Using the Vernacular to Teach the Standard1

Dr. John R. Rickford
Department of Linguistics
Stanford University
Stanford, CA 94305-2150

Introduction. I want to begin by congratulating California State University at Long Beach for holding this conference, particularly in a climate in which there is so much ignorance and hostility. I am especially glad to see so many Latino scholars here, indicating that an interest in the language and education of African American students does not imply any lack of interest in the language and education of Latino students, or any other population, for that matter. I begin with these remarks because the Director of an educational program in Northern California had invited me to speak on Ebonics recently, then wrote a little later to withdraw the invitation, because one or two of her Latino Board members felt that attention to Ebonics might detract from attention to (and funding for) bilingual education.

In response to that understandable but misplaced concern, I wrote to tell her what my friend Dr. Geneva Smitherman has also been emphasizing recently: A rising tide lifts all boats. We should not be squabbling over crumbs from the table. The needs of various groups of students in our schools are similar in some ways, but different in others. We should be concerned about the success of ALL students and work together to provide each group with the resources it needs to maximize its chances of success in school and life.

In early 1997, I was corresponding often with Don Trujillo, who was at the time a policy representative in Sacramento. He sent me lots of information on California Senate Bill 205, the so-called "Education: Equality in English Instruction Act." Had it been successful, this bill would have wiped out the Standard English Proficiency Program [SEP], which is specifically designed to improve the Standard English skills of Ebonics speakers. This would have been a devastating blow not only for schools in the Oakland area, but throughout the state.2 Trujillo also sent me information on California Assembly Bill 36, which would have gutted bilingual education in California of a lot of its key features, but failed to pass out of committee on April 23, 1997. People like Trujillo are commendable for their openness to and concern for the language-related educational challenges of ALL students of color--ALL children, regardless of ethnicity, for that matter--and your presence here indicates that you are like him in this respect. By contrast, California State Assemblywoman Diane Martinez successfully introduced on February 28, 1997 Assembly Bill 1206, which "prohibits school districts from utilizing, as part of a bilingual education program, state funds ore resources for the purpose of recognition of, or instruction in, any dialect, idiom, or language derived from English." This bill was clearly aimed at forestalling any attempt to use bilingual education funds for speakers of Ebonics or African American English, and it was eventually approved and signed into law. It represents that defensiveness and terror in the ranks which caused that Northern California Director to withdraw my invitation to speak on Ebonics. Hopefully we can dispel that unnecessary defensiveness and fear, and pull together for the good of ALL the students in California and across the nation.

Let me go on now to explain my title, which is "Using the vernacular to teach the standard." Ninety percent of what was written and said in the media after the Oakland Ebonics resolution of December 1996 represented a misapprehension of the nature of the problem and the nature of the solution which Oakland was proposing. Most writers and commentators made a big fuss of emphasizing how important it was for children to learn standard English in this society. To this Oakland might simply have replied, "Yes, we agree. But what's next? HOW are we going to do it?"

How (badly) schools have failed to educate African American students. Oakland's original aim was to extend the Standard English Proficiency [SEP] program, which had been in place since 1981 throughout the state. That program has as its goal using the vernacular to teach the standard. I want to get that point straight at the beginning. I also want to begin where Oakland began, which is with the facts of massive educational failure within the African American community. The fact is that existing methods, throughout the country, are not working. The insinuation of the many vocal critics of Oakland's Ebonics resolution over the last four months was that Oakland's innovations were misplaced, but that the existing situation in the rest of America was JUST FINE, thank you. However, the fact of the matter is that the status quo with respect to the teaching of African American children in American elementary, middle and high schools is far from satisfactory. One of the tragedies of the media coverage of this Ebonics issue is that it never really got to the kinds of problems which started Oakland thinking about Ebonics and other solutions in the first place.

We've read already, of course, of the kinds of failures among African American students which are evident in the Oakland School District. You can read about those at the following Web Site:, and you can get access to that and other interesting web sites on the Ebonics issue through "Jacqueline's Ebonics Information Page" ( I do however want to point to a number of other examples across the country so you don't think this is just an Oakland or a California problem. You know how people from other parts of the county sometimes think, "Those folks in California are different, and kind of weird, anyhow."

What I want to do, first, is have you look at the test scores from Palo Alto and East Palo Alto ("Ravenswood" school district") shown in figure 1. The tall bars at the back here represent the children of Palo Alto, California. Palo Alto is right in the middle of Silicon Valley; and includes a lot of professors' kids, many of the kids of computer scientists and other highly educated professionals. Palo Alto has some of the best public schools in the country. Looking first at reading, the first two tall bars in the back, you'll

Figure 1: CAP Test Scores For Palo Alto And Ravenswood, 1990see that Palo Alto kids in the third grade score at the 96th percentile on the California Assessment Program test; and by the sixth grade, they score at the 99th percentile. Scoring at the 99th percentile means, of course, that they are better than 99% of kids in the state, that is, everybody else! In writing, they score at about the 94th percentile in the third grade, and by the sixth grade, they are at the 99th percentile and they continue like that. If they were ever to slip to the 92nd percentile, Palo Alto would have a big national conference to figure out what's going wrong.

