History in Africa 32 (2005) 321-369
 
[Access article in PDF]

Bound to Africa:

The Mandinka Legacy in The New World

I

I offer here a theory of "cultural convergence," as a corollary to Darwin's natural selection, regarding how slave Creoles and culture were formed among the Gullah and, by extension, supported by other examples, in the Americas. When numerous speakers from different, and sometimes related, ethnic groups have words with similar sounds and evoke related meanings, this commonality powers the word into Creole use, especially if there is commonality with Southern English or the host language. This theory applies to cultural features as well, including music. Perhaps the most haunting example of my theory is that of "massa," the alleged mispronunciation by Southern slaves of "master."1 Massa is in fact the correct Bainouk and Cassanga ethnic group pronunciation of mansa, the famous word used so widely among the adjacent and dominant Mande peoples in northern and coastal west Africa to denote king or boss. In this new framework, the changes wrought by Mandinka, the Mande more broadly, and African culture generally on the South, are every bit as significant as the linguistic infusions of the Norman Conquest into what became English.

Long before studying the Mandinka as an anthropologist in west Africa, I was exposed to their legacy in the United States through my contact with the Gullah of Saint Simons Island, Georgia, my home town. The correlation between a white minority and the Mandification of the [End Page 321] English language during the slave era might be obvious to some and terrifying to others. My recently completed work on Mandinka oral traditions lays some of the groundwork for this hypothesis by providing texts that, on close examination, do seem to have some resemblance to select slave vocabulary and diction in America. I propose that the Southern accent, depsite all its varieties, is essentially an African-American slave accent, and possibly a Mandinka accent, with other African accents, along with the colonial British accent layered in.

The purpose of this paper is to consider the implications of an observation made about the practice of slavery in North America and to ask whether this view might be extended to the rest of the Americas. The observation is Philip Curtin's conclusion, after sifting through the immense number of sources available to him, that "South Carolina planters . . . had strong ethnic preferences in the Charleston slave market. They preferred above all to have slaves from the Senegambia, which meant principally Bambara and Malinke from the interior [both are Mande] . . . and they generally have a preference against short people" especially from the Bight of Biafra.2 In the present paper, Curtin's observation becomes the first in a chain of facts and informed speculation that reveal a pattern of Mandification of Southern English.

While the notorious Charleston market was not the only slave port in the U.S., it was a major port and was involved in North American slave trafficking early on, with a fairly wide regional influence into the rest of South Carolina and Georgia. Curtin notes that slave-buying proclivities in the Charleston slave market, emphasizing Mande and including the Mandinka of Senegal and Gambia, might have caused other states such as Virginia to have a slight preference for Senegambian slaves as well. When Curtin's Table 45 speculates that 13.3% of all slaves imported to North America were from Senegambia, 5.5% from Sierra Leone, and 11.4% were from the Windward Coast or Liberia, he emphasizes the regions of west Africa where large numbers of Mande still live today, including Mandingo, Mende, Malinke, Maninke, Mandinka, Susu, Bambara, Vai, and Dyula among others, distributed among non-Mande groups.3 How many Mande or Mandinka were really in these percentages? The linguistic map showing which ethnic groups in west Africa speak Mande-related languages is immense, with many groups on the coasts or relatively near slave ports.4

Of course the vast area of eastern Mali—the heartland—contains Mande-speakers. But from here the influence spread out all along the [End Page 322] Gambia River, the Pakao region of southern Senegal, northern Guinea-Bissau, major regions of Guinea and Sierra Leone, significant territory in Liberia, Ivory Coast, Burkina Faso, and even a border area of northwestern Nigeria. The seeming fragmentation of the Mande among so many regions and into slave era classifications that included geographic references to three, or sometimes four, seemingly disconnected areas—Senegambia, "Sierra Leone," "Guinea," and the "Windward Coast" (Liberia and Ivory Coast)—have worked to understate among scholars the Mande influence on the eighteenth- and nineteenth-century slave societies of the U.S., as if these geographic areas could not have a broad ethnic and linguistic group such as the Mande bound by a common language and history.

Further amplifying this seeming ethnic fragmentation is that one key slaving area—along the Gambia River—of vital importance to the slave markets of coastal South Carolina and Georgia, the Caribbean and throughout the New World in certain decades, became by far the smallest country in west Africa, The Gambia. Since the early seventeenth century the Mandinka have predominated in villages along both sides of this river, settling there after Manding (the ancient Mali empire) expanded and began to disintegrate toward the end of the fifteenth century.

II

In many ways William Pollitzer's The Gullah People and Their African Heritage is a vital source on the whole question of identifying the Mandinka contribution to Gullah culture and language, especially because he did the hard work of combing through colonial British and plantation records, and numerous mentions of slaves in colonial newspaper accounts, including ads for runaway slaves. However, Pollitzer's analysis of the Gullah suffers some by not fully appreciating the connectedness of Mande culture and language back in west Africa. Another rare defect in this important book is that his analysis of Lorenzo Turner's seminal Africanisms in the Gullah Dialect seems too literal in its reliance on Turner's African-language speakers of the 1940s who singled out the ethnic origins of the thousands of Gullah words collected.

Pollitzer in Table 16 thus notes that an astonishing 100% of the 92 words collected by Turner in Gullah stories, songs, and prayers are from Mende (69%), Vai (29%), Bambara (1.1%) and Mandinka (1.1%).5 These are all Mande ethnic groups (and most if not all were collected by Turner in Glynn and McIntosh counties, the two counties on the Georgia coast where I grew up). This concentration suggests the enormous power [End Page 323] of Mande music, prayer, and storytelling within Gullah culture, but surely other African ethnic groups made contributions as well.

Perhaps the greatest defect is that Pollitzer does not take into account the absence of a Gambian or Pakao Mandinka informant in Turner's listed group of African informants (more on this below), even though Pollitzer's historical data suggest that Mandinka slaves were often a first or second priority for slave buyers in Charleston and Georgia. A Mandinka would surely have found more Mandinka words in Turner's Africanisms, as I show below. Turner was also hampered by the absence of recent Mandinka dictionaries; David Gamble did not start publishing his Gambian Mandinka word lists until after Turner's work appeared. Pollitzer's Table 16, based on Turner's analysis, thus shows that Yoruba and Kongo have the highest percentage (15.9% and 14.5%) of 3595 Gullah words as personal names, while the following Mande groups as individual ethnicities seem to have far less importance: Mandinka and Mandingo are 4.2% and 1.6%; the Mende are 8.9%; Bambara are 6.6%; Vai are 4.5%; Malinke are 0.2%; and Susu are 0.1%.

However, the combined Mande total would be 26.1%, much higher than that for Kongo or Yoruba. For the 251 words Pollitzer notes in Table 16 that are used in Gullah conversation (as recorded by Turner), the 24.8% Kongo total seems higher than the following Mande groups: Mende 7.8%, Bambara 5.2%, Vai 7.2%, Mandinka 0.5%, Mandingo 2%, and Malinke 0.2%. However, the Mande together are 23.2% (while, curiously, Yoruba are only 3.2%). A modern analysis by Africans of all Turner's Gullah words might change these totals somewhat, as it clearly would for Mandinka.

In a similar way, Pollitzer's Table 18 takes a much-needed look at the 1940 WPA masterpiece Drums and Shadows, but almost certainly understates Mandinka and Mande influence by attempting to quantify the various magic practices of the Gullah in terms of an ethnic group and region of west Africa.6 As noted below, in the eyes of an anthropologist with considerable experience studying the Mandinka, the culture of this ethnic group seems to resonate virtually throughout Drums and Shadows, from both Muslim and non-Muslim Mandinka traditions. Also, this work, published in 1940, relied not on recent anthropological accounts of the Mandinka, but mostly on early explorers' accounts, such as Francis Moore's 1738 Travels (up the Gambia River), as main sources for comparative examples of Mandinka culture. Another problem that must be confronted and understood in appreciating Mandinka legacy in the New World, is that both Muslim and non-Muslim Mandinka slaves came to [End Page 324] this hemisphere in great numbers, adding to Mandinka cultural variety during the slave era.

