Gone with the wind? Evidence for 19th century

Links & Letters 5, 1998 127-145 Abstract For decades the variety of English spoken by African Americans in the United States has been a major focus of research in linguistics. Despite that, there is still considerable controversy over its past, and specifically whether there had formerly been a plantation creole which shaped the modern African American Vernacular English (AAVE) linguistic system as it emerged. Increasingly abundant evidence has now been assembled on the 19th century in the form of recordings of speakers born in the antebellum period, backed up by data from works of fiction. Taken together, this evidence strongly suggests that a variety of creole was indeed spoken alongside English, perhaps without clear separation, at least until the time of the Civil War. Key words: Afro-American Vernacular English, Gullah, Creole, 19th century, Slavery. Table of Contents 1. Introduction It seems impossible to begin an essay like this without first making reference to the anglicist-creolist controversy that has polarised the work of linguists researching African American Vernacular English (AAVE). Anglicists are those who think that the distinctive vernacular(s) spoken by Black Americans have developed directly from the various dialects of English brought from the 1. Introduction 2. AAVE and Gullah in the 19th century 3. The co-existence of AAVE and Creole 4. Creole microswitches in the ex-slave recordings 5. Conclusions 6. Endnote References (*) I would like to thank Joe Dillard very warmly for his encouragement and dedicate this article to him. Gone with the wind? Evidence for 19th century African American speech* David Sutcliffe Universitat Pompeu Fabra October 1997 128 Links & Letters 5, 1998 David Sutcliffe British Isles, while the creolists argue that there was also some input from a creole-like system. That is to say, there may have been a creole-like variety in use on Southern plantations at some time before the ending of slavery in the nineteenth century which helped to shape modern AAVE.1 In addition, there is a third point of view which overlaps the other two: the divergence/innovation hypothesis, put forward variously by Guy Bailey, John Myhill and William Labov,2 among others. That is, that the typical ways in which AAVE now differs from other dialects of English have emerged particularly over the last hundred years. This hypothesis would seem to fit well with the Anglicist viewpoint. However there are two theoretical complications here, which have arisen in response to what is after all a complex linguistic situation viewed across time and space. Firstly, one of the main proponents of the divergence hypothesis, Guy Bailey, now supports the creolist viewpoint (rather than the anglicist one) and argues that, regardless of what has happened in the 20th century, African American vernacular was, at some earlier period, conditioned by contact with an earlier plantation creole. Secondly, William Labov has come to the conclusion that the divergence which has occurred in AAVE over the 20th century, according to Bailey and others, has actually made AAVE more creole-like in its typology. So in a sense the plot thickens: some researchers (for example Bailey) see creole in AAVE's past, others in its present (Labov's current position), some see creole in both (the creolists) some in neither (the anglicists). It is hard to believe that these researchers are all working on the same language. But by definition AAVE is a variety of English that takes its place among the range of English dialects spoken in North America. At the same time, AAVE may also be regarded as an Afro-American language, one of the range of such languages spoken mainly by people of African descent in North, Central and South America. A majority of these Afro-American languages show some degree of creolisation. So I would argue that linguists necessarily have to be aware of the two frames of reference to be able to work towards negotiable consensus. Assuming, then that neither of the two main positions, the anglicist and the creolist, are likely to be right to the exclusion of the other, and that AAVE is best understood as having a double affiliation, deriving from a complex historical source, I can now broach the subject of this article: a review of the evidence we have for 19th century African American speech, and the extent to which it might have differed from its modern counterpart. 