Now step across the freeway to the Ravenswood School District in East Palo Alto. As figure 1 shows, the primarily African American and Latino kids here in the third grade score on the 16th percentile on the reading component of the California Assessment Program, but by the sixth grade, they've dropped to the 3rd percentile. (Statistics which I haven't included in this figure show that by the eighth grade, their reading scores have dropped even further, to the 2nd percentile.) If you look at writing, they score on the 21st percentile in the third grade, but by the sixth grade they're once again lower, at the 3rd percentile. This represents the regular pattern. Somehow the Palo Alto Schools are able to build on the skills and talents their primarily White children bring to the school and add value to them, so that very rapidly kids are performing at their maximum potential. Somehow schools in East Palo Alto, with African-American and other students of color, fail to do that, subtracting value instead: Kids come in with a certain level of achievement, and do steadily worse with each passing year. This is a forcible demonstration of the point which Claude Steele (1992:68), made in his important Atlantic Monthly article on "Race and the Schooling of Black Americans ": "The longer they [African Americans students] stay in school, the more they fall behind.

Lest you think that this is another weird California phenomenon, let us look at some recent data from predominantly African American schools in Philadelphia.3 In the 1995-96 school year, 41% of the students at one Elementary School (Birney) were reading at the Basic level or above as tested on the SAT-9, and the school's overall reading score was 56.9. At a high school in the same district, however (Benjamin Franklin), the percentage of students reading at or above the Basic Level was only 7.6%, and the overall reading score was 24.4. The 1996-97 statistics show a similar downward spiral, although the extent of the drop between the elementary and high school levels is smaller: 34.4 % of the students at Birney Elementary School read at or above the Basic Level, and the school's overall reading score was 52.7; at Benjamin Franklin High School, only 14% of the students read at the Basic Level or above, and the School's overall reading score was 41.9.4

More comprehensively, Michael Casserly, Executive Director of the "Council of Great City Schools," presented data before Senator Specter's US Senate Ebonics panel in January 1997 summarizing the performance of students in fifty large urban public school districts, including between them hundreds and hundreds of schools. Among other things, the data indicated that while White students in these schools show steady improvement in their reading achievement scores as they get older (60.7% read above the 50th percentile norm at the elementary school in 1992-93, and 65.4% did so by high school), African American students showed a steady decline (31.3% read above the 50th percentile norm at the elementary school level, but only 26.6% did so by high school). Moreover, data from the 1994 National Assessment of Educational Progress which he also presented show the same depressing trend in a different way: On a 500-point scale, African American students at the age of 9 are an average of 29 points behind the scores of their White counterparts; by the age of 13 they are 31 points behind; and by the age of 17, they are 37 points behind.

I cite these different data sets to make the point quite forcefully that whatever you may think of the Oakland School District and their Ebonics resolutions, the educational malaise of African American students in their District is very general across the United States, particularly in urban areas. Moreover the methods currently being used to teach Reading and the Language Arts to African American students--which the detractors of Oakland's Ebonics solution seem to be quite satisfied--are flat out NOT working.

Now clearly there are other factors that are involved in this kind of failure than language, or even methods of teaching reading Obviously, there are socioeconomic and class issues, and issues about the kinds of facilities which schools in primarily African American and White school districts tend to have. I was present at a meeting which the Rev. Jesse Jackson had with Board members of the Oakland Unified School District on December 30, 1996 (when he announced his revised position on their Ebonics resolution), and I was struck by his statement that the average US prison with large African American populations has better facilities than the average school with large African American populations. There's a frenzy of prison building, expansion and renovation across the country, as communities discover they're good business. There's not a similar frenzy of school building and improvement, so we should not be surprised at declining levels of school performance. And unfortunately, those who drop out of schools are more likely to end up in to prisons or otherwise fall into the clutches of the criminal "justice" system. As Jones (1995:9) has noted, drawing on a 1995 report by the Sentencing Project, a national non-profit organization, "one in three Black men between the ages of 20 and 29 are within the grasp of the criminal justice system."

