Suite for Ebony and Phonics

Published in Discover magazine, December 1997
(Final version is slightly different, reflecting editor's changes)

To James Baldwin, writing in 1979, it was "this passion, this skill, ... this incredible music." Toni Morrison, two years later, was impressed by its "five present tenses," and felt that "The worst of all possible things that could happen would be to lose that language." What these African American novelists were talking about was Ebonics, the vernacular or informal speech of many African Americans, which rocketed to public attention after the Oakland School Board approved a resolution in December 1996 recognizing it as the primary language of African American students.

The reaction of most people across the country--in the media, at holiday gatherings, and on electronic bulletin boards--was overwhelmingly negative. In the flash-flood of email on America Online, Ebonics was variously described as "lazy English," "bastardized English," "poor grammar," and "fractured slang." Oakland's decision to recognize Ebonics and use it to facilitate mastery of Standard English [SE] also elicited superlatives of negativity: "ridiculous, ludicrous," "VERY, VERY STUPID," "a terrible mistake." Linguists--the scientists who carefully study the sounds, words, and grammars of languages and dialects--were less rhapsodic about Ebonics than the novelists, but much more positive than most of the media and the general public. At their January 1997 annual meeting, members of the Linguistic Society of America [LSA] unanimously approved a resolution describing Ebonics as "systematic and rule-governed like all natural speech varieties," and referring to the Oakland resolution as "linguistically and pedagogically sound." In order to understand how linguists could have had such a different take on the Ebonics issue, we need to understand how linguists study language and what their studies of Ebonics over the past thirty years have led them to agree on (and what it has not).

Although linguists approach the study of language from different perspectives--some are keener on language change, for instance, while others are more interested in language as a formal system, in what language tells us about human cognition, or how language reflects social divisions--we agree on a number of general principles. One of these is that linguistics is descriptive rather than prescriptive, our goal being to describe how language works rather than to prescribe how people should or shouldn't speak. A second principle is that all languages have dialects--regional or social varieties which develop when people are separated by geographical or social barriers and their languages change along different lines, as they develop their own pronunciations, for instance, or their own ways of referring to things. When linguists speak of "dialects" they don't do so in the pejorative way that many non-linguists do. A dialect is just a variety of a language; everyone speaks at least one. A third principle, vital for understanding linguists' reactions to the Ebonics controversy, is that all languages and dialects are systematic and rule-governed. To some extent, this is a theoretical assumption--for if individuals made up their own sounds and words and did NOT follow a common set of rules for putting them together to express meaning, they would be unable to communicate with each other, and children would have a hard time acquiring the "language" of their community. But it is also an empirical finding. Every human language and dialect which we have studied to date--and we have studied thousands--has been found to be fundamentally regular, although its rules may differ from those of other varieties.

Now is Ebonics just "slang," as so many people have characterized it? Well, no, because slang refers just to the vocabulary of a language or dialect, and even so, just to the small set of new and (usually) short-lived words like chillin ("relaxing") or homey ("close friend") which are used primarily by young people in informal contexts. Ebonics includes non-slang words like ashy (referring to the appearance of dry skin, especially in winter) which have been around for a while, and are used by people of all age groups. Ebonics also includes distinctive patterns of pronunciation and grammar, the elements of language on which linguists tend to concentrate because they are more systematic and deep-rooted.

But is Ebonics a different language from English or a different dialect of English? Linguists tend to sidestep questions like these, noting, as the 1997 LSA resolution did, that the answers often depend on sociohistorical and political considerations rather than on linguistic ones. For instance, spoken Cantonese and Mandarin are mutually unintelligible, but they are usually regarded as "dialects" of Chinese because their speakers use the same writing system and see themselves as part of a common Chinese tradition. By contrast, although Norwegian and Swedish share many words and their speakers can generally understand each other, they are usually regarded as different languages because they are the autonomous varieties of different political entities (Norway, Sweden). Despite this, most linguists might agree that Ebonics is more of a dialect of English than a separate language, insofar as it shares most of its vocabulary and many other features with other informal varietes of American English, and insofar as its speakers can understand and be understood by speakers of most other American English dialects.