1. Creoles are languages which were created by speakers who initially adopted a common language as a lingua franca, often a European language such as English or French, and in the process more or less restructured it. Well-known Caribbean examples are Jamaican and Haitian creole. In fact, most of the vernaculars of the anglophone and francophone Caribbean are more or less creolised. 2. See Bailey 1995; Labov 1996. Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 129 This issue will necessarily include another, whether such differences were in the direction of a creole or not. Notice that despite the way that the notion of creole looms large in the debate about AAVE's past and present, practically no hard evidence has been adduced of an earlier creole, in terms of actual speech examples from the United States mainland, at least until very recently. The honourable exceptions to this are, of course, Stewart (1968) and Dillard (1972).3 Over the last decade we have seen an enormous enrichment of our understanding and knowledge of 19th century African American speech. It might seem surprising that we have much firm evidence on this question at all, let alone the kind of audio recordings that constitute the typical form of evidence on spoken speech varieties. However, we are fortunate in having the ex-slave recordings, a corpus of mechanically recorded speech of twelve former slaves whose birth dates range from 1844 to 1861 and who in most cases were recorded as part of the Works Project Administration scheme between 1935 and 1942.4 We also have some reliable fictional writing featuring AAVE which will take us back to the very beginning of the 19th century. And there are still other sources, including for example the Hyatt corpus (Hyatt 1978) which Wolfgang Viereck brought to our notice. What we find is that while evidence for a creole has continued to elude researchers, at least until very recently, there is clear evidence of a widespread and vigorous use of English —African American Vernacular English— which did not differ greatly from the AAVE of today. Apart from Louisiana French Creole, the only substantial exception to this generalisation is the Gullah coast and the Gullah language still spoken today along the coastal strip of Georgia and South Carolina, and especially on the Sea Islands. Gullah is of course an Atlantic Creole and closely resembles the Anglophone creoles of the Eastern Caribbean. As is well known, the Atlantic creoles were created during the 17th century, when hundreds of thousands of enslaved Africans were obliged to learn a form of English or of other European languages, mainly to be able to communicate with each other and forge a collective identity. In the process, the European language was more or less completely restructured. This at least is what happened in Jamaica and elsewhere in the Caribbean, and indeed along the Gullah coast. The question which then naturally arises is: to what extent did this happen in the United States outside the Gullah area? Did the 17th and 18th century ancestors of modern African Americans throughout the old South restructure or creolise English? And if so, to what extent did African American speech still retain signs of earlier restructuring or creolisation at the time of the Civil War? 3. The evidence adduced was necessarily not statistical and mainly limited to literary quotations some of which have been representations of African-born slaves. 4. This crucially important corpus was brought to scholars' attention by Jeutonne Brewer (see Brewer 1991). More such sound archives have now been found, and it is likely that the total number available has doubled since 1991. 130 Links & Letters 5, 1998 David Sutcliffe 2. AAVE and Gullah in the 19th century At this point it would be worthwhile looking informally at the similarities and differences between 19th century Gullah and 19th century Black English (AAVE). The following two short passages are samples of the two varieties, and are actually taken from two versions of the same story: how Brer Rabbit ate Brer Wolf's cow and got away with it. The incident related here is the basic motif of the tale, where Brer Rabbit, after eating the meat, tricks Brer Wolf by sticking the cow's tail in the ground and telling Brer Wolf that the cow is disappearing into the nether regions. (1) a. 19th century Harris AAVE «Run yuh [run here] Bruh Wolf! Run yuh! Yo cow gwine in de groun! Run yuh!» When ole Bruh Wolf got dar, which he come uh scootin, dar wuz Bruh Rabbit holin on tuh de cow-tail, fuh tuh keep it fum gwine in de groun. Bruh Wolf, he kotch holt, an dey gin uh pull uh two [they gave a pull or two] en up come de