable of a variety. Nonetheless, after the six lessons, the results from both the sociolinguistic measures and the psycholinguistic measures indicate that the participants did grow to believe that African American English is an actual linguistic variety aside from slang and that it is worth valuing it as a variety. They also grew to believe that the speech samples from the African American speaker did not sound as negative as they did in the Pre Psycholinguistic measure.
208 Implicit Versus Explicit Measures A question I had was whether participants would respond in mo re socially desirable ways on the explicit, sociolinguistic measures than they would on the implicit, psycholinguistic measure. Although the two different methodologies of measuring attitudes and opinions were gathering different sets of data, I did not s ee a marked difference between the two different types of measurement. This may be in part due to the anonymity of the participants, since they were assigned numbers that they wrote on the forms instead of their names. If they had to write their names on the forms they may not have been as candid as they were. Participants were very clear and direct on the pre sociolinguistic measures indicating that they thought that African American English was an unacceptable way of speaking. While I do not have any way of knowing how they may have socially tuned their responses towards me, it seems to me that they expressed their true feelings in light of the fact that many of these feelings were very negative. For example, one participant said of the Psycholinguist ic Experiment: literate soundi
209 Certainly not all of the responses to the open ended questions were negative; however, I have highlighted several of the negative ones to demonstrate that I do not believe that the majority of the participants were attempting to please me as a researcher since I find their comments to be very negative and uninformed. Furthermore, in terms of the implicit measure, participants had more positive responses in terms of number of positive adjectives selected on the Post Test than on the Pre Test. This indicates that they were not completely tuning their beliefs towards me when answering the Pre Test measures. Additionally, because their responses were time locked, their potential desire to please me with positive selections was mit igated by the time in which it took them to make their decisions. Is New York Ready for African American English? As mentioned in Chapter 4, when I was attempting to find a White African American English speaking woman to record the stories, my advertiseme nts on Craigslist were repeatedly deleted. On the other hand, my advertisements for African American women who could speak African American English were not deleted. There is no way to determine who was flagging my advertisements and causing them to be d eleted; however, I assume that the individual or individuals were also seeing my advertisements for an African American woman because they were chronologically next to each other. I can never know why the advertisements kept being flagged, however I belie this was a somewhat unusual racial mix that I was discussing. When I began teaching deg ree of resistance from the participants. In particular, several openly expressed that they thought that the topic of the sessions was racist. I assume that whoever flagged
210 my advertisements and caused them to be deleted also assumed that my motivations w ere negative, although I will never have a way of knowing. Simply the fact that they were deleted, though, indicates the difficulty I encounter when broaching this subject, and it is also potentially a sign of why there are very few individuals or program s that exist to address the linguistic issues that were addressed throughout this research. It is a sensitive topic and one that is met with great resistance. American English, my res ponse would be that it is ready to address the topic and learn strategies to assist students to become effective code switchers, yet I do not believe results indicated that while the participants believed that African American English is an effective language for communication and that it is important for it to be accepted as a valid variety, they did not believe at significant levels that African American English sounds as good as Academic English. The participants are thus willing to accept that African American English is a legitimate, rule governed variety that has a place in willing t o say that it sounds as acceptable as Academic English. Certainly there was no context associated with this question on the Attitude Questionnaire, however on the open ended response sheet that participants were given, several did state that they believed that while valuing African American English was important, Academic English was nonetheless needed for academic and career purposes. During the lessons I emphasized that giving students access to Academic English would help them
211 academically, economicall y, and career wise, and these views were reflected back in the responses. For example: opening experience. I had no idea that African American English was an entirely different language with its own rules. Now that I know, I see it d ifferently. However, I think that African American English speakers should be made aware of and gradually taught to use the Academic English for formal purposes. This is not only because that is what is accepted in the society but also because we need a standard platform to weigh accomplishments and achievements in this never ending race. I really enjoyed these experiments and African American English. I st ill believe that in an academic setting, the use of mean that African American English is not accepted. To prepare for the future success of our students, we need to give our stude nts the highest academic standard, what educators would expect to see at each step of the education Charity (2008), Mallinson & Charity Hudley (2010), and Mallinson et al. (2011) scholars and practitioners across disciplines tackle questions about how the social They offer suggested ways for linguists to become a part of the community in which they live in order to share linguistic findings and apply what is known wit hin the field of what we know in academia and use it to benefit society. My current r esearch is in line with what Charity proposes. In this research I educated teachers or soon to be teachers about the linguistics of African American English and gave them specific skills
212 and strategies to use to help their students become adept at code sw itching. I have also founded a business, Linguistic Consulting, through which I give professional development sessions to teachers in the New York City area around issues of linguistic diversity. I want to ensure that all students have access to Academic English for academic, social, and economic mobility, and I want all students to be taught by teachers who understand and value linguistic diversity and see the interconnectedness between language and identity. Implications Because the experiment was so s uccessful on all fronts, additional research and their academic progress. It would be valuable to know whether the lessons I taught in a measurable way. Though that was outside the academic benefits as well as benefits to their self esteem in terms of language and identity. My larger hope is that th ey would also see benefits later in life with easier access to university admission and more success in career interviews. I am currently that I can determine how effec tive the teaching sessions are for the students themselves. As I mentioned in Chapter 2, I collected pilot data in a high school in Harlem, New York, and that data showed excellent results from the lessons I taught those teachers. I am continuing this wo rk and hope to affect much larger numbers of students through exploring language varieties, our ideologies towards these varieties, and ensuring that all students are given access to Academic English.