At the same time, Ebonics is one of the most distinctive varieties of American English, differing from Standard English [SE]--the educated standard--in several ways. Consider, for instance, its verb tenses and aspects. ("Tense" refers to WHEN an event occurs, e.g. present or past, and "aspect" to HOW it occurs, e.g. habitually or not.) When Toni Morrison referred to the "five present tenses" of Ebonics, she didn't give examples, but it is probably usages like these--each one different from SE--which she had in mind:

1. Present progressive: He Ø runnin (=SE "He is running" or "He's running")

2. Present habitual progressive: He be runnin (=SE "He is usually running")

3. Present intensive habitual progressive: He be steady runnin (=SE"He is usually running in an intensive, sustained manner."

4. Present perfect progressive: He bin runnin (=SE "He has been running")

5. Present perfect progressive with remote inception: He BIN runnin (=SE "He has been running for a long time, and still is")

The distinction between events which are non-habitual or habitual, represented in 1 and 2 respectively by the non-use or use of an invariant be form, can only be expressed in SE with adverbs like "usually." Of course, SE can use simple present tense forms (e.g. "He runs") for habitual events, but then the meaning of an ongoing or progressive action signalled by the "-ing" suffix is lost. Note too that bin in 4 is unstressed, while BIN in 5 is stressed. The former can usually be understood by non-Ebonics speakers as equivalent to "has been" with the "has" deleted, but the stressed BIN form can be badly misunderstood. Years ago, I presented the Ebonics sentence "She BIN married" to twenty five Whites and twenty five Blacks from various parts of the US, and asked them, individually, if they understood the speaker to be still married or not. While almost all the Blacks (23, or 92%) said "Yes," only a third of the whites (8, or 32%) gave this correct answer. In real life, a misconstrual of this type could be disastrous!

OK, so it's not just slang, but an English dialect, sharing a lot with other English varieties, but with some pretty distinctive features of its own. What of characterizations of Ebonics as "lazy" English, as though it were the result of snoozing in a hammock on a Sunday afternoon, or the consequences of not knowing or caring about the rules of "proper" English? Well, if you remember the Linguistics principle that all languages are rule-governed, you'll probably be ready to reject these characterizations as a matter of general principle, but you can also challenge them on specific grounds.

One problem with statements like these is that they fail to recognize that most of the "rules" we follow in using language are below the level of consciousness, unlike the rules that we're taught in grammar books or at school. Take for instance, English plurals. Although grammar books tell us that you add "s" to a word to form a regular English plural, as in "cats" and "dogs," that's only true for writing. (Let's ignore words that end in s-like sounds, like "boss," which add "-es," and irregular plurals like "children.") In speech, what we actually add in the case of "cat" is an [s] sound, and in the case of "dog" we add [z]. (Linguists use square brackets to represent how words are pronounced rather than how they are spelled.) The difference is that [s] is voiceless, with the vocal cords in the larynx or voice box in our throats (the Adams apple) spread apart, and that [z] is voiced, with the vocal cords held closely together and noisily vibrating. You can hear the difference quite dramatically if you put your fingers in your ears and produce a "ssss" sequence followed by a "zzzz" sequence followed by a "ssss" sequence: sssszzzzssss. Everytime you switch to "zzzz" your voice box switches on (voiced), and everytime you switch to "ssss" your voice box switches off (voiceless). Now, how do you know whether to add [s] or [z] to form a plural when you're speaking? Easy. If the word ends in a voiceless consonant, like "t," add voiceless [s]. If the word ends in a voiced consonant, like "g," add voiced [z]. Since all vowels are voiced, if the word ends in a vowel, like "tree", add [z]. Because we spell both plural endings with "s," we're not aware that English speakers make this systematic difference everyday, and I'll bet your English teacher never told you about "voiced" [z] and "voiceless" [s]. But you follow the "rules" for using them anyway, and anyone who didn't--for instance, someone who said "book[z]"--would strike an English speaker as sounding funny.