Nevertheless, William Pollitzer's wonderful historical research both supports and broadens the preferences noted by Curtin, concluding that in South Carolina the order of preference for slaves was "Gold Coast, Gambia, Windward Coast, and Angola; Ibo from Calabar or Bonny in the Bight of Biafra were considered worst."7 Pollitzer cites a 1755 letter from Henry Laurens, a founding father and leader of colonial South Carolina, saying that slaves from "Gold Coast or Gambia are best." Another letter from 1756 states that "[t]he Slaves from the River Gambia are preferred to all others with us save from the Gold Coast." Compared with the latter, "Gambians were similarly tall, strong and very dark. Senegalese were considered most intelligent and esteemed for domestic service. Mandingoes were gentle in demeanor but sinking under fatigue."8

In what years or decades did these Mandinka slaves enter the Americas? In order to demonstrate the influence of any African ethnic group, we need to know the numbers of slaves arriving, and when. Pollitzer gives us the best sense of this for Charleston, indicating that 1636 Senegambian slaves were sold there during 1716 to 1744 ("early period"), representing 7.4% of the total and 11.2% of slaves identifiable by geography. This number swells to 15,951 slaves from 1749 to 1787 ("middle period"), representing 25.2% of the total and 31.8% of identifiable slaves (the largest percentage from any geographical area). In 1804 to 1807 ("final period") the number of Senegambian slaves diminishes to 506 slaves, representing 1.7% and 1.9%. (In contrast, slaves from Angola represent 51%/77% for the early period in Charleston, 14.6%/18.4% in the middle period, and 52%/56.6% in the final period).9 If we add to Senegambians the slaves brought from Sierra Leone and the Windward Coast, (the three areas comprising the Mande region), the totals rise, to 48.5%/ 61.2% in the crucial middle period, when more than half of all legal importation into Charleston occurred (68,701 slaves out of 121,464). [End Page 325]

As noted, Mande slaves came not just from Senegambia but from Sierra Leone (especially the Mende in the final period) and the Windward Coast; other ethnic groups in this large area likely had at least a few Mande-language speakers, and the culture of non-Mande groups such as the Jola and Bainouk, among many others, may have been influenced by Mande. Pollitzer hypothesizes that a "homogenous group" arriving in South Carolina and Georgia "first and in large numbers had an opportunity to establish their common speech and culture" while later groups had to adjust.10 While Pollitzer uses such an analysis to imply that Angolans from this "Gola" region of Africa came through Charleston in greatest numbers in the early period (1716-44), and thus influenced the "Gullah" name more than the Gola of Liberia, this concept of a homogenous group coming relatively early, could just as easily apply to the Mandinka and the Mande more broadly in the middle period (1749-87).11

On the Georgia coast, from 1755 until 1798, the presence of Gambia slaves was just as significant as those coming through Charleston, if not more so, within a group of 6539 estimated by Donlan. Of these, 2038 slaves came from the Caribbean; out of 3680 from a known area of Africa, 43% came from Gambia and 44% from Sierra Leone or the Windward coast.12

The several rice and cotton plantations in South Carolina owned by the influential Ball family provide a rare case study where the probable ethnicity of the slaves was documented in the eighteenth century; by far the largest number of these slaves came from Gambia, implying a Mandinka preference by the Ball family and suggesting that a knowledge of rice cultivation was important for selecting their slaves.13 An ad in the South Carolina Gazette for 1785 noted 152 slaves from Gambia for sale and proclaimed, "[t]he Negroes from this part of the coast of Africa are well acquainted with the cultivation of rice."14 There are today large concentrations of rice-cultivating Mande ethnic groups living in the area of Senegambia, Sierra Leone, and the Windward Coast, the areas that provided 61% of all slaves imported into Charleston from the middle period (1749 to 1787). Indeed, the broad area encompassing Liberia was sometimes known in the colonial era as the "Grain Coast or Rice Coast."15

Eighteenth-century ads in Charlestown newpapers for runaways tell us that slaves from the Senegambia (and Guinea) were the tallest and (along [End Page 326] with the Ibo) lighter-skinned. Many runaways spoke foreign languages, implying mobility in Africa, and played musical instruments, including the violin. Senegambians appeared perhaps less often in ads as runaways because they seem to have been used more as house servants with less chance of running away than a field hand. Slightly more than half the African names in the South Carolina Gazette from 1732 to 1775 seem to have been Tshiluba names, implying a Bantu heritage, but names from Angola and Gambia were significant. However, figuring out the ethnic heritage of African names requires linguistic sophistication. When Pollitzer points out that the name Keta is a common name in Yoruba, Hausa, and Bambara, and written by a Southern owner as Cato, I would speculate this is very likely a reference to Keita, the name well-known to Mande Africans of the highly-influential ruling clan of ancient Mali. As if referring to a veritable incubator and laboratory for jazz, in 1886 George Washington Cable fancifully described the Place Congo in New Orleans as the scene of exuberant music, dance, and singing by a variety of a dozen identifiable ethnic groups, including tall, well-built Senegalese and Gambia River Mandingo, who were slightly less well-built but cunning and lighter-skinned.16

A key component of the Mandification of Southern English is that back in Africa, Mande traders, warriors, and emigrants were already spreading their influence throughout much of west Africa. Judith Carney points out that "seven hundred years of Mande empire formation, however, would leave a pronounced legacy on the linguistic and cultural map of West Africa." This resulted in "the widespread diffusion of Mande languages as well as selected cultural practices throughout West Africa, a cultural process that Paul Richards has referred to 'Mandingization'." Carney suggests the process of Mandingization in west Africa began at the dawn of empire-building, at least by 700 CE with two types of knowledge, cultivating glaberrima rice and smelting iron. Accorded the powers of magic, the caste of smiths migrated into forest areas in search of charcoal, and the arrival of iron implements spread rice cultivation.17 The advent of Islam in the Mande heartland area by about 1000 CE, the conversion to Islam by rulers of the Mali empire prior to 1400, and the spread of Islam, aided by jihad, into even the first years of the twentieth century no doubt amplified the process. Mande warriors, urged on by their clerics, proselytized among and battled non-Muslim ethnic groups, including the large number of non-Muslim Mandinka who had not yet converted even in the nineteenth century. [End Page 327]

Because of Mandingization, the Mande and neighboring ethnic slaves sold into the Americas came here already equipped with a kind of linguistic and cultural homogeneity or anchor. Mande culture in its broadest sense could help them both in communicating with fellow slaves and in creating new societies, within the cultural hodge-podge in the South and elsewhere. In such a free-for-all, Mande commonality achieved at least a modest cultural and linguistic dominance, capable of influencing other slaves, their owners, and other white people. At least one major Southern slave owner, Thomas Spalding of Sapelo Island, Georgia, purchased African slaves from Mande-preferring Charleston, based on these slaves' ability to communicate with each other and the Africans already running his plantation, to make it easier to train them in the work of the plantation.18

III

Before proceeding further, I must admit to a Mandinka bias, having done nearly two years of field work with them in 1972 and 1974-75, and returning briefly in 1980. At same time, my expertise with Mandinka culture and language hopefully makes it easier to identify possible areas of their influence in this hemisphere. But any comparisons between Mandinka usage today and slave Creoles in America must also bear in mind that change is a constant, wherever people live. The French words used by the Mandinka during my field work, such as Commandante (leader) and Anglais (the Gambia), appearing in the legends of Djinns, Stars and Warriors, are but one example of an ongoing linguistic change. The same process of Creolization no doubt happened in the U.S. South, as slaves poured in directly from Africa or via the Caribbean. Among numerous other geographic and ethnographic classificatory difficulties, did contemporary observers from the sixteenth though the nineteenth century even know if a "Mandingo" was really a "Mandingo," as distinguished from a Wolof or Fulani, not to mention hundreds of ethnic groups farther south in Africa?