tail. Den Bruh Rabbit, he wink his off eye en say, sezee: «Dar! de tail done pull out en de cow gone,» sezee. Adapted5 from How Mr.Rabbit saved his meat (Harris 1982: 113) b. 19th century Gullah Den Buh Rabbit run ketch de cow tail en mek luk e duh try fuh pull [makes like he is trying to pull]. E grunt en e strain, e grunt en e strain. Turrecly [directly]] e tell Buh Wolf say, «Buh Wolf, all uh try [as much as I try] uh couldn moobe um.» Den Buh Wolf gone dey e ketch de tail, en time e sta-a-t fuh pull, de cow tail come out de ma-a-sh en Buh Wolf gone SHUPOW on e back een de ma-a-sh en muddy up all e self. Adapted from Cow tail een de ma-a-sh (Stoddard 1949: 12) The language of passage (1a) is a conservative form of African American English extracted from the prolific writings of Joel Chandler Harris (born in Putnom County, Georgia). Harris, the creator of the Uncle Remus stories, was evidently very successful in recreating the AAVE of central Georgia, and the latter can be taken as broadly representative of the vernacular English dialects of African Americans throughout the South at that time (1880).6 The Gullah passage (1b) on the other hand is a creole according to all the canons of what an Atlantic creole is, and can be seen to be operating within a creole grammatical system. Even so it will be noticed that the AAVE of passage (1a) has something in common with the Gullah of (1b). A family likeness is evident, for example, in the similar shape of certain morphemes across the two 5. The spelling has been adapted to make comparisons with the Gullah passage easier. Firstly er has been changed to uh (both spellings represent the sound [Ã] as in «run»). Secondly, the apostrophes indicating «missing» letters have been removed. 6. On Harris's bidialectal ability, and his plantation upbringing, see Hemenway (1982). Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 131 varieties and the relative lack of standard English inflections in both. Less evident from the above passages is the fact that both language varieties have vowels with phonemic length —not a feature of any mainstream varieties of English in North America— and in fact both have remarkably similar phonological systems. This is so, even though the two varieties are separated geographically by a major linguistic divide. This runs down the Atlantic seaboard from just south of Myrtle Bay, South Carolina, to the Florida border.7 To the east or seaward side of the line is the area where Gullah was, and to some extent still is spoken. Gullah has many overtly creole features scarcely attested anywhere else in the United States, even in 19th century fictional data. To the west or inland side of the line, English (African American English) has been in vigorous use on the plantations since as far back as reliable records can take us, and was apparently used even for in-group interaction. This is not to say that there were no creole features in the AAVE on that side. 7. There is some Gullah further north in coastal North Carolina (Tometro Hopkins, personal communication, and confirmed by Hyatt's observations and transcripts of his recordings, Hyatt 1978). I am particularly indebted to Tometro for her help. TEXAS NEBRASKA IOWA ILLINOIS MISSOURI KENTUCKY NORTH SOUTH GEORGIA ALABAMA FLORIDA MISSISSIPPI ARKANSAS OKLAHOMA LOUISIANA KANSAS TENNESSEE INDIANA OHIO WEST PENNSYLVANIA VIRGINA DEL. N.J. R.I. CONN. SOUTH DAKOTA Gullah line ➋ Chandler Harris (Putmon County, Georgia) ➋ ➊ Fountain Hughes (Charlottesville, Virginia) ➊ ➌ Celia Black (Tyler, E. Texas) ➍ Billy McCrea (Jasper, E. Texas) ➎ Laura Smalley/Harriet Smith (Hempstead, S-E. Texas) ➏ Charlie Smith (Texas, unspecified) ➌➍ ➎ Map 1. Position of the Gullah line circa 1900. 132 Links & Letters 5, 1998 David Sutcliffe But, for example, Harris himself was at pains to emphasise the difference he observed between the Gullah of the coastal rice-growing areas and SeaIslands, and the dialect spoken by Remus in «Mid Georgia«. So the difference must have been in place at least several decades before that date (Harris (1882) 1982: 43; 192). However, in the light of very recent research findings, the seemingly evident conclusion that there was no fully creole variety —no plantation creole— west of the Gullah line may now have to be revised. As a result of work begun in late 1995, there is now good reason to believe that at least some plantations west of the Gullah Line (in some cases nearly a thousand miles west of it!) were bidialectal communities. In other words, while English was clearly in widespread and flourishing use, individual speakers may have had access to a more creole-like variety that was being transmitted to succeeding generations until well into the 19th century. Furthermore, this variety seems to have been similar to coastal Gullah. We will now look at the evidence. 