There are also problems in terms of the kinds of teachers which most urban school districts are able to attract, and the training they have, and these problems are related to the fact that urban schools tend to pay lower salaries and have more challenging working conditions. There are also problems in terms of books and supplies. My wife Angela, a reading specialist, was doing a demonstration lesson in the teaching of reading recently at an urban school in the San Francisco Bay area, and she asked the teacher for a storybook to read to the class. The teacher said, "Story book?" She didn't have any! The classroom lacked the shelves and tables of gaily-colored and attention-grabbing story books which are customary in suburban schools. Luckily, one of the kids in the classroom had in her backpack a book she had happened to bring to school, and that was the book Angela used to demonstrate the teaching of reading. (By contrast, I recently visited Los Angeles schools participating in the Language Development Program for African American Students, run by Noma LeMoine, and I was impressed by the ready availability of books in each classroom, many of them about African Americans.) Finally, teachers in schools with primarily African American and other ethnic "minority" populations tend to have lower expectations for their students, and to ask less challenging questions, and the evidence is overwhelming (see Tauber 1996, A. Rickford 1998) that teacher expectations are closely tied to student achievement. If teachers expect you to do badly, you are more likely to do badly; and if they expect you to do well, you are more likely to do well.

The relevance of Ebonics. While factors like facilities, supplies, teacher pay and training, teacher expectations, parental involvement and others are indisputably relevant, and I would add my voice to those of others urging that these receive greater attention (see Irvine 1990, Comer 1993, 1997, Cose 1997), I would strongly dispute the claim of Ellis Cose (in Newsweek, January 13, 1997, p.80) that Ebonics--the language which many African Americans bring to school--is "irrelevant."

On theoretical grounds alone we would assume that the language of African American kids plays SOME role in the level of success they achieve in school, since language is so closely connected with cognitive abilities and with performance in other school subjects. As we know, kids who do well in English tend to do well in a variety of subjects across the curriculum, and those who don't do well in English, don't do well in most other subjects either.

But there is empirical evidence that language might be related. We know, for instance, that most of the kids who fall behind in reading and otherwise fail in inner city schools (see above) are Working Class kids, rather than Middle-Class kids. And we know that the distinctive pronunciation and grammatical features of African-American Vernacular English or Ebonics are used most commonly by members of the Working and Lower Class. Consider table 1, which summarizes data from Wolfram's (1969) study of Detroit.5 Except for consonant cluster simplification and absence of plural -s, every other Ebonics feature in that table is far more frequent among the Working Class groups than among the Middle Class groups; for instance. the Lower Working Class uses multiple negation 78% of the time, while the Upper Middle Class does so only 8% of the time.


FEATURE Lower Upper Lower Upper

Working Class Middle Class

Consonant cluster simplification NOT in past tense (p 60) 84% 79% 66% 51%

Voiceless th [O] --> f, t or Ø (p. 84) 71% 59% 17% 12%

Multiple negation (p. 156) 78% 55% 12% 8%

Absence of copula/auxiliary is, are (p. 169) 57% 37% 11% 5%

Absence of third person present tense -s (p. 136) 71% 57% 10% 1%

Absence of possessive -s (p. 141) 27% 25% 6% 0%

Absence of plural -s (p. 143) 6% 4% 1% 0%


TABLE 1: Use of selected AAVE features in Detroit, by social class (from Wolfram 1969)Actually, the Detroit figures for working-class Ebonics usage are kind of tame compared to the data we have from East Palo Alto (source of the Ravenswood school district figures in figure 1). In the latter community we have recorded working class teenagers (see Rickford 1992) with copula absence figures of 81% and 90%, compared with 57% and 37% in Wolfram's Detroit study, and with third singular present tense -s absence of 96% and 97%, compared with 71% and 57% in Wolfram's Detroit study. So there's definitely a socioeconomic class boundary which operates with respect to Ebonics usage,6 and the fact that Working and Lower Class African American kids tend to do worse in school than their Middle Class counterparts may well be related to differences in their language use, or to teacher's attitudes and responses to their language use.

The relevance of negative teacher attitudes to Ebonics was a key element in the 1979 ruling of Justice Joiner that the Ann Arbor, Michigan school district had failed to take adequate measures to overcome the barriers to equal education posed by the language of the African American children at Martin Luther King Jr. elementary school (Smitherman 1981, Labov 1982). But the evidence concerning negative teacher attitudes and responses to the vernacular of African American children had existed even earlier. Williams (1976) reported from a series of experiments that there were regular correlations between teachers' assessment of the relative "standardness" and "ethnicity" of students' speech and their ratings of the children's status, and their confidence or eagerness: Kids who sounded more non-standard and/or non-White were also rated as being less promising or effective students. What was worse, Williams and his associates also found in a separate experiment that prospective elementary teachers' perceptions of the relative standardness of children's speech were also affected by the children's race; the same sound track, when accompanying a videotape of an African American or Mexican American child, was rated as less standard than when accompanying a videotape a White child (Williams 1976:105). So students of color got a double whammy negative effect in terms of how teachers perceived and evaluated them in terms of race and language.