Yet, as Pollitzer, Curtin, and others make abundantly clear, many slave traders and plantation owners in South Carolina and Georgia had a fascination for ethnicity which cannot be easily dismissed. At least some eighteenth and nineteenth century observers, such as Henry Laurens above and the Jamaica slave-ownesr Bryan Edwards or Thomas Spalding mentioned below, seem to have been keenly aware of various west African ethnic groups. At some point in the future, computer linguistic [End Page 328] modeling and genetic tracing methods, of the kind that show the origins of British islanders, might confirm or disprove the patterns below that emerge from historical and linguistic analysis. Perhaps Mandinka/Mande cultural and linguistic influence in the Americas will then seem even more significant and discernible, alongside the influence of other ethnic groups from Africa. The identification of Mande influence in the South, the Caribbean, and Brazil, must also be conditioned with a huge reality—ethnic diversity. Slaves from hundreds of ethnic groups from all over western Africa came into the South and the rest of the Americas along with the Mandinka/Mande. At least some of these groups, especially larger ones such as Yoruba, Kongo, and Angolans were also fairly widely diffused back in Africa, and their influence has been discerned among the Gullah in the South.19

IV

Even the name "Mandingo" has a certain, if varied, cachet in different parts of the Americas. An economist from Argentina told me that in his country the term Mandinga traditionally was a "black devil" or a person of African origin with mysteriously threatening or magical qualities. In modern Brazilian Portuguese, a Mandinga is a fetish, a kind of material Vodou object capable of causing either good or evil, or it is a charm worn to protect the body (like the grigri charms worn by Muslim slaves in the 1835 Salvador revolt or worn by Mandinka Muslims historically through the present day.) A Mandingueiro/a in Brazilian Portuguese is a kind of sorcerer (or Gullah "root doctor") who "makes Mandinga" (fazer Mandinga in Brazilian Portuguese.) Mandingar is to bewitch or to use sorcery. Mandingaria is witchcraft or the practice of sorcery.

Curtin's and Pollitzer's data give at least some statistical basis for the kind of old Southern rumor Kyle Onstott used in writing the novel Mandingo, implying an ethnic group of African-American slaves considering themselves superior to other slaves and considered superior by some Southern slaveowners. The appearance of Alex Haley's Roots and the Mandinka slave Kunta Kinte, did not necessarily detract from this myth, but once the hype and storytelling are set aside, convey the idea of certain Mandinka linguistic traditions being passed on through several generations in one biological family. The takeover of the slave ship Amistad by Mende slaves must also have enhanced the reputation for leadership of this ethnic group in the ante-bellum United States. [End Page 329]

What happens when a whole lot of similar traditions from African-born slaves get passed on together in the South, on a scale that occurred after the Norman conquest of England, when hundreds of French words were introduced into English and indeed the whole way of speaking English changed? It would seem logical that the introduction of at least a couple hundred thousand African-born slaves into the South, owned by the trend-setting Southern elite, would not just introduce new vocabulary, but influence diction and even accent. My contention is not only that this happened, but that we can begin to identify the Mandinka influence in these changes.

Within the broad Mande group, the Mandinka in particular, along the Gambia River and in the center of southern Senegal's Casamance region, are from one of the earliest areas of west Africa to be extensively slaved, beginning in the mid-fifteenth century. This region was the first one reached as Portuguese and other explorers proceeded southward along the west coast of Africa. Because of the westward bulge in west Africa, this Mandinka region lies comparatively close to the Caribbean and the U.S. Situated on the westernmost point of Africa, Goreé Island became a major port of embarkation in the trans-Atlantic slave trade. A great many Mande slaves, but certainly not all, came through this important port, although there were several other slave ports in Guinea-Bissau, Guinea, Sierra Leone, and Liberia through which Mandinka and other Mande slaves were shipped.

Aided by prevailing wind patterns, slave traders plying the Gambia, the Casamance and other nearby rivers within a few hundred miles south could maintain that their ships reached the Caribbean and Charleston relatively faster with, if one can even use the adjective, healthier slaves. These rivers also lie relatively close to the Mande heartland in the western Mali and the trans-Saharan trade routes terminating there, so geography influenced not only slave trading patterns, but the whole process of Islamization in west Africa. Geography—access both to rivers closest to the Americas and to trans-Saharan trade—favored Mande expansion in Africa and their influence in the Americas.

However, geographic arguments favoring Mandinka influence can be pushed only so far. Politics certainly intervened. After 1807 when U.S. law outlawed the importation of African slaves, the British navy aggressively intercepted slave ships off the coasts of Senegal, Gambia, Guinea-Bissaou Sierra Leone, and Liberia, tending to push the focus of slave trade further south, toward regions such as the Congo and Angola. The ever-rising demand for slaves in Brazil, closer to Nigeria, the Congo, and Angola, also increased the importance of these regions in the slave trade. [End Page 330] Betty Kuyk does a good job of suggesting that a large illegal slave trade arose in a state like Georgia after 1808 until 1858, where efforts to suppress the trafficking were less intense, and this trade heavily favored the Kongo people. As she says, "[i]n the Sea Islands, Kongo people came late and stayed in large numbers."20 She suggests that the numerous west Africans already established along coastal Georgia and South Carolina, with their accent and speech patterns already set, were joined by a large number of Kongo people smuggled in.21

At first glance, Curtin's numbers suggest that only about 5% of all slaves coming into the Americas were Senegambian Mande, and slightly more than this if Mande from Sierra Leone and Liberia are included. However, limited evidence suggests that Mande and Senegambian slaves were more significant in the earliest years of the slave trade and at a few other times and places, including, for example, middle eighteenth-century Georgia and South Carolina. While Pollitzer's work above shows that Angolans rather than Senegambians dominated the early period (1716-44) of slave importation into Charleston, this is perhaps not true for two early colonies in theAmericas—Mexico and Peru. Curtin's list of 207 African-born slaves in Peru compiled from 1548 to 1560, while perhaps not statistically significant, shows 74% of the total were from Senegambia and present-day Guinea-Bissau, an area where early Portuguese slavers concentrated, and that 7.2% were Mandinka (15 slaves). A list of 83 African-born slaves (from 123 born in Africa), compiled from the Mexican estate of Hernán Cortés, shows 10.8% Mandinka (9 slaves) and 88% from Senegambia and Guinea-Bissau.22

The ethnomusicologist Michael T. Coolen focuses on another area, Georgia between 1765 and 1775, where a relatively high percentage of slaves, 53%, were from Senegambia, with significant percentages from other years of the late nineteenth century.23 Paralleling Curtin and Pollitzer, his research concludes that the planters of both Georgia and South Carolina had a clear preference for Senegambian slaves, especially those along the Gambia River, where there were large concentrations of Mandinka (and to a much lesser extent Fula and a scattering of ethnic groups, including the Wolof). Coolen observes that planters liked the fact these slaves were often expert horsemen and traders, and that traders could boast of relatively short sailing times. [End Page 331]

To focus on one fairly distant colony for which early numbers are available, Peru, Stephen Bühnen concludes that Mandinka contributed 9% of the slaves in Peru between 1548 to 1650, larger than most groups, but paling by comparison with the Bran (27.4%) and Biafra (17.3%), whose higher percentages are explained by their proximity to Portuguese towns on the coast of Guinea-Bissau. The Banol (Bainouk) average of 10.7% also exceeds the Mandinka. Interestingly, the Bainouk percentage rises sharply from single digits to 17.9% in 1595, and remains mostly higher than their 10.7% average until 1625 (21.3%), raising the possibility that a major war might have doubled the number of their people entering slavery.24

In fact, oral traditions of the Mandinka in Pakao, in Senegal's central Casamance region, recount that their forbears defeated the Banol in roughly this time frame, resulting in these people being called the pejorative Mandinka appellation Bainouk—"those chased away"—from the Mandinka bai, meaning "to chase out." Bühnen goes on to suggest that the whole trans-Atlantic slave trade was relatively confined to the coasts and that for the period 1560-91 more than half of all African slaves (54.2%) and 67.2% of Upper Guinea slaves came to the Americas from a miniscule area of some 20,000 square kilometers reaching from the lower Casamance River to the Kogon River—basically southern coastal Senegal and northern Guinea-Bissau. This relatively small area, which includes the Bainouk and Cassanga regions of western Pakao overrun by the Mandinka, was an early ground zero in the slave trade. Bühnen implies a significant Mandinka influence on both the neighboring Cassanga and the Banol (Bainouk) by noting how the later two groups have a witch-detecting Mama Jombo mask, which in both name and function seems roughly similar to the legendary Mandinka Mumbo Jumbo. The Mumbo Jumbo, first reported by the English explorer Francis Moore in 1738, is an eighteenth- and nineteenth-century mask roughly synonymous, and concurrent, with the Mandinka kangkurao mask, in which the wearer covers himself with the blood-red bark of the fara tree.25

Following the zenith of the ancient Mali empire from ca. 1250 to ca. 1350, Mandinka emigrant/traders, warriors, and their Islamic proselytizing marabouts, spread westward along the Gambia River and into the upper Casamance River, expanding Mandinka influence among neighboring [End Page 332] ethnic groups, some of whose people they helped to enslave. The Muslim Mandinka were especially oppressive toward their geographical neighbors—non-Muslim Mandinka and the Banol (Bainouk), Cassanga and Jola peoples along the coast from the Casamance River up to the Gambia River.