3. The co-existence of AAVE and Creole Some of the most solidly-based evidence for the above assertion is indirect. As we have said, the 19th century AAVE dialect itself was not bereft of creolelike features, and the presence of such features suggests contact with a creole and/or language shift away from a creole. To find out more about this, I began analysis of the Harris/Uncle Remus corpus, mentioned above. This runs into 10 volumes and nearly half a million words. To date only the first volume has been processed and one important and representative feature looked at: the extent to which the simple verb is inflected for tense after done perfective and aint negative marker. These are points of focal interest since in both cases the basilectal creole pattern (for example in Gullah) would be to Figure 1. Copy of corrections made by Joel Chandler Harris to the first edition of Nights with Uncle Remus (1880). Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 133 leave the lexical verb invariantly uninflected for tense (done tell, aint see, etc) regardless of the context. We have room here to look at the done construction: whatever happened in the past, in modern AAVE done is virtually always followed by the inflected verb (done told, for example) presumably under pressure from other varieties of English. However, when we look at the figures for Harris, volume one (see Table 1) what we find is that the verb after done follows the modern AAVE pattern 63.6% of the time (e.g. done told, etc.) and follows the creole pattern 36.4% of the time (e.g. done tell, etc.). Figure 1, above, is a reproduction of some corrections made by Joel Chandler Harris to the first edition of volume one (Nights with Uncle Remus, Harris 1880). Here, and in two other instances, Harris corrected done heard to done year (done + uninflected verb). These are just some of the many corrections he made to render Uncle Remus' language «more authentic». Most of these were not finally incorporated. The figures for done plus following verb in Table 1, therefore, refer to the text without the changes. Had they been included there would have been an even higher rate of occurrence of the creole pattern.8 So, in this respect at least, Harris's AAVE takes up a transitional position in linguistic space between Gullah on the one hand and contemporary AAVE on the other. Notice, incidentally, that the marking of the verb after done is not a minor feature, it involves a major component of the tense-aspect system, and reflects on the whole typology of the language. This finding alone suggests that there was a creole substrate or adstrate exerting influence on the African American English of the 19th century, even in areas far outside the Gullah area. But then, given such indirect proof, what direct proof is there for the occurrence of a plantation creole west of the Gullah Line? What recorded or transcribed data do we have which actually features it? As was mentioned earlier, until very recently there was almost no direct evidence of this kind. The newly available information comes largely (not entirely) from the ex-slave recordings (ESR). These are recorded interviews of aged speakers who had spent their ear8. My thanks go to the Woodruffe Library archive, Emory University (Atlanta) for supplying me with photocopies of all the relevant pages from their annotated first edition. Table 1. Rate of inflection after preverbal done. Modern AAVE Harris AAVE 19th century Gullah N% N% N% CATEGORICAL 22 63.6% 17* CATEGORICALLY ABSENT Sources: Harris 1982; Stoddard 1949; Weldon 1994 * 16 ambiguous, «don't-count» tokens also occurred. 