We got an even more powerful demonstration of the relevance and role of children's language--and how teachers respond to it in school--in Piestrup's (1973) study of over two hundred first-graders in predominantly African American classrooms in Oakland, California. One of the things she found is that there is a very strong inverse correlation between reading score and vernacular dialect score: The lower your dialect score, that is, the less of the vernacular you use, the higher your reading score, that is, the better you do on standardized tests of reading. This is interesting, but not unexpected, given what we know of the relationship between vernacular English usage and other factors like socioeconomic background which themselves correlate with school success. More interesting, because less well-documented, is the relationship she found between children's reading scores and the different ways in which teachers responded to the vernacular in the classroom. In what Piestrup dubbed the "Black Artful" approach, (Piestrup, 131) teachers

"used rhythmic play in instruction and encouraged students to participate by listening to their responses . . . attended to vocabulary differences and seemed to prevent structural conflict by teaching children to listen to standard English sound distinctions. Children taught with this approach participated enthusiastically with the teacher in learning to read."

By contrast, teachers using the "Interrupting" approach (ibid.) "asked children to repeat words that were pronounced in dialect many times and interpreted dialect pronunciations as reading errors. Teachers in this group presented standard English sounds for discrimination without ensuring accuracy of response." Some children taught by the Interrupting Approach (131-2) "tediously worked alone at decoding without reading as if they understood; others seemed to guess at almost as many words as they were able to read. Some children withdrew from participation in reading, speaking softly, and as seldom as possible." The latter result was not surprising, because each time they opened their mouths, they were met with rebuke, reprimand, or correction.

Figure 2 shows more concretely the difference between these two approaches (and four other approaches which we don't have time to consider) in terms of their correlations with dialect and reading scores. Note that children taught by the Black Artful teachers had higher reading scores overall than children taught by the Interrupting teachers. Moreover, if you look at the slopes for the two groups of teachers (lines 5 and 6), you'll see that the kids with the highest dialect scores (i.e., who spoke the most dialect), when taught by the Artful approach, read about as well as the kids with the lowest dialect scores (i.e., who spoke the least dialect) when taught by the Interrupting teacher. This is very clear evidence that the way in which teachers respond to and build on the vernacular can have a powerful effect on the level of success in reading which African American children attain.

The sad fact, however, is that most teachers do NOT build artfully and skillfully on the vernacular. And most members of the public support them in this. In the hue and cry of the Ebonics controversy in December 1996 and the first few months of 1997 the predominant public response was: "Stamp out Ebonics, or if you can't do that, ignore it, leave it alone, and hope and pray that it will go away. Bury your head in the sand; cover your ears with mufflers. Hear nothing. Don't let that virus anywhere near the classroom." The undeniable fact, however, is that most African American children come to school fluent in the vernacular. It WILL emerge in the classroom, and HOW teachers respond to it can crucially affect how the students learn to read, and how well they master Standard English. Ignoring or condemning the vernacular is not a particularly successful strategy, as shown in Piestrup's study, and as suggested by the massive educational failure associated with this approach nationwide.

If you asked then, "How might the vernacular of African American children be taken into account in efforts to help them do better in schools?" I would say that there are basically three different approaches.

Figure 2: Correlation between reading scores, dialect scores, and teaching strategies (Piestrup 1973: 162)1. The linguistically informed approach. The first is what I call the "linguistically informed' approach. This encompasses the specific suggestions made by Labov (1995) based on decades of research on Ebonics or African American Vernacular English. One of these is that teachers should "distinguish between mistakes in reading and differences in pronunciation"; so the kid who reads "I missed him" as "I miss him" should not automatically be assumed to have misread, in the sense of not being able to decode the letters. On the contrary, he may have decoded the meaning of this Standard English sentence correctly, but he may then have reproduced its meaning according to the pronunciation patterns of his dialect, in which a consonant cluster like [st]--the final sounds in "missed"--is often simplified to [s]. Labov (ibid.) also suggests giving more attention to the ends of words, where Ebonics pronunciation patterns have a greater modifying effect on Standard English words than they do at the beginnings of words. He also suggests that words be presented in contexts that preserve underlying forms, for instance, words that are followed by a vowel which favors retention of final consonants--testing or test of--rather than test in isolation. He also suggests using the full forms of auxiliary verbs (e.g. "He will be here," He is tall") and avoiding contractions (e.g. "He'll be here," "He's tall"), because of evidence that once you go through a contraction stage, Ebonics is much more likely to proceed to deletion ("He Ø be here," "He Ø tall"). These are sound ideas that should not be terribly controversial, but how much of an impact they will make on reading instruction for African American kids is not yet clear, since no one has systematically implemented them or assessed their effects.