V

Bühnen provides a stunning example of how an adjacent but non-Mandinka ethnic group could have helped carry a Mandinka linguistic and cultural concept through slaves into the Americas through the Mumbo Jumbo. He also points out that massa is the Cassanga and Banol pronunciation of the Mandinka word mansa, meaning king or boss.26 (The name of the Casamance River is derived from Cassanga and mansa, meaning rulers of the Cassas, and the rulers of both ancient Mali and subsequent Mandinka kings along the Gambia River were all called mansa).

One can easily imagine a cultural convergence with the English word master powering the Mandinka word mansa and the related Cassanga/Banol massa into the eighteenth- and nineteenth-century slave vernacular in North America. Massa is not a mispronunciation of master but the correct pronunciation of a widely used West African word for king or boss. If master, mansa and massa (and monsieur in Haitian or Cajun Creole) all sound the same and mean something similar, the chances rise for that word's vernacular to be used interchangeably.

In this demographic and linguistic sense, cultural convergence in the development of American Creole language and culture is a variation or corollary of Darwin's theory of natural selection, suggesting that multiple sources increase the chance a word, expression, or cultural feature gets adopted and survives. Lorenzo Turner implies this in Africanisms of the Gullah Dialect by noting in his lengthy wordlist that numerous African words have very roughly the same meanings among sometimes widely disparate ethnic groups. Pollitzer picks up on this by saying that "[t]hose linguistic features understood by the largest number of slaves and shared with English were most likely to survive."27 However, it is best to think of cultural convergence flexibly; sometimes there is convergence with English, sometimes not. The key is probably that roughly similar sounds and cultural features from different African ethnic groups, and meaning roughly the same things, have a better chance to be powered into an [End Page 333] American Creole. Numerous examples of cultural convergence in music, language, and culture are discussed below.

Coolen directs us to a musical parallel that could serve as another stunning example of cultural convergence by noting that the adjacent Fula, Mandinka, and Wolof all used musical instruments "strikingly similar" to the fiddle and banjo, two of the most popular slave musical instruments.28 The Mandinka today call their instrument of this sort the halam, which is held sideways and played like a banjo. Coolen communicated to me by e-mail that during an interview in Dakar with Abdoulai Ndiaye, a gewel and instrument-maker of Tukulor heritage, a discussion emerged about the U.S. banjo. Ndiaye then mentioned this U.S. instrument was like the old plucked lute called the "banjar." Mandification in the U.S. South would have ended this term with a vowel, thus "banjo," if this term did not already exist in eighteenth-century Mandinka.

The Swedish banjo historian Ulf Jagfors helps build a fascinating case for this cultural convergence in music by focusing his search for origin of the American minstrel banjo on the long-necked akonting banjo of the Jola, with at least some influence, apparently, from the neighboring Mandinka.29 The Jola are a non-Muslim people in the coastal Casamance region of Senegal and northern Guinea-Bissau, who are bordered on the north by Gambian Mandinka and on the east by Casamance Mandinka. Jagfors shows how the constrruction of akonting, its up-picking style of play, some of the songs played, and even the common Jola names of Sambo and Juba (Jibba) all point to the akonting as the nearest relative, among several related Senegambian instruments, to the American banjo. Jagfor's Jola informant Daniel Jatta asserts that the name banjo comes from the Mandinka word bangoe, for the local papyrus used in making the long neck of the akonting banjo and that resembles Asian bamboo in its qualities of hardness.

During my field work in the early 1970s, what the Mandinka in the Pakao region of Senegal called bung or bungo, looked for all intents and purposes like bamboo. This bamboo was also a critical building material when split and woven to make raised platforms for sleeping and conversation; walls around the washing areas of round, mud-brick houses; rice-threshing baskets and conical hats, among other things. Jagfors notes how the musical historian Samuel Charters visited the Gambia about 1980 and recorded a song by the prominent Mandinka griot, Alhaja Fabala Kanutheh, about how the Portuguese sailed to the Gambia to buy slaves in the fifteenth century. One line notes that the Portuguese found [End Page 334] people chopping down "sticks they called 'bang' and the Europeans asked them, 'What are you cutting?' and they said they were cutting the sticks called 'bangjola,' and the Europeans wrote down the name."

As it happens, the name of Gambia's capital city is Banjul, a Mandinka expression referring to the island where the papyrus or bamboo grows and where the Portuguese encountered the locals.30 Jagfors believes the Jola's habit of drumming at night after work while playing music and getting drunk under a palm tree made them an easy catch for slavers, who needed musicians to perform on slave ships, so the slaves could dance for exercise to remain healthy. Jagfor's informant, Daniel Jatta, was warned as a boy by his parents not to play the akonting in the forest or the devil would take him away forever. Several Jola akonting players told Jagfors of this and how, even today, Jola drummers insist on playing only within the safety of a village.31

A late eighteenth-century watercolor, The Old Plantation, in the Rockefeller Folk Art Center in Williamsburg, shows a primitive banjo that looks remarkably like a Jola akonting, with its characteristic long neck.32 The ethnomusicologist Michael Coolen tells me the dance in this painting looks like the stick dance still performed by the Jola. Seated next to the banjo player, the drummer has his head covered in a turban (like a Mandinka) and plays a drum held between his legs like a Mandinka jembe, beating it with two drumsticks, more like a Mandinka tama or tantango would be played.33

Amplifying the theme of cultural convergence among adjacent west African people, Coolen points out that the Wolof word for slave is jam and that the Mandinka word for slave is jon.34 Is there a Wolof derivation for "jam session" and the verb "to jam," perhaps reinforced by the Mandinka? Jams as a verb or the plural of the noun could easily have been pronounced "jazz" in African-American slave dialect. Is it plausible to see these west African words for slave embedded somehow in the origin of the word jazz? The concept of cultural convergence allows other potential derivations of jazz, like the French chasse, or other African "j" words, to join in powering this important word into use. One of these, Coolen tells me, is "jas," another Wolof word, meaning "to mix up," offering reinforcement as yet another possible source for jazz. Pollitzer [End Page 335] notes that jazz may come from the Hausa word jaiza, which describes the sound of drumbeats.35

VI

Cultural convergence is not necessarily required for an African word to become popularized in African-American slang. The Wolof, for example, a coastal people often at war with the adjacent Mandinka in the eighteenth and nineteenth centuries, offer a number of other potential sources for terms not necessarily reinforced by Mandinka. Michael Coolen tells me that these Wolof words include the terms "hip, hep, hipkat and hepkat." Hep is "someone who comes to know something." Kat is an agentive suffix, making a verb into a noun. Wolof, by implication, is behind such important slang expressions as "hip" and "cool cat." The linguist David Dalby told me that "Digana Wolof?" or "Degana Wolof?" was the source of "dig" in U.S. slang, as in "Can you dig it," meaning "Do you understand?" In fact the Wolof repeat the word "Dinga" or "Denga" to ask someone with emphasis, "Do you understand me?"