One reason people might regard Ebonics as "lazy English" is its tendency to omit word-final consonants, especially if they come after another consonant, as in "tes(t)" and "han(d)." But if one were just being lazy or cussed or both, why not also leave out the final consonant in a word like "pant"? This is NOT permitted in Ebonics, and the reason (building on your newly acquired knowledge about voicing) is that Ebonics does not allow the deletion of the second consonant in a word-final sequence unless both consonants are either voiceless, as with "st," or voiced, as with "nd." In the case of "pant," the final "t" is voiceless, but the preceding "n" is voiced. Not only is Ebonics systematic in following this rule, but even its exceptions to the rule--negative forms like "ain'," and "don'"--are non-random. In short, Ebonics is no more lazy English than Italian is lazy Latin. To see the (expected) regularity in both we need to see each in its own terms, appreciating the complex rules that native speakers follow effortlessly and unconsciously in their daily lives.

Talking about native speakers naturally brings up the question of who speaks Ebonics. If we made a list of all the ways in which the pronunciation or grammar of Ebonics differs from that of SE, we probably couldn't find anyone who uses all of them 100% of the time. There is certainly no gene that predisposes one to speak Ebonics, so while its features are found most commonly among African American speakers ("Ebonics" is itself derived from "Ebony" and "phonics," meaning "Black sounds"), not all African Americans speak it. Ebonics features, especially distinctive tense-aspect forms like those in examples 1-5 above, are more common among working class than among middle class speakers, among adolescents than among the middle aged, and in informal contexts (a conversation in the street) rather than formal ones (a sermon at church) or writing. These differences are partly the result of differences in environment and social network (recall our point about geographical and social conditions forging dialects), and partly the result of differences in identification. Lawyers and doctors and their families have more contact than blue collar workers and the unemployed do with Standard English speakers, in their schooling, their work environments, and their neighborhoods. Moreover, working class speakers, and adolescents in particular, often embrace Ebonics features as markers of Black identity, while middle class speakers (in public at least), tend to eschew them.

What about Whites and other ethnic groups? Some Ebonics pronunciations and grammatical features are also found among other vernacular varieties of English, especially Southern White dialects, many of which have been significantly influenced by the heavy concentration of African Americans in the South. But other Ebonics features, including copula absence, habitual be, and remote BIN are rarer or non-existent in White vernaculars. When it comes to vocabulary, the situation is different. Partly through the influence of rap and hip hop music, a lot of African American slang has "crossed over" to whites and other ethnic groups, particularly among the young and the "hip" (derived from Wolof hipi "be aware"). Expressions like givin five ("slapping palms in agreement or congratulation") and Whassup? are so widespread in American discourse that many people don't realize they originated in the African American community. This is also true of older, non-slang words like tote ("carry," derived from Kongo -tota, Swahili -tuta).

By this point, some readers of this article might be fuming. It's one thing to talk about the distinctiveness and regularity of Ebonics and its value as a marker of Black identity and hipness, you might say, but don't linguists realize that nonstandard dialects are stigmatized in the larger society, and that Ebonics speakers who cannot shift to SE are less likely to do well in school and on the job front.? Well, yes. As the January 1997 LSA resolution emphasized, "there are benefits in acquiring Standard English." But there is experimental evidence both from the United States and Europe that the goal of mastering the standard language might be better achieved by approaches that take students' vernaculars into account and teach them explicitly to bridge the gap to the standard than by conventional approaches which ignore the vernacular altogether. (Most conventional approaches show a shockingly poor success rate, I should add.) To give only one example: At Aurora University, outside Chicago, African American inner-city students taught by a Contrastive Analysis approach in which SE and Ebonics features were systematically contrasted through explicit instruction and drills showed a 59% REDUCTION in their use of Ebonics features in their SE writing after eleven weeks, while a control group taught by conventional methods showed an 8.5% INCREASE in such features. Despite ambiguities in their original wording, what the Oakland school board essentially wanted to do is help their students increase their mastery of SE and do better in school through an extension of the Standard English Proficiency program, a contrastive analysis approach widely used in California and already in use in some Oakland schools. It was considerations like these that led the Linguistic Society of America to endorse the Oakland proposal as "linguistically and pedagogically sound."