2. Contrastive Analysis. The second approach is to do some form of Contrastive Analysis where you draw students' attention specifically to the differences between the vernacular and the standard language. One of the best examples of this was some work that was done by Hanni Taylor (1989). She's at Aurora University, just outside Chicago, and she was faced with a number of students from inner city Chicago who used a lot of Ebonics features in their standard English writing. She divided her students into two groups. With the control group, she used conventional techniques of teaching English and made no reference to the vernacular . But with the other, experimental group, she used Contrastive Analysis, specifically drawing their attention to the points on which Ebonics and Standard English were different. What she found after eleven weeks was that the kids who were using traditional techniques showed an 8.5 percent increase in their use of Ebonics speech in their writing while the kids who had benefited from Contrastive Analysis showed a 59 percent decrease in their use of Ebonics features in their writing. This is a very dramatic demonstration of the fact that even if we agree with the pundits across the country that you want kids to increase their mastery of Standard English, the Contrastive Analysis approach --essentially what Oakland wanted to do--is more likely to be successful than the conventional approaches that are currently being used. If I can give a very specific example, one of the features that she looked at was third person -s absence, as in "He walkØ" , instead of "He walks.". Taylor found that students taught by traditional techniques did show a small reduction (-11%) in the use of this feature over the course of eleven weeks, but the kids who were taught by Contrastive Analysis showed a massive decrease in the use of this feature (91.7). The point Taylor made overall is that this process of comparing the two varieties seems to lead to much greater metalinguistic awareness of similarities and differences between the vernacular and the standard and allows kids to much more effectively negotiate the line between the two.

There are at least two other instances in which this approach has been successfully used to help Ebonics speakers improve in Standard English and reading. Parker and Christ (1995)--both extol the virtues of the Bi-dialectal Contrastive Analysis approach in teaching minorities to play the corporate language game. In this approach you try to respect the home variety of the kids and help them negotiate between that variety and the standard language, teaching them about appropriate contexts for different varieties of speech. They have used this approach successfully with vernacular speakers in Tennessee and Chicago at the preschool, elementary, high school and college levels. There's also a program which I just visited in DeKalb County, Georgia, just northeast of Atlanta; it's the brainchild of Kelli Wright Harris, and involves use of Contrastive Analysis to help fifth and sixth grade students switch between home speech and school speech. According to Cummings (1997), the program "has won a 'Center of Excellence' designation from the National Council of Teachers of English. Last year, students who had taken the course had improved verbal scores at every school" So we have evidence from these programs that Contrastive Analysis works.

3. Introducing reading in the vernacular, then switching to the standard. The last kind of approach I want to talk about is one in which you actually begin by teaching kids in the vernacular--introducing them to reading in the vernacular and then switching to the standard.7 This follows a principle that was established from research dating back to the 1950's. A classic work is Cheavens' (1957) dissertation on Vernacular Languages in Education. Cheavens reported on studies around the world which showed that when you began by teaching kids in their vernacular or native language before switching to a second language which was not their vernacular, they tended to do better than if you began by teaching them in that second language directly. One of the most dramatic examples was a study done between 1948 and 1954 in fourteen schools in Iloilo Province in the Philippines (see Orata 1953). In this study, half of the kids were taught completely in English for four grades while other kids were first taught for two years in Hiligaynon, their native Philippine language, and then switched to English. What the researchers found is what other researchers have found in many other studies, namely that the kids who began in their own vernacular, when they switched to the second language, very rapidly caught up with the kids who started in English, and even surpassed them. The kids who started in the vernacular were outperforming in English the kids who started in English, in subjects ranging from reading to social studies, and even arithmetic. This was a massive study done over a fairly long period of time.

The closest parallel to this in terms of the United States and Ebonics or African-American English, is the "Bridge" study reported on in Simpkins and Simpkins (1981). This was a study involving five hundred and forty kids in the United States in twenty-seven different schools in five different parts of the United States. Four hundred and seventeen of the kids were taught with an experimental series of "Bridge" readers which began with narratives and exercises written in Ebonics, went through a transitional series written in a variety intermediate between Ebonics and English, and ended up with a final series written entirely in Standard English . A control group of one hundred and twenty-three kids was taught entirely in Standard English using conventional methods, without the "Bridge" readers. What the researchers found, after four months of instruction and testing, is that the kids who were being taught by the conventional methods showed only 1.6 months of reading gain, which would be consistent with the evidence presented earlier that the longer African American kids stay in school with existing methods, the further they fall behind. By contrast, the kids that were being taught with the Bridge Readers showed 6.2 months of reading gain after four months of instruction. The experimental evidence was dramatically in support of the approach--the method offered the hope that African American kids would finally be able to read above and ahead of the norm rather than below it. But the inclusion of the vernacular in some of the "Bridge" readers elicited knee-jerk negative reactions similar to those which emerged in the Oakland Ebonics debacle of 1996. The publisher of this innovative series of readers, embarrassed by the negative reactions, quickly decided against continuing production of the "Bridge" series, and this very innovative and promising experiment came to an abrupt end despite its dramatically demonstrated pedagogical success.8