Pollitzer and Turner give numerous examples of how non-Mandinka words could have influenced English. Vodou in Haiti and hoodoo in Gullah come from vodou, a god or demon in Fon and a good or bad spirit in Fon (although I must point out the Mandinka of Pakao labeled as bunyu furo a fetish made of chicken feathers and a shard of cracked pot). Arabic tabix for cement and the Wolof tabi for earth or a similar hard material, perhaps gave rise to tabby, the Southern English word for a cement made of limestone and oyster shell (although I must point out again that in Mandinka tabi means "to cook," as in stirring a broth in an iron pot, perhaps evoking the process of making tabby). Shindu, noise made by feet in Gullah and Kongo, may have engendered the word shindig, perhaps mixing in the Wolof dig.36

Even the modern rap-singing of black musicians might find some origins in a curious Mandinka tradition, the fino, or rapping/chanting griot. The entire Pakao village of Sumbundu is composed of such griots, who rhythmically chant their praise without accompanying musical instruments. While griots who sing, drum, or play the kora, a 21-string calabash instrument, are well-known among Mande peoples, the fino or rapping griot is far less known in the West. Sumbundu elders talk about their [End Page 336] origins directly from Manding, the ancient Mali empire, with an original founder, of the clan name Kamara, who brought in the fino tradition with him. As Pakao evolved, the fino tradition became popular because it was less overtly musical, performed without instruments, and thus more acceptable to the conservatism of Islam. Did this formal Mandinka chanting by their fino griots find its way into African American slave traditions? Guy Johnson in Drums and Shadows reports Gullah shouting in Church, accompanied by drumming.37 Mandinka fino rapping and Gullah shout-singing may have influenced rap music through another precursor such as African-American prison "toasts" and hustler poetry from the 1950s and earlier, which resemble both rap and gangsta rap.38

VII

Mande musical traditions including the griot caste musicians with their kora, balafong (wooden xylophone), and halam (small guitar or banjo), and the various Mandinka drums—the tabala and tama and jembe—all could have helped Mandinka slaves have great impact in the slave culture of the United States and elsewhere in Americas. The Maninka concept of ngara, the master oral historian—or storyteller, singer, and musician—pervades Mande/Maninka culture.39 This seems to have left a deep imprint on slave society when we recall that all of the stories, prayers, and songs in Turner's Africanisms were from Mande ethnic groups. Such influence seems especially likely when considered in the broader Senegambian context where there were similar musical instruments and traditions among the Jola, Wolof, and Fulani. Rich Mandinka oral traditions by their jeli or caste musicians (griots) about descent from the kings of ancient Mali might have given slaves from this ethnic group an inherent confidence that so impressed one notable slave owner in Jamaica, Bryan Edwards. Edwards found that his Mandinka slaves considered themselves superior to the other slaves and attributed this in part to the ability of some of them to write Arabic, impressing other slaves and even their sometimes illiterate owners:

Most, if not all, the nations that inhabit that part of West Africa which lies to the northward and eastward of Sierra Leone, are Mahometans, and following the means of conversion prescribed by their prophet, are, as we are told, perpetually at war with such of the surrounding nations as refuse to adopt their religious tenets. The prisoners taken in these wars furnish, I [End Page 337] doubt not, a great part of the slaves which are exported from the factories on the Windward coast, and it is probable that death would be the fate of most of the captives, if purchase were not to be met with.

But the Mandingoes have frequent wars with each other [Muslims versus non-Muslims], as well as such nations as they consider enemies of their faith… [Edwards had another Mandingo slave] who could write with great beauty and exactness the Arabic alphabet, and some passages from the Alcoran.

The advantage possessed by a few of these people, of being able to read and write, is a circumstance on which the Mandingo Negroes in the West Indes pride themselves greatly among the rest of the slaves; over whom they consider that they possess a marked superiority; and in truth they display such gentleness of disposition and demeanor, as would seem the result of early education and discipline.40

The education Edwards refers to would have come not just from Qur'anic schools around campfires for the boys, but the traditional (non-Muslim) education and discipline infused by circumcision and minimum of two weeks seclusion of both boths and girls, not to mention age-grades and other Mandinka secret societies. One can sense from Edwards' remarkable observations that the Mandinka slaves he knew were admired by other slaves, yet feared by owners (Mandinka were "more prone to theft than any of the African tribes").41 The Mandinka reputation among fellow slaves would also surely have been enhanced by the legends of great Mande kings sung by griots, by Mande prowess in war and trading, and by the devotion of an important group of Mandinka to Islam. The sense of superiority, and indeed manifest destiny, pushing the Mandinka from their Mali heartland roots out toward the northern coast of west Africa, might have helped them assume leadership roles in the slaves societies of the Americas.

VIII

West African Islam had an enduring and proselytizing quality that was transported by west African slaves to the Americas. Several notable Arabic script documents written by west African slaves have shown up all over the Americas.42 In Brazil we even find numerous medicinal charms [End Page 338] comprised of Arabic script messages encased in leather, as worn among Mandinka in Pakao and all over west Africa. In 1835 these charms were carried into a Muslim revolt or jihad by mostly Nigerian Yoruba (or Nago slaves) erroneously called Malês (meaning literally "from Mali"), a term Reis believes is from the Yoruba imale, meaning Muslim. Reis conjectures that the term was brought by Mande marabouts emigrating southward from Mali.43

We can infer from Allen Austin's work that perhaps as many as 10% of the slaves coming into the Americas were Muslim, coming mainly from the Mande, Fula, Hausa, and a few other, mostly west African ethnic groups. Austin includes biographies of several of these west African-born Muslim slaves.44 One of them, Bilali, left to posterity a 13-page manuscript in Arabic that was translated for me by a Mandinka Jakhanke descendant of Pakao's 1843 jihad leader Syllaba.45 Bilali was said to have been a Fula from a Fulani capital Timbo in the kingdom Futa Jalon, where Mandinka live in close proximity, but nothing in the style of writing in the manuscript suggests that Bilali himself was Fula. Bilali's writing seems more in the Mande style, resembling the writing among the Pakao Mandinka, and perhaps is Susu, according to a preliminary report e-mailed to me by Nikolai Dobronravine, of St. Petersburg University. Bilali, who prayed to Allah daily, read from his Qur'an, and practiced his Muslim faith openly, was at the very least a chief slave-driver or acting foreman for the Sapelo Island planter Thomas Spalding. Bilali attracted considerable fame among his fellow slaves, and even the white general public, for, among other things, saving the island population from the great hurricane of 1824 and drilling a local slave militia, armed by Spalding with rifles, to prevent a British incursion on the island during the War of 1812.

As in Timbo, Mandinka Muslims and Fulani Muslims lived side-by-side in the Pakao area of present-day southern Senegal. The Mandinka and the Fulani made alliances to wage jihads against non-Muslims, as happened in Syllaba's Pakao jihad. But the Mandinka and Fulani were also enemies who sometimes enslaved each other's people. This complex [End Page 339] relationship led the contemporary observer Bryan Edwards to consider the Mandinka as a broad ethnic group—incorrectly—actually including the Fula. He says "Mandingoes . . . consist . . . of very distant tribes, some of which are remarkably tall and black and one tribe (called 'Phulies' or Fula) a link between Moors and Negroes properly so-called."46 Edwards seems to use the tribal name "Mandingo" generically to denote any west African Muslim, just like Malê (meaning "from Mali") was used in the first half of the nineteenth century in Brazil to refer to any west African Muslim, including some, like the Nago Yoruba who were of course not Mandinka.

IX

Linguistic evidence also supports the notion that Mandinka slaves were significant among west African Muslim slaves brought to the Americas. In Africanisms Turner found more than a dozen Mandinka Muslim names and Mandinka Arabic religious words, along with numerous other Mandinka words among thousands of African words he identified among the Gullah of coastal Georgia and South Carolina.47

The Muslim personal names Turner found include Ibrahima, Mariama, Siaka, and Mamadu, among numerous examples, all identified with at least one Mande ethnic group, and which are, incidentally, widely used among the Pakao Mandinka in southern Senegal, and presumably among the Gambian Mandinka. Turner's tendency to associate Muslim personal names with the Mande, rather than with Fulani, Wolof, or other groups at least partially Islamized in the eighteenth and nineteenth centuries, might be correct. But it also perhaps reflects a mistaken bias of both eighteenth- and nineteenth-century observers and Southern slave owners to associate any Muslim slave, especially one writing a little Arabic, to be a "Mandingo" (i.e., a Mande, if not Mandinka).

Turner found several other Islamic words among the Gullah, whom he identifies as coming from one of the panoply of ethnic groups considered by Western observers to be among the Mande. These include Kitimu, an important Muslim festival, karamo (Muslim teacher), laila (oh God), moriba (Muslim saint, also great marabout), days of the week such as arjuma (Friday), and Muslim prayer times of the day such as alansaro (3 P.M. prayer) and fitero (6 P.M. prayer). Other Mandinka words that Turner finds include several terms for animals that are also clan totems: bamboo (crocodile, a totem of the Mamburi), bida (the black or spitting [End Page 340] cobra, a totem of a "noble clan," in fact the Drame), and jati (lion, the totem of another "noble clan," probably the Keita).