213 My hope in a wider scope of implications is that eventu ally all students will be taught by teachers who have been exposed to the idea that African American English or other non standard varieties are rule governed, logical linguistic varieties. In the movement to close the achievement and opportunity gaps I b elieve that the language students bring to school is a factor that is being somewhat overlooked, and that if we were to systematically educate all teachers about non academic achievement scores would improve. This would be be attitudes would become more positive, which has been demonstrated to positively affect achievement, and teachers would also have a repertoire of specific skills and strategies to use to enable their students to code switch between the ir home variety and Academic levels; however, language is one important facet, and I believe that placing more emphasis on this area would greatly benefit students. It is my goal to work with as many teachers as possible and to ultimately affect a large number of students across the country. Additionally, this work has notable implications in terms of whether it is valuable for mono dialectal speakers of African Ame rican English to acquire Academic English. The results from the experiment indicate that when interacting with people who have been trained in the linguistics of African American English, the acquisition of Academic English is beneficial. Certainly in ou r society most people have not been trained in the linguistics of African American English, and this research is limited to measuring the attitudes of people who were exposed to the linguistic rules and patterns. Nonetheless, current prevailing ideologies indicate that speaking Academic English is beneficial, and
214 most professional workplaces require the use of Academic English to a certain degree, or at the minimum during the job hiring process. As I reflect on the experiment and all of the informatio n gathered, I think back to strategies to teach my students ways to code switch. Now I do have the strategies, and attitudes towards the variety. This research wove several strands of research together, and it is my intent to continue in this line of research and work. Raising the achievement levels of students across the country is an important task, and examining the language that students are using and educating teachers about non standard varieties is one of the crucial steps on the path towards educational equity in the United States.
2 15 APPENDIX A PRE ASSESSMENT ART___________ Pre Assessment Date: ____________________________________ Professor: ________________________________ Class Meeting Time: ________________________ Directions: Please answer the questions to the best of your ability. 2. Are you familiar with the Oakland, California School Board Controversy of 1996? If so, please describe what you remember. 3. What does the following mean in African American English? John be running. Alexis been done ate.
216 4. Do you know of any ways to help children who speak African American English master Academic (Standard) English? If so, please describe. 5. Is Academic English better than African American English? Please describe.
217 APPENDIX B DEMOGRAPHICS QUESTIONNAIRE Demographic Questionnaire Do not put your name on this. ID Code :_____________________ Date: Professor: The following questions are needed for demographic information only and will not be associated with your name. Please take a moment to answer the following questions. 1. Gender: _________________________ 2. Age:_____________________________ 3. Where are you from? ___________________________ If you are from New York City, which borough? _____________________ 4. Your race/ethnicity: ______________________________ 5. Race/ethnicity of your Mother: _____________________________ 6. Race/ethnicity of your Father: ______________________________ 7. What was your average household income growing up? _______________________
218 8. What is your first language the language you grew up speaking at home? ________________ 9. What other languages do you speak, if any? __________________________________________ 10. Are you currently a teacher? ___________________ If so, what race/ethnicity are the majority of your students? ______________________ What language do they primarily speak? ______________________________ 11. How many years have you been teaching? _____________________________ 12. What grade level(s) do you teach? ______________________________ 13. Do you encounter language issues with your students, or any special language circumstances you would like to comment on? Please describe. Thank you for your responses. Your name will not be associated with them.