134 Links & Letters 5, 1998 David Sutcliffe liest years on the plantations in the South until «freedom broke». This material has already received much scholarly analysis in the ground-breaking volume The Emergence of Black English (Bailey et al. 1991). So the work presented here is, in effect, gleaning after the reapers have passed. It is hardly surprising that the main transcript from Emergence of Black English (hereafter EBE) made collectively by the contributors to that volume, should contain omissions and other features which could be rectified upon further listening. After all, the editors themselves refer to the transcripts as «analogs» of the original verbal performances. In many cases originally unintelligible sequences can be clarified simply by accessing the right frame of reference. For example, in the following extract, Charlie Smith's interviewer (Elmer Sparks) refers to the title of a publication, Believe it or not. The latter is a compendium of amazing facts collected by author Robert Ripley. But without the necessary background knowledge, none of the EBE transcribers was able to decipher this. (2) Interviewer: Uh, were you, uh, interviewed by Robert Ripley? Do you remember that, Robert Ripley? CS: Robert… Interviewer: Ripley. He's right there. That's «Believe it or not.» (EBE transcribes: Ripley. He's right there. That's [unintelligible]…) EBE Charlie Smith (Texas): 259-262 Smith's great age had obviously been the reason for Ripley's interest, together with the fact that he had been enslaved and brought from Africa (Galina, Liberia). At the time of the interview in 1974, Smith claimed to be 144 (the EBE gives his age as 130), and he had been living for many years in Texas, working as a cowboy. Where proper names are involved, the problem is evidently lack of contextualisation without which it is apparently difficult to «hear» what is said, let alone parse it correctly. And in somewhat the same way, an important step towards clarifying some obscurities in the original EBE transcript was simply to enlarge the linguistic frame of reference (of the transcriber) to include creole. To expect, in other words, creole-like forms to occur. 4. Creole microswitches in the ex-slave recordings The very nature of the interview situation in which these former slaves were placed would have sharply decreased the chances of the most markedly ingroup vernacular from occurring. Accordingly, what we have by way of evidence are fleeting intrusions of creole morphemes from a linguistic level which in general the speaker did not intend to use. Three such useful markers of this kind are the use of the duh equative copula and preverbal aspect marker, creole-like personal pronouns, and the Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 135 typically creole -dem pluralising suffix. The review of these features, given below, concludes with a short note on locative copula de and on zero tense marking. 4.1. DÙH preverbal marker/copula There are many hundreds of examples of this low-toned dùh particle in the ESR materials, used preverbally in a number of unexceptional uses, for example as an infinitive marker. In such contexts it is plainly an allomorph of tuh (‘to’).9 In sharp contrast we have found a very small number of examples (perhaps ten) where dùh is used unmistakably and unambiguously as a continuative aspect marker, just as in Gullah and other creoles: (3) But they was sure fine white folks over there, …whuh I dùh wórk at. where I ASP work at ‘where I used to work’ (EBE transcribes:…where they work at.) EBE Laura Smalley (Hempstead, S-East Texas): 506 When I first listened to the Laura Smalley interview, I passed over this little sentence without paying it special attention. Luckily something registered subconsciously, and this resulted in my going back to look for it on the following day. I realised then (a) that the EBE transcription makes no referential sense in the context, since the only employee Mrs. Smalley is talking about at this point is herself; (b) the form transcribed as where ended on an [] with high pitch and was preceded by a lengthened [Ã:], a very odd pronunciation for where. Segmentation, however, gives us whuh (where) plus a first person pronoun I (pronounced [Ã] as in Gullah), which not only makes referential sense in the context but forces a grammatical reinterpretation of [də] as the continuative aspect marker.This, too, makes greater contextual sense than the EBE transcription. In one of the most telling passages of her interview, Smalley again uses dùh aspect particle when she describes the brutal flogging of a young woman by the latter's own father, Uncle S., who was plantation overseer. Mrs. Smalley concludes wryly: (4) If you pa dùh whíp you half a day, you aint want eat If your father ASP whip you half a day, you UNTENSED NEG want eat ‘If your father is whipping you for half a day you don't want to eat.’ (EBE transcribes: If they whip you half a day, you ain't want to eat.) EBE Laura Smalley: 226-227 9. Here and elsewhere I have greatly benefited from insights drawn from Mufwene (1991). 