Turner lists several Mandinka clan names (using his spelling): Gojan, Sougko, Touri, Sane (noble clans), Dabo (a clan of "petty traders," a noble clan in Pakao), and "Keyita" (not identified by Turner, but they are the royal clan of medieval Mali). Among the lower castes—including artisan-praise-singers, griots, and slaves—Turner includes Dafi (a clan of caste leatherworkers), Tungkara, and Kijera (clans of caste goldsmiths and blacksmiths), Suso (a clan of caste drummers and jesters), and Danso (a slave clan of weavers).

Already we see a pattern of religious words, Muslim given names, clan names cutting across all castes, and important animals that happen to be totems—exactly what one might expect, words handed down representing something fundamental and precious from Mandinka slaves' lives back in Africa. Turner goes on to identify additional words that play central roles in Mandinka culture: fa (father), lula (5), konondo (9), and other words for numbers, jambo (leaf), jiyo (water), juso (liver, a "good liver" commonly means good-natured today in Pakao), kidola (gun), kemu (man), kodo (silver, incidentally also money in Pakao), musolu (woman), musonding (girl), sajano (harvest season), safero (to write), sali (to pray), sama (rainy season), sani (gold, to purchase), solo (leopard), somanda (morning), yiro (tree), tiyo (master), warata (large), tilo (sun), tilibo (eastern land, in Pakao ancient/medieval Mali, and an indirect reference to Mecca), tana (totem), tamu (own, in Pakao to walk on), tambo (spear), taba (edible fruit, also pronounced tabo, the most revered tree in Pakao), koima (white), suto (night), kongko (hunger), kuntingo (hair), mala (shame, from Pakao this appears to be an important pre-Islamic concept), minto (where are you), mirango (gourd), and sining (tomorrow; in Pakao siningding means day after tomorrow).

Turner's list of words is breathtaking, almost painful to contemplate, when one considers the process by which these words got from west Africa to Turner's notebook. The words, distilled through the unique torture and deculturization of slavery, show what is important and fundamental to the Mandinka in a most profound way, and also how their culture maintained linguistic vitality despite the horrors of enslavement.

Unfortunately, Turner used only Mande informants who were Bambara (presumably originating from Mali), Mende from Sierra Leone and Liberia, and Vai from Liberia. He does not report a Mandinka from the Gambia River or Pakao region in his interesting list of named informants.48 A Mandinka informant would surely have noticed quite a [End Page 341] number of African words that Turner failed to associate with the Mandinka and instead associates with other ethnic groups, including doko (work or younger sibling), hadi (yes), jalo (griot or praise-singer), kelo (war), bangko (land, country), ko (salt), kono (stomach), bada (forever, from Arabic and thus almost religious, as used in Pakao), mali (ancient Mali), mansa (king), misera (small mosque, often the first one established in founding a village), nomo (a slave caste name), namanole (male and female circumcision novices), saba (three), safo (listed as the last Muslim prayer of the day, but could also be amulet or written charm), singa (the circumcised or purgatory, a pre-Islamic concept), and Keyita (the royal clan of ancient Mali).

While boro is listed as a Mande word (both Vai and Bambara), Turner fails to list this as the important Mandinka word for both medicine and poison. Turner identifies Kiang as an "ancient African kingdom," but fails to note this is a kingdom along the Gambia; he also lists Combo and Wuli but fails to note they also are important Gambian Mandinka kingdoms. Jarume is listed as a Fula word, but it is also an important village in Pakao Mandinka village system. Turner lists the word samba for elephant (sama or samo in Pakao Mandinka means elephant). However, in Pakao samba is widely used for bring or brought, take, or sent as in slaves brought (samba) to the land of the white men.

Turner does identify this last group of Gullah words with other, non-Mande, ethnic groups from western Africa. Yet these words were commonly used in Pakao. This too must be an example of a cultural convergence or overlap of similar words/sounds, that is an important linguistic concept for understanding how in the slave era in the Americas, a momentum could have been created for certain words, phrases, or ideas to be powered into broader English.

As noted, the lyrics for all of the several songs Turner identified are also Mande.49 Despite the lack of a Mandinka informant and contemporary dictionaries, Turner was able to write a special section on Mandinka influence, singling it out alongside several other notable African ethnic groups whose language heritage is seen in Gullah, but nevertheless understating the Mandinka impact.50

Curiously, a number of positive, uplifting words appear in this Mandinka list from Turner, as if the very need to survive included hopeful words—all the religious words, yes, "liver" as in good natured, purgatory (a hopeful and fundamental of pre-Islamic religion, giving people a [End Page 342] second chance), and shame (the positive Mandinka quality needed before penitence allows forgiveness).

Some words are double-edged. These include boro (the word for both poison and medicine), spear and war (both defense/offense, but often a precursor for enslavement), and the intriguing word samba, which in Mandinka usage implies both voluntary ("to bring") and involuntary ("brought") as if to recognize that they both sent their own people into slavery, and were also taken their against their own will. We can only guess which words from Turner's list were non-Muslim Mandinka names. There are at least several; since the Pakao Mandinka are fully Islamized today, their pre-Islamic culture had to be inferred during my fieldwork from their witchcraft beliefs, circumcision rituals and songs, and ethnomedicine, among other things.

Turner tells us that many of these African words were used among the Gullah as personal names or nicknames spoken semi-privately among themselves as a language kept secret from the outside world. I can vouch for this, remembering my 1955 visit as a child to Sapelo Island with my father, a veterinarian called over to vaccinate horses, and not being able to understand a single word of Gullah spoken in our presence. Sapelo, where Bilali lived, remains a Geeche (Gullah) heartland. Turner explains that the Gullah of the 1930s spoke more understandable English to outsiders, but that the more he got to know them over the years, the more they used African words, obviously doing so among themselves.51

If the Mandinka Gullah words Turner lists have anything in common, one can imagine it is their everyday importance back in west Africa, as if they became in America haunting recollections too precious to lose—village names, religious words, personal and given names, clan names, and totemic animal names. Turner went way beyond proving that Gullah was not primitive pidgin or baby-talk. He showed conclusively that Gullah was a heavily west Africanized Creole, and also a new language. He also allows us to infer how at least several English words commonly used in America today might have had African origins. Among such words with at least some Mande influence, are kunu meaning boat (Bambara), tote meaning to carry or lift (Mandinka and other Mande/non-Mande languages), yam or yambi meaning sweet potato (Mandinka and other non-Mande languages; in Pakao, nyambo).52 and bubu meaning any insect [End Page 343] whose wound is poisonous—thus a small wound (from Mandinka and other Mande/non-Mande languages). Gullah words heard in English, where Turner finds no Mande connection, include tabi meaning a building material (tabby) (from Wolof and other non-Mande languages), gumbo meaning gumbo (from Tshiluba in Congo and Umbundu in Angola), bidi bidi meaning itty-bitty (Kongo, a non-Mande language), and gola or gula, meaning Gullah (either a Liberian or an Angolan ethnic group and language).53

Turner allows us to glimpse the process of Africanized thinking and culture seeping into Southern English and from there into mainstream American English. He forces us to go back and take a second look at American English, and start asking deeper questions about its African content. One west African linguist who has done this was David Dalby, among the earliest to point out that the widespread traditional Mandinka usage of "OK" mirrored its similar usage as one of the most characteristically American words in existence. Therefore, Dalby suggests, the very American expression "OK" must have seen usage first among Mandinka slaves in the South, who passed the expression on to the rest of us.54

In my fieldwork in Pakao, I found the Mandinka expressions OK, OK kuta and OK kuta bake (OK, very OK and very, very OK) to be widely used.55 The Mandinka signature on this expression, accenting heavily the second syllable, and often using the expression with the common Mandinka words kuta and kuta bake, help convince me this is not some absorption from twentieth-century America, but rather a descendant of the African precursor to U.S. usage. Even if a telegraph operator helped put the expression into common usage in America, then the expression could have been reinforced by usage among Mandinka slaves and their descendants, in the kind of cultural convergence already discussed above for mansa and massa. Turner himself does not single out "OK" as one of the Gullah expressions. It was so common he may not have thought to include it. [End Page 344]

However, Turner's discussion of the west African syntax in Gullah speech patterns provides a model for thinking about a west African derivation for other expressions commonly associated with Southern English. The widely used "y'all" may be another example of a cultural convergence, in this case between the English "you all" and the Mandinka "al," meaning "you all," or "y'all" and often followed by a verb. Thus the Mandinka say al ta for "Y'all go" or "Y'all git." They say al ku for "Y'all wash" and al jinan for "Y'all come down here." See this latter expression in Kadri Drame's account of Deskaleri the Mysterious.56 The Mandinka also use fo as their word for "for" in the sense of "until," for example, "I went fo the house" as in Southern diction. In his tale about "The Bwa or Cannibal-witch, Kadri Drame says that djinns "can only harass someone until [fo] their time of death has come."57 Fo also would be an example of a cultural convergence. Several of the Mandinka legends in Djinns, Stars and Warriors also use quotations one after another in rapid fire, preceded by "he said/says" or "I said/say," which was also a feature of Southern storytelling that I heard growing up.