219 APPENDIX C ATTITUDE QUESTIONNAIRE ART:_________ Questionnaire Please circle the response that most closely reflects your opinion. Circle the response ABOVE the question. We are interested in your first response please do not deliberate over the questions. Strongly Some what Neutral Somewhat Strongly Agree Agree Disagree Disagree 1. The scholastic level of a school will fall if teachers allow African American English to be spoken. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 2. African American English is a misuse of Standard English. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 3. Attempts to eliminate African American English in school result in a situation that can be psychologically damaging to African American children. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 4. Continued usage of a nonstandard dialect of English would accomplish nothing worthwhile for society.
220 Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 5. African American English sounds as good as Standard English. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 6. Teachers should allow African American students to use African American Engl ish in the classroom. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 7. African American English should be discouraged. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 8. African American English must be accepted if pride is to develop among African American people. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 9. African American English is an inferior language system.
221 Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 10. African American English is cool. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 11. African American English should be considered a bad influence on American culture and civilization. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 12. African American English sounds sloppy. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Dis agree 13. If use of African American English were encouraged, speakers of African American English would be more motivated to achieve academically. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 14. African American English is a clear, thoughtful, and expressive language.
222 Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 15. African American English has a faulty grammar system. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 16. When teachers reject the native language of a student, they do him/her great harm. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 18. In a predominantly African American school, African American English as well as Standard English should be taught. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 19. Widespread acceptance of African American English is imperative.
223 Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 20. The sooner we eliminate nonstandard dialects of English, the better. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 21. Acceptance of nonstandard dialects of English by teachers will lead to a lowering of standards in schools. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 22. Nonstandard English should be accepted socially. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 23. Nonstandard English is as effective for communication as is Standard English. Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 24. One of the goals of the American school system should be the standardization of the English language.
224 Strongly Somewhat Neutral Somewhat Strongly Agree Agree Disagree Disagree 25. One successful method for improving the learning capacity of speakers of African American English would be to replace their dialect with Standard English.
225 APPENDIX D END OF TASK RESPONSE (AFTER PRE PSYCHOLINGUISTIC EXPERIMENT) ART:_________ End of Task Response Do you have any further comments on the experiment? If so, please describe.
226 APPENDIX E END OF TASK RESPONSE (AFTER POST PSYCHOLINGUISTIC EXPERIMENT) ART:_______ End of Experiment/Class Sessions Response Do you have any further comments on the Experiment or the Class Sessions? If so, please describe.
227 APPENDIX F POST ASSESSMENT ART___________ Post Assessment Date: _____________________________________ Professor: ________________________________ Class Meeting Time: _____________________ Directions: Please answer the questions to the best of your ability. 2. Are you familiar with the Oakland, California School Board Controversy of 1996? If so, please describe what you remember. 3. What does the following mean in African American English? John be running. Alexis been done ate.
228 4. Do you know of any ways to help children who speak African American English master Academic (Standard) English? If so, please describe. 5. Is Academic English better than African American English? Please describe. 6. Have you implemented any of the suggested linguistic activities? If so, which ones, how, and what were the results?