Let us turn now to the issue of the origins of Ebonics, on which there is much less agreement among linguists The Oakland resolution referred to the influence of West African languages as the source of Ebonics' distinctive features, and as one reason for its recognition. The African ancestors of today's African Americans came to America mostly as slaves, and mostly between 1619 and 1808, when the slave trade officially ended. Like the forebears of many other Americans, these waves of African "immigrants" spoke languages other than English. Their languages were from the Niger-Congo language family, especially the West Atlantic, Mande and Kwa subgroups spoken from Senegal and Gambia to the Cameroons (e.g., Wolof, Mandingo, Twi, Ewe, Yoruba, Igbo), and the Bantu subgroup spoken further South (e.g. Kimbundu, Umbundu, Kongo). Arriving in an American milieu in which English was dominant, the slaves learned English. But how quickly and completely, they did so, and with how much influence from their African languages, are matters of dispute.

One view, the Afrocentric or Ethnolinguistic view, is that most of the distinctive pronunciation and grammatical features of Ebonics represent transfers or continuities from Africa, since West Africans acquiring English as slaves restructured it according to the patterns of Niger-Congo languages. On this view, Ebonics simplifies word final consonant clusters ("pas'") and omits linking verbs like is and are ("He Ø happy") because these features are generally absent from Niger-Congo languages, and Ebonics creates verbal forms like habitual be and remote BIN because these tense-aspect categories are present in Niger-Congo languages. However, most Afrocentrists don't specify the particular West African languages and examples which support their argument, and given the wide array of languages in the Niger Congo family, some historically significant Niger-Congo languages don't support them. For instance, while Yoruba does indeed lack a linking verb like is for some adjectival constructions, it has another linking verb rí for other adjectives, and SIX other linking verbs for non-adjectival constructions where English would use is or are. Moreover, features like consonant cluster simplification are also found among other English vernaculars (for instance, in England) which had little or no West African influence, and this weakens the Afrocentric argument. Many linguists acknowledge continuing African influences in some Ebonics and American English words (direct loans like hip and tote were cited earlier, and we can add to these loan-translations of West African concepts into English words, as with cut-eye "a glance of derision or disgust"). But when it comes to Ebonics pronunciation and grammar, they want more specific proof.

A second view, the Eurocentric or dialectologist view, is that African slaves learned English from White settlers, and that they did so relatively quickly and successfully, with little continuing influence from their African linguistic heritage. Vernacular or non-SE features of Ebonics, including consonant cluster simplification and habitual be, are seen as transfers from vernacular dialects spoken by colonial English, Irish, or Scotch Irish settlers, many of whom were indentured servants, or as features which developed in the 20th century, after African Americans became more isolated in urban ghettoes. (Habitual be appears to be commoner in urban than in rural areas.) However, as with Afrocentric arguments, we still don't have enough details of the putative source features in British and settler English varieties, and crucial Ebonics features like the absence of linking is appear to be rare or non-existent in them, so they're unlikely to have been the source. Moreover, even with relatively low proportions of Blacks to Whites in the early colonial period, and the fact that they worked alongside each other in households and fields, particularly in the North, the assumption that slaves rapidly and successfully acquired the dialects of the Whites around them requires a rosier view of their social relations and interactions than the historical record and contemporary evidence suggest.