X

The little known ante-bellum memoir of Ophelia Troup Dent of Hofwyl-Broadfield Plantation in Glynn County near Brunswick, Georgia, tells us that her slaves used "My little aunt" to address a wet-nurse of presumably lesser importance and age, and "My big aunt" to address the main female house servant.

Writing during her old age in 1902, Dent says she especially remembers two of her grandmother's women slaves:

one, a small brown woman who nursed all the babies born in our house for a month. She had the care of the old Broadfield House (not the work), which was occupied by my father and uncles, our headquarters being Darien [near Hofwyl-Broadfield.] She was called "My Little Aunt" by our servants; but the big brown woman, who ruled our yard with a rod, was called "My Big Aunt.'" We children, and everyone else I knew, [including dozens of known slaves and slave descendants on this plantation] except my father and mother, called her "Mom Betty." She carried the keys when my mother was confined to her room, and in the spring made us [End Page 345] sassafras beer as in Charleston. She was the most scornful woman, black or white, I ever knew. She took care of the Darien house in the summer. She lived to a great age and died at Broadfield during the war.58

While a great many west African kinship concepts may have converged to produce this vernacular system, I must at least point out that Mandinka women in Pakao commonly used the word ba for mother to describe an important village leader such as a circumcision queen. Chief Fode Ibrahima Drame spoke about one such woman, Ture Nyako, in Pakao's Dar Silame. "All the women of Dar Silame chose her as their common mother [ba or baa]."59

In addition to "Mom Betty," the Dent slaves used expressions like "My Big Aunt" and "My Little Aunt;" would they have also said, "my big brother" or "my little brother" or "my big sister" or "my little sister?" Such expressions are in wide use in Southern English. Both Ophelia Troup Dent and her slaves seem to have used "big" and "little" to distinguish kin on the basis of relative age and importance. This was done among the Pakao Mandinka to distinguish between older and younger brothers, sisters, and other relatives with the widely used kinship terms koto or doko, (older or younger sibling). Pakao Mandinka also usually preface their use of kinship words with "my" (n), as in nba or nbama for "my mother" or nkoto for "my big sister, or "my big brother" or ndoko for "my little sister" and "my little brother." "Little" and "big" are west Africanized ways of translating "younger" and "older."

The expression "Mom Betty" is especially fascinating. Among the Mandinka, relatives through the mother, especially the mother's brother, are more important. In the Mandinka kinship system, young men try to marry their mother's brother's daughters, or matrilateral cross cousins (i.e., to marry any woman with the same clan name of the husband's mother). The Mandinka kinship vocabulary favors this preference, because the Mandinka word for mother's brother, mbaring, is also the word for father-in-law, so that the father of every bride in effect also becomes the husband's mother's brother, even if the preferred kinship did not exist before the marriage. This Mandinka kinship system, favoring the "mother" idiom and preferred matrilateral kinship in a man's marriage [End Page 346] partner, is quite old. Ibn Batuta visited Mali in 1352 and reported a similar though more radical matrilateral kinship system in which men claimed descent, not through the father, but through their mother's brother. A man's heirs were his sister's sons, not his own sons.60 In this sense we can see hints of an ancient Mande kinship system pushing through slavery into a Southern idiom, into the fabled "black Mammy," influencing the use of the maternal "Mom Betty" by Ophelia Troup Dent and her family's slaves.

XI

When we consider Drums and Shadows, the classic WPA study of Gullah religious beliefs in South Carolina and Georgia from the 1930s, collected/edited by Guy Johnson, Mary Granger, and others, Mandinka culture and Mande culture broadly reverberate on seemingly every page. At the same time, what cultural features appear Mandinka or Mande, such as drumming, might also be similar to cultural features of other ethnic groups, including those adjacent to the Mandinka in Senegal and Gambia such as Wolof and Fula, or distant groups including those from Ghana, Nigeria, Angola, and elsewhere, mostly on the western side of Africa.

Drums and Shadows relates how the Gullah drummed in church and also to communicate with each other.61 For example, they "beat the drum signaling them to gather, then all sing and dance in a circle to the accompaniment of the drum."62 Gullah women often used to dance in a circle to drums while clapping, like Pakao Mandinka today.63 Mandinka villagers still beat the tabala or large bass drum to summon people to important funerals or meetings; the jembe and hourglass squeeze drums called tama (tantango) are used for dancing, and when these are not available, a large gourd bowl is turned upside down in a tub of water and used as a drum. While drumming no longer occurs during mosque because of Islam's conservatism, the men and boys of Pakao during my fieldwork would sing [End Page 347] Muslim songs all night, accompanied by drumming, once the harvest was gathered. In the legend "Ture Nyako and Her Time," Fode Ibrahima Drame relates a traditional tale where an imam Kang Siaka said "don't bring in jembes [tantango.] But the village chief said: bring on the drums[tangtango]."64 In his tale, The Fall of Kunkali, Drame similarly associates the demise of the village with too much drumming (tangtango).65 Tam-tam or tantango, meaning drum or drumming, are Mandinka words that may have helped give us "tam-tam"or "tom-tom" in English.

Setting aside the question of similarities with the culture of other African ethnic groups, numerous additional similarities with Mandinka culture abound in Drums and Shadows. The handmade Gullah banjo, figure IIIa, looks like a Mandinka halam, also evoking what Drums describes as a gourd ("goad") guitar.66 These Gullah gourd guitars or banjos may have been influenced by either the Mandinka halam, the 21-string Mandinka gourd kora, the Jola akonting mentioned above, and various other musical instruments from the region of Senegal and Gambia. When referring to the "guitah" or banjo, Drums says that the local people "makes em from goad," using the plural "em," implying multiple, commonly-made instruments. The Gullah goatskin covered log drum, figure IV d., evokes both the Mandinka tam-tam or jembe used more for music or dancing and the larger, deeper-sounding tabala used to summon villagers.

Elsewhere, Drums reports Gullah baptismal candidates being dressed in white robes, and wading into the river to be immersed behind a preacher with a long robe.67 Such Gullah/Geeche baptisms must have seemed culturally logical to the Mandinka, who were used to the traditional white circumcision costumes and and also to"riverwash" (batakuo), when for the first time in a week after circumcision, the male and female novices ritually bathed in a river away from the village. Drums reports Sapelo Island Geeche oral traditions about the piety with which the Muslim Bilali and at least one of his wives prayed, including a repeated use of the Muslim word Ameen to punctuate prayers.68 Bilali's piety obviously impressed other slaves on Sapelo and perhaps on the mainland. [End Page 348] Does the tradition of an Amen corner in Southern black Christian churches owe some of its piety to the Mandinka Muslim tradition of punctuating prayer with Ameen, an Islamic and Arabic-inspired Mandinka word?

When Johnson says Gullah traditions report several slaves who flew back to Africa, is this not consistent with the widespread Mandinka belief that people and various spirits can fly or that people can change their shape into animals that can fly?69 One Gullah informant, George Smith, also reported fox and rabbit stories recounted among his people to their children.70 Similar traditions might have influenced Joel Chandler Harris' Uncle Remus tales. I recorded several "Hyena and Hare" (Br'er Fox and Br'er Rabbit) stories in Pakao, where these creatures were referred to with the alliterative words suluo and sula." Since the Mandinka don't have a fox, their equivalent would be the hyena, suluo.