For many if not most of you, this kind of information about the positive effects of taking the vernacular into account in education is probably brand new, even though you may have followed media discussions of the Ebonics issue for months. That's in part because of what Noam Chomsky has called more generally the manufacturing of consent (see Achbar 1994), the manipulation of information by the media to present certain sides of issues and exclude others. In keeping with Chomsky's insistence that "the responsibility of intellectuals is to tell the truth and expose lies," several linguists (I know of Geoffrey Pullum, Salikoko Mufwene, and the film-maker Gene Searchinger besides myself) submitted Op Ed pieces on the Ebonics issue to major national newspapers like the New York Times, Washington Post, and the Los Angeles Times. We have all been declined. Some of us have managed to get our voices heard (see Rickford 1996) in less well-known papers like the San Jose Mercury News. But by and large it has been an uphill struggle to get anything like a pro-Ebonics or pro-vernacular perspective aired. Sometimes the newspapers will say, "Well, the issue is passé." But the next weekend you'll see another editorial or Op Ed ranting and raving about the horror that Ebonics represents or the wrongness of the Oakland resolutions, so it's clearly not the timeliness of the issue that's in question but the take on it which linguists represent. Having seen what the media do with an issue you know well really makes you wonder about their other coverage of other issues. You take on faith that they will follow certain principles of neutrality and objectivity and truth, but once you see the betrayal and misrepresentation and manufacturing of consent on one issue, it makes you question all the others.

Some Caribbean and European parallels. Lest I get entirely wrapped up in lambasting the media and forget my focus on education, let me give you some quick parallels from the Caribbean and Europe to give you a sense that the ways of taking the vernacular into account which I sketched out above are not completely novel. I am originally from the Caribbean, and we speak varieties of Creole English there that are very similar to African-American English in many respects, and in fact I have argued in a number of publications (see Rickford 1977, 1997) that there is a historical relation between these varieties. Way back in the 1950's, Robert Le Page, a well-known British linguist, after going to Jamaica and noticing the appalling failures in the teaching of English and other subjects in the public schools, proposed that the first year or two should be taught in Creole before standard English is introduced. One reporter in a local newspaper damned it as an insulting idea (cited in Cassidy 1970:208)--in fact, if you read some of the press coverage on this issue in Jamaica from the 1950s, it sounds like press coverage of Ebonics from the 1990s in California. But as Le Page (1968) argued, there was a problem with the teaching of English across the "English speaking" Caribbean: the percentage of students from each county who passed in English on the 1962 GCE "Ordinary" level exam ranged from 10.7% to 23.1%. Le Page argued that there was systematic interference in the students' English from the Creole which was not being recognized by the teachers or the educational system, and an approach that recognized and dealt with this interference would be more effective.

There was similar controversy in Trinidad in 1975, when a new English language curriculum that took Creole usage into effect was introduced (See Carrington and Borely 1977). More recently, people working with West Indian students in North American schools similarly feel the need to take their English creole vernaculars into account; the Toronto folks have been particularly innovative in this respect (see Coelho 1991).

In terms of the European scene, I will briefly refer you to two studies, although there are others that are relevant. The first is Osterberg's (1961) study of Swedish dialects and education. Osterberg conducted an experiment for a few years in which he began teaching one set of kids in their vernacular dialect of Swedish and then switching to standard Swedish. A second set of students was taught entirely in standard Swedish for the same period. As you may already have recognized, this was essentially a dialect version of the most famous sets of work that Cheavens had looked at earlier in terms of languages. Again, after 35 weeks, what Osterberg found was that the dialect methods showed itself superior--both in terms of reading quickly and rapidly assimilating new matter. The same positive results applied to reading and reading comprehension.

Later on a scholar named Tove Bull (see Bull 1990) did a similar study in Norway. She did this between 1980-1982 with ten classes of beginning students, with nearly 200 students about seven years old. Again, the same kind of design--dialect varieties of Norwegian versus standard Norwegian. The results showed that the vernacular children read significantly faster and better than the controlled subjects, and this was particularly true of the children who were doing worse to begin with. Bull attributed this in part to the same kinds of factors that Hanni Taylor talked about--that the vernacularization of the medium of initial teaching made students better able to analyze their own speech-- increasing and improving their metalinguistic consciousness, and it did so more than the principles according to which reading and writing were traditionally taught..