142 Links & Letters 5, 1998 David Sutcliffe The microswitches we have been looking at here are evidently small, scarcely noticed slips into a variety that the speakers might have spoken fluently and consistently in other circumstances. On this evidence, it now seems likely that a Gullah-like variety of creole was spoken alongside English in the nonGullah areas of the South from Virginia and Maryland to Texas. This finding further supports the substrate or adstrate creole influence found in the Remus stories by Joel Chandler Harris. At the time of the Civil War this Gullah-like variety was apparently still in widespread use, or at least had a wide geographical spread. Evidence from the Laura Smalley and Billy McCrea interviews shows that it was established in South East Texas, even though the latter was not settled by Anglos and Blacks until two decades or so before the outbreak of the Civil War. It is probable that most of these people came from Louisiana and Mississippi. From the John Copeland quote and other attestations it seems that the Lower Mississippi was an area where a Gullah-like variety could have survived until very late. Lastly, the occurrence of creole features in the ESR interview with Virginian Fountain Hughes and in Dance (1978), plus an attestation in Hancock (1987: 292), all suggest that a creole-like variety was in use in 19th century Virginia. At all events, this ‘hinterland’ Gullah was probably weakening and merging with English even before the Civil War. By 1940-1942 when most of the ex-slave recordings were made of very elderly speakers it must have been on the point of extinction in most areas. This evidence obviously gives added support to the creole hypothesis, and therefore challenges the anglicist hypothesis —unless the latter is restated in a weaker form, that is that varieties of British English were adopted and adapted by African American speakers alongside a more creole-like variety. As regards the divergence/innovation hypothesis, the existence of an earlier creole might well ‘explain’ the creation and expansion of apparently creole structures in modern AAVE as a resurgence, in modified form, of earlier patterns. This would perhaps involve the covert transmission of typological principles which could be activated at a later date. 6. Endnote I should not like to close without paying brief tribute to Laura Smalley, Billy McCrea and the other former slaves who are the subjects of the exslave interviews. As several of the examples quoted here show, there was much brutality and danger in their lives, even after the abolition of slavery. However, the human spirit seems to be indomitable. Mrs Smalley for example is scrupulously fair (perhaps too fair) in her assessment of her exemployers, Mr and Mrs P. These are the ‘fine white folks’ in example (3). She analyses the man as being ‘fractious’ (he was evidently prone to dangerous fits of anger) and yet a good man, and she communicates this without a hint of irony or apparent accommodation to the white man. I think she is Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 143 too old and too dignified to stoop to this. Her interviewer John Henry Faulk in turn shows a complete lack of prejudice or patronage. Even more strikingly, there are other passages, for example the interview with McDonald and an unnamed woman,12 where either the speakers are absolute masters of accommodation (traditional ‘shucking’) or there genuinely is an indomitable optimism —and I think it is the latter— which communicates across the years like a blaze of light. References BAILEY, G. (1987). «Are black and white vernaculars diverging?». American Speech 62: 32-40. — (1993). «A perspective on African-American English». In PRESTON, D. (ed.). American Dialect Research. Amsterdam/Philadelphia: John Benjamins. BAILEY, G.; MAYNOR, N.; CUKOR-AVILA, P. (1991). The emergence of Black English: texts and commentary. Amsterdam: John Benjamins. BAILEY, G.; THOMAS, E. (In press). «Parallels between vowel subsystems of African American Vernacular English and Caribbean anglophone creoles». To appear in the Journal of Creole Studies. BREWER, J. (1991). «Songs, sermons and life stories». In BAILEY; MAYNOR; CUKORAVILA, 155-171. BAUGH, J. (1991). «A reexamination of the Black English copula». In EDWARDS, W.; WINFORD, D. (eds.). Verb phrase patterns in Black English and creole, 32-59. Detroit: Wayne State University Press. DANCE, D. (1978). Shuckin' and jivin'. Bloomington: University of Indiana Press. DILLARD, J.L. (1972). Black English, its history and usage in the United States. New York: Random House. EDWARDS, W. (1991). «A comparative description of Guyanese creole and Black English preverbal aspect marker don». In