A third view, the Creolist view, is that many African slaves, in acquiring English, developed a simplified fusion of English and African languages which linguists call a pidgin or creole, and that this influenced the subsequent development of Ebonics. A pidgin is a contact vernacular, used to facilitate communication between speakers of two or more languages. Native to none of its speakers, a pidgin is a mixed language, incorporating elements of its users' native languages, and it is also has a less complex grammar and a smaller vocabulary than its input languages. A creole, as traditionally defined, is a pidgin which has become the primary or native language of its users (e.g. the children of pidgin speakers), expanding its vocabulary and grammatical machinery in the process, but still remaining simpler than the original language inputs in some respects. Most creoles, for instance, don't use inflectional suffixes to mark tense ("he walked"), plurality ("boys") or possession ("John's house").

Where are creoles common? All over the world, but particularly on the islands of the Caribbean and the Pacific, where large plantations brought together huge groups of slaves or indentured laborers, speaking various ethnic languages, and smaller groups of colonizers and settlers whose European languages (English, French, Dutch) the former had to learn. Under such conditions, with minimal access to European speakers, new restruc-tured varieties like Haitian Creole French and Jamaican Creole English arose. These do show African influence, as the Afrocentric theory would predict, but where the patterns of various African languages were conflicting, the Creolist theory would provide for elimination or simplification of more complex alternatives, like the seven linking verbs of Yoruba referred to above. Within the United States, one well-established English creole is Gullah, spoken on the Sea Islands off the coast of South Carolina and Georgia, where Blacks constituted 80% to 90% of the local population in places. When I did research on one of the South Carolina Sea Islands some years ago, I recorded the following creole sentences, much like what one would hear in Caribbean Creole English today:

6. E.M. run an gone to Suzie house. (=SE "EM went running to Suzie's house.")

7. But I does go to see people when they Ø sick. (=SE "But I usually go to see people when they are sick."

8. De mill bin to Bluffton dem time. (=SE "The mill was in Bluffton in those days.")

Note the characteristically creole absence of past tense and possessive inflections in 6, the absence of linking verb are and the presence of unstressed habitual does in 7, and the use of unstressed bin for past and dem time (without s, but with pluralizing dem) in 8.

What about creole origins for Ebonics? One way in which creole speech might have been introduced to many of the American colonies is through the large numbers of slaves who were imported in the 17th and 18th centuries from Caribbean colonies like Jamaica and Barbados where creoles definitely did develop. Some of those who came directly from Africa may also have brought with them pidgins or creoles which developed around West African trading forts. _Moreover, some creole varieties--apart from well-known cases like Gullah--might have developed on American soil. While the percentages of Blacks in the local population might have been too low in 18th century New England and Middle Colonies for creoles to develop (3% and 7% respectively, compared with 50% to 90% in the early Caribbean), they were higher in the South (40% overall, 61% in south Carolina), where the bulk of the Black population in America was concentrated. There are also observations from travelers and commentaros through the centuries to Black speech being different from White speech (contra the Eurocentric scenario), and repeated textual attestations of Black speech with creole-like features. Even today, certain features of Ebonics, like the absence of linking is and are are widespread in Gullah and Caribbean English creoles,.while rare or nonexistent in British dialects.

My own view, perhaps evident from the preceding, is that the creolist hypothesis most neatly incorporates the strengths of the other hypotheses, while avoiding their weaknesses. But there is no current consensus among linguists on the origins issue, and research from these competing perspectives is proceeding at fever pitch. One of the spinoffs of this kind of research is the light it sheds on aspects of American history which we might not otherwise consider. Whatever the final resolution of the origins issue, we should not forget that linguists from virtually all points of view agree on the systematicity of Ebonics, and on the potential value of taking it into account in teaching Ebonics speakers to read and write. That position may strike non-linguists as unorthodox, but that is where our science leads us.