229 APPENDIX G RENDITION: Yesterday I was my bruvver close. The subsequent exchange between student and teacher sounds something like this: washed .] S: Wash. T: No. Look at it again. What letters do you se remember what we say when we see those letters on the end of a word? T: OK, but in this case we say wash ed Can you say that? S: Wash ed T: Good. Now read it again. S: Yesterday I wash ed brother .] S: Bruvver. T: No. Look at these letters in the middle. [Points to th .] Remember to read what you see. Do you remember how we say that sound? Put your tongue between your teeth and say / th/
230 LIST OF REFERENCES Adger, C.T. (2005). Language varieties in the school curriculum: Where do they belong and how will they get there? In Ramirez, J. D., Wiley, T.G, de Klerk, G., Lee, E., & Wright, W.E. (Eds.), Ebonics: The Urban Education Debate (2 nd Edition, pp. 96 108). Clevedon, UK: Multilingual Matters LID. Au, K. H. (1980). Participation structures in a reading lesson with Hawaiian children: Analysis of a culturally appropriate instructional event. Anthropology and Education Quarterly, 11 91 115. Ball, A., & Lardner, T. (1997). Dispositions toward language: Constructs of teacher knowledge and the Ann Arbor Black English case. College Composition and Communication, 48 469 485. Baugh, J. (1995). The law, linguist ics, and education: Educational reform for African American language minority students. Linguistics and Education, 7 87 106. Bell, Y., & Clark, T. (1998). Culturally relevant reading material as related to comprehension and recall in African American c hildren. Journal of Black Psychology, 24 455 475. Blair, I. V., Ma, J. E., & Lenton, A. P. (2001). Imagining stereotypes away: The moderation of implicit stereotypes through mental imagery. Journal of Personality and Social Psychology, 81 828 841. Blake, R., & Cutler, C. (2003). African American English and variation in teachers' attitudes: A question of the school philosophy? Linguistics and Education, 14 163 194. Bowie, R. L., & Bond, C. L. (1994). Influencing future teachers' attitudes towar d Black English: Are we making a difference? Journal of Teacher Education, 45 112 118. Brehm, J. W. (1966). A theory of psychological reactance New York: Academic Press. Brice Heath, S. (2009). Ways with words: Language, life, and work in commun ities and classrooms Cambridge, UK: Cambridge University Press. Casesnoves Ferrer, R., & Sankoff, D. (2004). The Valencian revival: Why usage lags behind competence. Language in Society, 33 1 31. 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 1340 1356.
231 Charity, A. H. (2008). Linguists as agents for social change. Language and Linguistics Compass 2 923 939. Cole, P.A., & Taylor, O. L. (1990). Performance of working class African American children on three tests of articulation. Language, Speech, and Hearing Services in Schools, 21 171 176. Da Silva, C. D., Huguley, J. P., Kakli, Z. & Rao, R. (2007). The opportunity gap: Achievement and inquality in education Cambridge, MA: Harvard Educational Review. Delpit, L. (1995). Other people's children: Cultural conflict in the classroom New York, NY: Norton. Delpit, L. (2008). No kinda sense. In Delpit, L., & Kilgour Dowdy, J. (Eds.), The skin that we speak: Thoughts on language and culture in the classroom (pp. 31 48). New York, NY: The New Press. Fasold, R. (1999). Ebonic need not be english. ERIC Clearinghouse on Languages and Linguistics, Washington, DC, 1 3. Fishman, J. (1977). Language, Ethnicity, and Racism. In Saville Troike, M. (Ed.), Georgetown University Round Table on Languages and Linguistics (pp. 297 309). Georgetown, VA: Georg etown University Press. Fogel, H., & Ehri, L. C. (2000). Teaching elementary students who speak Black English Vernacular to write in Standard English: Effects of dialect transformation practice. Contemporary Educational Psychology, 25 212 235. Fordham S. (1999). Dissin "the standard": Ebonics and guerilla warfare at Capital High. Anthropology and Education Quarterly, 30 272 293. Fordham, S., & Ogbu, J. (1986). Black students' school success: Coping with the "burden of 'acting white.'" Urban Review 18 176 206. Gawronski, B. (2007). Attitudes can be measured! But what is an attitude? Social Cognition, 25 (5), 573 581. Gawronski, B., & Bodenhausen, G. V. (2006). Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change. Psychological Bulletin,132 692 731. Gawronski, B., Deutsch, R., Mbirkou, S., Seibt, B., & Strack, F. (2008). When 'just say no' is not enough: Affirmation versus negation tr aining and the reduction of automatic stereotype activation. Journal of Experimental Social Psychology, 44 370 377.