I could go on to cite studies, but my time is up. The point I want to make in closing is that there is much better evidence for the kind of approach that Oakland advocated than you might have supposed from reading newspapers and magazines and listening to your television and radio. In closing, I would like to turn on its head a comment which the Rev. Jessie Jackson made in his initial phase on the Ebonics issue, before he saw the light and changed his mind. He said at first that the kind of approach that Oakland was advocating represented "an unconditional surrender, borderlining on disgrace--an Unconditional Surrender Bordering on Disgrace. I would argue that to continue with existing methods in the light of the dramatic failure rates that are associated with those represents to me an unconditional surrender, bordering on disgrace.


1 This is a revised version (March 25, 1998) of remarks delivered at the California State University Long Beach [CSULB] Conference on Ebonics held on March 29, 1997, and will appear in the proceedings (ed. by Gerda DeKlerk), to be published by CSULB in 1998 I am grateful to the organizers, including Robert Berdan and Gerda DeKlerk, for inviting me to take part.

2 Fortunately, the bill was defeated in committee on April 2, 1997, although there have been subsequent attempts to resuscitate it in a significantly revised form. For further information on this and other California State or Assembly bills cited here, see, and consult Richardson (1998) for information on other legislative responses to the Ebonics controversy of 1996-97 at the state and federal levels.

3 These schools were deliberately picked to provide a comparison with data from the Philadelphia Inquirer of July 25, 1976 which were cited in Labov (1995).

4 One interesting aspect of the Philadelphia data for 1995-96 and 1996-97 is that the reading data from Cooke Middle School actually show an improvement over those from Birney Elementary School in terms of percentage reading at or above the Basic Level both years (47.5% and 40.1% respectively) although not in overall reading scores (53.1 and 51.2 respectively). This is somewhat encouraging since the 1976 data on reading and math combined which Labov (1995) cited show a steady and precipitious decline from the elementary level (31% of Birney students scored below the 16th percentile) through the middle school (50% of Cooke students scored below the 16th percentile) to the high school (75% of Franklin students scored below the 16th percentile).

5 Unfortunately, we don't have good class-based studies of African American communities beyond the 1960s; it is an area in urgent need of empirical research. I keep encouraging graduate students to do it, but they tend to be daunted by the time, effort and resources which a randomized study of class in an urban African American speech community would require.

6 The gap in Ebonics use between the working and middle class helps to explain the tremendous denial and condemnation evidenced by African-Americans in 1996 and 1997 in relation to Ebonics. By and large, the people that the media interviewed were not from the African American working and under-classes; Kweisi Mfume, Maya Angelou, Bill Cosby et al were very much upper middle class "representatives of the race," and what they had to say about Ebonics was decidedly influenced by their backgrounds.

7 Note that this is NOT the approach that the Oakland School Board advocated in 1996.

8 McWhorter (1997) has pointed to a series of studies done in the early 1970s in which "dialect readers were shown to have no effect whatsoever on African American students' reading scores." I think it is important to re-examine and even replicate those studies, but it should be noted that they all differ from the "Bridge" study insofar as they lacked any time depth. The studies cited by McWhorter were one-time studies of the effects of using vernacular or Standard English stimuli on decoding or reading comprehension in the relatively brief (e.g. 30 minute) session or sessions needed to conduct the experiment, rather than studies of the effects of teaching children in the vernacular or in Standard English over an extended period of time, as was the case with the "Bridge" study. This crucial difference may account for the success of the latter study and the failures of the earlier studies. This much is suggested by the authors of one of the most comprehensive earlier studies, Simons and Johnson (1974), who note (p. 355) that "Another limitation of the present study concerns the length of the experiment and the number of reading texts employed. It may be the case that the treatment may have been too brief to show a difference in reading."


Achbar, Mark, ed. 1994. Manufacturing consent : Noam Chomsky and the media : the companion book to the award-winning film by Peter Wintonick and Mark Achbar. Montreal ; New York : Black Rose Books, c1994.

Bull, Tove. 1990. Teaching School Beginners To Read and Write in the Vernacular. In Tromsø Linguistics in the Eighties 11:69-84. Oslo: Novus Press.

Carrington, L.D., and C.B. Borely. 1977. The Language Arts Syllabus, 1975: Comment and Counter Comment. St. Augustine, Trinidad: University of the West Indies.

Cassidy, Frederic G. 1970. Teaching Standard English to Speakers of Creole in Jamaica, West Indies. In James E. Alatis, ed., Report of the 20th Annual Round Table Meeting on Linguistics and Language Studies: Linguistics and the Teaching of Standard English to Speakers of Other Languages or Dialects, 203-214. Washington DC: Georgetown University Press.