EDWARDS; WINFORD, 240-255. EDWARDS, W.; WINFORD, D. (eds.) (1991). Verb phrase patterns in Black English and creole. Detroit: Wayne State University Press. ESCURE, G. (1984). «The Belizean copula: a case of semantic shift». In CARRINGTON, L. (ed.). Studies in Caribbean language. University of the West Indies, St. Augustine. Society for Caribbean Linguistics. HANCOCK, I. (1987). «A preliminary classification of the anglophone Atlantic creoles, with syntactic data from thirty-three representative dialects». In GILBERT, G. (ed.). Pidgin and creole languages. Essays in memory of John E. Reinecke, 264-334. Hawaii: University of Hawaii Press. HANNAH, D. (1997). «Copula absence in Samaná English». Qualifying paper. Department of Linguistics: Stanford University. HARRIS, J.C. (1982). Uncle Remus, his songs and his sayings. Harmondsworth: Penguin. [First published New York: D. Appleton and Co.] 12. Perhaps her name was Carol: the last words that interviewer John A. Lomax utters before turning off the machine are not transcribed in EBE. But he clearly says «Well Miss Carol…» 144 Links & Letters 5, 1998 David Sutcliffe HEMENWAY, R. (1982). «Introduction: author, teller, and hero». In HARRIS, J.C. (author) Uncle Remus, his songs and his sayings. Harmondsworth: Penguin. HYATT, H.M. (1978). Hoodoo, conjuration, witchcraft, rootwork. Los Angeles: UCLA/Alma Egan Hyatt Foundation. LABOV, W. (1998). «Coexistent systems in African American English». In MUFWENE, S.; RICKFORD, J.; BAUGH, J. (eds.). African American English. London: Routledge. MCWHORTER, J. (1997). Towards a new model of genesis. New York: Peter Lang. MIGGE, B. (1997). «Vindicating the African element of the copulas [d/n]a and de». Paper given to the Conference of the Society for Pidgin and Creole Linguistics, University of Westminster. MONTGOMERY, M. (1991). «The linguistic value of the ex-slave recordings». In BAILEY; MAYNOR; CUKOR-AVILA, 173-190. MUFWENE, S. (1991). «On the infinitive in Gullah». In EDWARDS; WINFORD, 209-222. MYHILL, J. (1991). «The use of invariant be with verbal predicates in BEV». In EDWARDS; WINFORD, 101-113. POPLACK, S.; TAGLIAMONTE, S. (1989). «There's no tense like the present: verbal -s inflection in early Black English. Language Variation and Change, 1 (1): 47-84. Cambridge: Cambridge University Press. — (1991). «African American English in the diaspora: evidence from old line Nova Scotians». Language Variation and Change, 3: 301-340. Cambridge: Cambridge University Press. PRESTON, D. (ed.) (1993). American dialect research. Amsterdam/Philadelpia: John Benjamins. RICKFORD, J.; BALL, A.; BLAKE, R.; JACKSON, R.; MARTIN, N. (1991). «Rappin' on the copula coffin: theoretical and methodological issues in the analysis of copula variation in African-American Vernacular English». Language Variation and Change, 3 (1): 103-132. Cambridge: Cambridge University Press. SINGLER, J. (1991a). «Liberian settler English and the ex-slave recordings: a comparative study». In BAILEY; MAYNOR; CUKOR-AVILA, 248-274. — (1991b). «Copula variation in Liberian Settler English and American Black English». In EDWARDS; WINFORD, 129-164. SPEARS, A. (1982). «The Black English semi-auxiliary come». Language, 58: 850872. STEWART, W.A. (1968). «Continuity and change in American Negro dialects». Florida FL Reporter, 6(I): 3-4, 14-16, 18. STODDARD, A. (1949). Animal tales told in the Gullah dialect (I-III). Washington: Library of Congress, Folklore Collection (Long playing record L46). SUTCLIFFE, D. (1992). System in black language. Avon: Multilingual Matters. — (1996). «Recutting and the history of the Black English copula». Bamberg: Dialectologia et Geolinguistica, 4: 59-86. — (forthcoming). «Breaking old ground: African American English and the search for its past». In Ramisch, H.; Wynne, K. (eds.). Language in time and space: Festschrift in honour of Wolfgang Viereck. Bamberg: University of Bamberg Press. SUTCLIFFE, D.; TURELL, M.-T. (1990). «Verb consecutivization in Black English narrative». Paper given at NWAVE-18, University of Pennsylvania. Gone with the wind? Evidence for 19th century African American speech Links & Letters 5, 1998 145 TAGLIAMONTE, Sali (in preparation). «Has it ever been perfect? Uncovering the grammar of early Black English». York Papers in Linguistics. VIERECK, W. (1988). «Invariant be in an unnoticed source of American early Black English». American Speech, 58: 99-113. WELDON, T. (1994). «Variability in negation in African American Vernacular English». Language Variation and Change, 6 (3): 359-397. Cambridge: Cambridge University Press.