232 Gawronski, B., Hofmann, W., & Wilbur, C. J. (2006). Are 'implicit' attitudes unconscious? Consciousness and Cognition: An International Journal, 15 485 499. Goodman, K. S. (1969). Dialect Barriers To Reading Comprehension. In Baratz, J. C., & Shuy, R. S (Eds), Teaching Black children to read ( pp. 14 28). Washington, DC: Center for Applied Linguistics. Gopaul McNicol, S., Reid G., & Wisdom, C. (1998). The psychoeducational assessment of Ebonics speakers: Issues and challenges. The Journal of Negro Education, 67 16 24. Green, L. (2002). A descriptive study of African American English: Research in linguistics and educati on. International Journal of Qualitative Studies in Education, 15 673 690. Green, L. (2007). African American English: A linguistic introduction Cambridge, UK: Cambridge University Press. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74 1464 1480. Harper, F. D., Braithwaite, K., & LaGrange, R. D. (1998). Ebonics and academi c achievement: The role of the counselor. Journal of Negro Education, 67 (1), 25 34. Heller, M., & Barker, G. Conversational strategies and contexts for talk: Learning activities for Franco Ontarian minority schools. Anthropology and Education Quarte rly, 19, 20 47. modification on memory and liking for the person described Journal of Experimental Social Psychology, 14 363 378. Hollie, S. (2001). Acknowledging the la nguage of African American students: Instructional strategies. English Journal, 90 54 59. Topic associating and topic centered styles. Linguistics and Education, 6 1 21 152. Labov, W. (1972). Language in the inner city Philadelphia, PA: University of Philadelphia Press.
233 Labov, W. (1982). Objectivity and commitment in linguistic science: The case of the Black English Trial in Ann Arbor. Language in Society, 11 165 201. Labov, W. (1995). Can reading failure be reversed: A linguistic approach to the question. In V. L. Gadsden & D. A. Wagner (Eds.), Literacy Among African American Youth (pp. 39 68). Cresskill, NJ: Hampton Press, INC. Ladson Billings, G (2000). Fighting for our lives: Preparing teachers to teach African American students. Journal of Teacher Education, 51 206 214. Ladson Billings, G. (2004). It's not the culture of poverty, it's the poverty of culture: The problem with teacher educat ion. Anthropology and Education Quarterly, 37 104 109. Lai, M. L. (2007). Exploring language stereotypes in post colonial Hong Kong through the matched guise test. Journal of Asian Pacific Communication, 17 225 244. Lambert, Wallace E. (1967). A social psychology of bilingualism. Journal of Social Issues, 23 91 150. Lowery, B. S., Hardin, C. D., & Sinclair, S. (2001). Social influence effects on automatic racial prejudice. Journal of Personality and Social Psychology, 81 842 855. MacFarlane, A., & Stuart Lingua, 122 764 778. Maddahian, E. & Sandamela, A. P. (2000). Academic English mastery program: 199 8 1999 evaluation report Los Angeles Unified School District, CA: Research and evaluation branch. Mallinson, C., & Charity Hudley, A. H. (2010). Communicating about communication: Multidisciplinary approaches to educating educators about language variation. Language and Linguistics Compass 4 245 257. Mallinson, C., Charity Hudley, A., Rutter Strickling, L., & Figa, M. (2011). A conceptual framework for promoting linguistic and educational change. Language and Linguistics Compass, 5 441 45 3. to literacy. Language in Society, 10 423 442. Monteith, M. J., Sherman, J. W., & Devine, P. G. (1998). Suppression as a stereotype control strategy. Persona lity and Social Psychology Review, 2 63 82.
234 National Center for Education Statistics. (2010). Status and Trends in the Education of Racial and Ethnic Minorities Retrieved October 17, 2012, from http://nces.ed.gov/pubs2010/2010015/indicator3_11.asp Newell, C. R. & Chambers, W. J. (1982). Meeting the educational needs of students who speak a vernacular English in the public sch ool setting A conference report. Washington, D.C.: National Institute of Education. Ogbu, J. (1999). Beyond language: Ebonics, proper English, and identity in a Black American speech community. American Educational Research Journal, 36 147 184. Pa ndey, A. (2000). TOEFL to the test: Are monodialectal AAL speakers similar to ESL students? World Englishes, 19 89 106. Gore, J. C., & Banaji, M. R. (2000). Performan ce on indirect measures of race evaluation predicts amygdala activation. Journal of Cognitive Neuroscience, 12 729 738. Piestrup, A. M. (1973). Black Dialect interference and accommodation of reading instruction in first grade (Monographs of the La nguage Behavior Research Laboratory No. 4) Berkeley, CA : University of California. Plant, E. A., & Devine, P. G. (1998). Internal and external motivation to respond without prejudice. Journal of Personality and Social Psychology, 75 811 832. Plant, E. A., & Devine, P. G. (2001). Responses to other imposed pro Black pressure: Acceptance or backlash? Journal of Experimental Social Psychology, 37 486 501. Plant, E. A., Devine, P. G., & Brazy, P. C. (2003). The bogus pipeline and motivations to respond without prejudice: Revisiting the fading and faking of racial prejudice. Group Processes & Intergroup Relations, 6 187 200. Purnell, T., Idsardi, W., & Baugh, J. (1999). Perceptual and phonetic experiments on American English dialect ident ification. Journal of Language and Social Psychology, 18 10 30. Rickford, J.R. (1997). Unequal partnership: Sociolinguistics and the African American Speech Community. Language in Society, 26 161 197. Rickford, J. R., & Rickford, A. E. (1995). Dialect readers revisited. Linguistics and Education, 7 107 128.