Cheavens, Sam Frank. 1957. Vernacular languages and education. Ph.D. dissertation, University of Texas, Austin.

Coelho, Elizabeth. 1991. Caribbean students in Canadian schools, Book 2. Markham, Ontario: Pippin Publishing and the Caribbean Student Resource Book Committee.

Comer, James P. 1993. School power : implications of an intervention project, second edition. New York : Free Press.

_____. 1997. Waiting for a miracle : why schools can't solve our problems, and how we can. New York : Dutton.

Cose, Ellis. 1997. Color-blind : seeing beyond race in a race-obsessed world. New York: Harper Collins.

Cummings, Doug. 1997. A different approach to teaching language. The Atlanta Constitution, January 9, 1997, p. B1.

Irvine, Jacqueline Jordan. 1990. Black students and school failure: Policies, practices, and prescriptions. New York: Greenwood Press.

Jones, Charisse. 1995. Crack and punishment: Is race the issue? The New York Times, Saturday October 28, 1995, pp. 1, 9.

Labov, William. 1982. Objectivity and commitment in linguistic science: The case of the Black English trial in Ann Arbor. Language in Society 11:165-201.

_____. 1995. Can Reading Failure be Reversed: A Linguistic Approach to the Question. In V. Gadsden and D. Wagner (eds.), Literacy among African American youth, 39-68. Creskill, NJ: Hampton Press.

Le Page, Robert B. 1968. Problems to be faced in the use of English as a medium of education in four West Indian territories. In Joshua A. Fishman, Charles A. Ferguson, and Jyotirindra Das Gupta (eds.), Language problems of developing nations, 431-43. New York: John Wiley and Sons.

McWhorter, John. 1997. Wasting energy on an illusion: six months later. The Black Scholar 27.2:2-5.

Orata, Pedro T. 1953. The Iloilo experiment in education through the vernacular. In Unesco, The use of vernacular languages in education, 123-131. Paris.

Osterberg, Tore. 1961. Bilingualism and the First School Language--An Educational Problem Illustrated by Results from a Swedish Dialect Area. Umeå: Väster-bottens Tryckeri.

Parker, Henry H., and Marilyn I. Crist. 1995. Teaching Minorities to Play the Corporate Language Game. Columbia, South Carolina: National Resource Center for the Freshman Year Experience and Students in Transition, University of South Carolina.

Piestrup, Ann McCormick. 1973. Black Dialect Interference and Accommodation of Reading Instruction in First Grade. University of California, Berkeley: Monographs of the Language Behavior Research Laboratory, No. 4.

Richardson, Elaine. 1998. The Anti-Ebonics Movement: 'Standard English-Only.' Journal of English Linguistics 26.2.

Rickford, John R. 1977. The question of prior creolization in Black English. In Albert Valdman (ed.), Pidgin and creole linguistics, 190-221. Bloomington: Indiana University Press.

_____. 1992. Grammatical variation and divergence in Vernacular Black English. In Internal and external factors in syntactic change, ed. by Marinel Gerritsen and Dieter Stein, 175-200. Berlin and New York: Mouton.

_____. 1996. Ebonics succeeds where traditional methods do not. San Jose Mercury News, December 26, 1996, page 8B.

_____. 1997. Prior creolization of AAVE? Sociohistorical and textual evidence from the 17th and 18th centuries." Journal of Sociolinguistics 1.3:315-336.

Simons, Herbert D., and Kenneth R. Johnson. 1974. Black english syntax and reading interference. Research in the Teaching of English 8:339-358.

Simpkins, Gary A., and Charlesetta Simpkins. 1981. Cross cultural approach to curriculum development. In Geneva Smitherman (ed.), Black English and the education of Black children and youth: Proceedings of the national invitational symposium on the King decision, 221-40. Detroit: Center for Black Studies, Wayne State University.

Smitherman, Geneva, ed. 1981. Black English and the education of Black children and youth: Proceedings of the national invitational symposium on the King decision, 11-36. Detroit: Center for Black Studies, Wayne State University.

Steele, Claude. 1992. Race and the Schooling of Black Americans. The Atlantic Monthly, April: 68-78.

Tauber, Robert T. 1997. Self-fulfilling prophecy : a practical guide to its use in education. Westport, Conn. : Praeger.

Taylor, Hanni U. 1989. Standard English, Black English, and bidialectalism. New York: Peter Lang.

Williams, Frederick. 1976. Explorations of the linguistic attitudes of teachers. Rowley, Massachusetts: Newbury House.

Wolfram, Walt (1969). A linguistic description of Detroit Negro speech. Washington, DC: Center for Applied Linguistics.