235 Rickford, J. R. (1999). African American Vernacular English: Features, evolution, educational implications Oxford, UK: Blackwell. Rickford, J.R. (1999). The Ebonics controversy in my experiences and reflections. Journal of Sociolinguistics, 3 267 275. Rickford, J. R. (2005). Using the Vernacular to teach the Standard. In Ramirez, J. D., Wiley, T.G, de Klerk, G., Lee, E., & Wright, W.E. (Eds.), Eboni cs: The Urban Education Debate (2 nd Edition, pp. 18 40). Clevedon, UK: Multilingual Matters LID. automatic biases: The malleability of implicit prejudice and stereotypes. Journal of P ersonality and Social Psychology, 81 856 868. Schmidt, R. W. (1990). The role of consciousness in second language learning. Applied Linguistics, 11 129 158. Shelton, J. N., & Richeson, J. A. (2005). Intergroup contact and pluralistic ignorance. J ournal of Personality and Social Psychology, 88 91 107. Siegel, J. (1999). Stigmatized and standardized varieties in the classroom : Interference or separation? TESOL Quarterly, 33 701 728. Singleton, G. E., & Linton, C. (2006). Courageous conve rsations about race: A field guide for achieveing equity in schools Thousand Oaks, CA: Corwin Press. Smitherman, G. (1971). Black idiom. Negro American literature forum, 5 88 91 & 115 117. Smitherman, G. (1972). English teacher, why you be do English Journal, 61 59 65. Smitherman, G. (1983). Language and liberation. The journal of Negro education, 52, 15 23. TESOL Quarterly, 32 139 143. Stewart, W. A. (1969). On the use of Negro dialect in the teaching of reading. In Baratz, J. C. & Shuy, R. W. (Eds), Teaching Black children to read ( pp. 156 219). Washington DC: Center for Applied Linguistics. Taylor, H. U. (1991). Stan dard English, Black English, and bidialectalism New York, NY: Peter Lang.
236 measured by the language attitude scale. In Shuy, R. W., Fasold, R. W. (Eds), Language at titudes: Current trends and prospects Washington, D.C.: Georgetown University Press. Terry, N. P. (2006). Relations between dialect variation, grammar, and early spelling skills. Reading and Writing, 19 907 931. Wiley, T. G., & Lukes, M. (1996). English Only and Standard English ideologies in the U.S. TESOL Quarterly, 30 511 535. Wolfram, W. (1970). Sociolinguistic alternatives for teaching reading to speakers of Non Standard English. Reading Research Quarterly, 6, 9 33. Wolfram, W. (1998). Language ideology and dialect: Understanding the Ebonics controversy. Journal of English Linguistics, 26 108 121. we speak: Thoughts on langua ge and culture in the classroom (pp. 203 219). New York, NY: The New Press.
237 BIOGRAPHICAL SKETCH Caroline Kennelly Latterman holds a PhD in linguistics from the University of Florida, a MSEd with a specialization in intercultural communication from the University of Pennsylvania, and a BA in linguistics and languages (French, Spanish) from Haverford and Swarthmore Colleges. She specializes in sociolinguistics and psycholinguistics and applies her academic background to the educational and corporate fie lds. She founded and owns a consulting firm, Linguistic Consulting (www.linguistic consulting.com), and she works with schools and businesses in a variety of linguistic capacities in New York City as well as nationally and internationally.