Several Gullah informants report the affliction, common among the Mandinka, of being "ridden by a hag," described as being short of breath and feeling the sensation of being pinned to the bed by a mysterious force. Drums associates this with a Mande ethnic group, the Vai, but one of my Mandinka informants Kadri Drame also reported it from Pakao in southern Senegal, a few hundred miles north of the Vai.71 Drums reports a fear of owls among the Gullah, as if these birds are messengers of death and the very incarnation of evil; a similar fear and belief is widespread among the Mandinka.72 Pakao Mandinka reported an almost phobic fear of their bwa or cannibal-witch, and explain that bwa is also their word for owl, just as it was in the 1730s when Francis Moore so reported in his Gambian Mandinka word list. During my fieldwork, if a Pakao Mandinka heard an owl screeching, he or she went inside mortally fearing imminent death to themselves or a close kinsmen.

Throughout Drums and Shadows are reports of "root doctuhs" putting evil charm medicine or "conjuh" on people, causing death and disease that can only be undone by the greater good medicine, also "conjuh," of another "root doctuh."73 This pattern is similar among Pakao Mandinka, where marabouts or Islamic priests/witch-doctors make and unmake evil spells with their written amulets (safo), which are also considered [End Page 349] boro (the Mandinka equivalent of "conjuh"—either medicine or poison.)74 I also found during my fieldwork with the Mandinka an abundance of healing plant medicine and evil fetishes, such as bunyu furo made of cracked pottery and chicken feathers tied onto a millet stalk, evoking both Gullah remedies and the lethal spells of their marabout-like "root doctuhs." The parallel between Gullah religion, with root doctuhs and charms from various objects, and Mandinka marabouts and their written Islamic charms, causes me to wonder if Gullah religion of the 1930s is a glimpse back in time to an eighteenth-century pre-Islamic Mandinka religion, when importing African slaves into the US was widespread and legal. The danger posed by "root doctuh" witches among the Gullah, is paralleled throughout the discussion on witchcraft in Djinns, Stars and Warriors, where the Pakao Mandinka inhabit a similarly dangerous world invaded by various djinns (jinno) and cannibal-witches (bwa), but mediated and protected by all-seeing wizards (kumfanute) and marabouts.75

Mandinka Muslims entering North American slavery in the eighteenth and nineteenth centuries could easily have believed in, and contributed to the Gullah witchcraft system Drums recorded. Telling us why, the marabout Fodali Cisse discussed the flexibility of Mandinka Islam in accommodating witchcraft and other non-Islamic beliefs: "The Koran says these . . . are illicit beliefs but we humans don't reject these beliefs because they are our custom." In his account of the origin of the bwa or cannibal-witch, Cisse says: "Let us not reject the word of the Koran, but let us not follow it too closely."76

XII

Turner's Africanisms reminds us that while the Mandinka may have been a significant influence, numerous other ethnic groups from western Africa also left some linguistic imprint. At least one scholar who recently evaluated Turner's material, Frederic Cassidy, "found Congo-Angola elements strongest in the word-lists and Nigerian elements stongest in the texts."77 However, as noted above, Turner relied solely on mostly older Mandinka dictionaries, and not on a Mandinka informant from the Gambia River area or Pakao. [End Page 350]

Betty Kuyk notes that between 1733 to 1807, two-fifths of slaves imported to South Carolina were from Kongo groups.78 She proceeds to analyze Gullah culture of the Sea Islands of Georgia and South Carolina in terms of Kongo influence, and finds a impressive array of examples. These include secret societies, "big" as eldest, multiple and often secretive names, white symbolizing purity, traditional baptism, smaller prayer houses on many plantations before and after the Civil War, white gowns and headcloths, hags, intiation novices and a palm lodge, processions to cemeteries, masked figures, and owls as cultural symbols.79

However, such cultural features are also notable among the Mandinka: the smaller, founding mosques or misero in villages as a prayer-house model on a plantation; the kangkurao masked figure; circumcision lodges made of millet stalks; novice costumes and headcloths of white, riverwash as a phase of initiation; hags; a fear of owls; etc. This is in no way intended to undermine Kuyk's analysis, but more to suggest how cultural convergence may be at work here, with cultural survivals perhaps occurring more readily where there is overlap in the cultural features of ethnic groups in slave societies. Yet tones favoring Mandinka or Kongo influence can be identified. While I see as distinctively Mande the Gullah fear of owls as "messengers of death," Kuyk would see owls in Kongo society as messengers and symbols or "old-time folks," something perhaps equally appropriate in viewing the Gullah.80 Even if Kongo slaves were smuggled into the South disproportionately after 1808, the Mande preference noted by Curtin and Pollitzer, established mostly before the termination of legal importation from Africa, raises interesting questions not just about vocabulary, but also the very accent of Southern speech.

XIII

My first insight into the possibility of significant Mandinka content in the Southern accent occurred in one memorable conversation in Ziguinchor during 1972 with Buli Drame, the Mandinka from Suna Karantaba who guided me to the four villages I emphasized in studying Pakao. We proceeded to converse in French and he asked where I was from. After I told him, he slowly repeated after me, "St. Simons Island," pronouncing the words with such a strong Southern drawl that a chill ran up my spine. After years at college and graduate school away from the South, my own Southern accent had mostly disappeared. Yet Buli pronounced these and [End Page 351] other English words with a strong, seemingly perfect Southern accent, certainly an accent of the Georgia coast where Africanisms of The Gullah Dialect and Drums and Shadows both suggest a strong Mande influx and influence. One can debate how much a coastal Georgia accent resembles variable accents elsewhere in the South, but the accents of Charleston and coastal South Carolina and Georgia, spoken by both slaves and elite whites, were established before much of the inner deep South was settled.

This is not to say that a British accent or accents from African groups other than the Mande are not also present in certain Southern accents. Several informants from the 1930s in Drums and Shadows, from different ethnic groups as far south as Congo, a long way down the coast from Mande groups, note a strange system in which red flags were used, often hoisted onto slave ships anchored close to shore, as a method for attracting and capturing themselves or other unsuspecting children.81 Because these informants would have come from the very end of U.S. slave importation from Africa, Drums and Shadows perhaps implies this wildly random tactic was employed in the latter stages of the trafficking, as demand continued, but African importation into the U.S. had become illicit and, as Kyuk notes, many Congo were imported into Georgia. Buyers during the illegal era clamored for slaves, and slavers were so desperate they would resort to any measure, including red flags, to get captives on board regardless of ethnicity. After 1808 the old system of ethnic preferences in the slave trade began breaking down.

In any event, after that conversation with Buli I began to visualize and hear a heavy Mandinka content in the Gullah accent and thus in the "Southern accent" with all its variety. Pollitzer's slave importation demographics above favoring the Mande regions of Senegambia, Sierra Leone and the Windward Coast during the middle period (1749-87), and his literal analysis of Turner's Africanisms, showing the collective importance of Mande groups in Gullah speech, tends to support the idea of a predominant Mandinka and Mande content in the Southern accent, with the various other accents layered in (even without Mandinka informants identifying additional words, or the concept that the Mande influenced nearby ethnic groups in West Africa). Accent follows the vocabulary and demographics consistent with a Mande preference in Charleston and Georgia.

In various locales in South Carolina and Georgia, slaves so outnumbered white people, it is inconceivable for white English not to have been influenced by a West African accent. Turner noted some sections of South [End Page 352] Carolina where black families outnumbered white families twenty to one.82 Thomas Spalding's grandson, the ex-Confederate Captain Charles Spalding Wylly, wrote that the ratio on Sapelo Island was one hundred slaves to one white person, and asserts that these slaves had close, family-like relationships with their owners, implying close, verbal exchanges. "I have so often referred to the slave that I think it may gratify curiosity to tell in what manner these men and women fresh from Africa would with any safety be taken into the life of the family where in all probability there were not three white men to three hundred of their own race."83 Parrish notes there were 4,000 blacks and only 700 whites in Glynn County in 1845.84 A visitor to South Carolina in 1737 found the area more resembled "a negro country" than one settled by "white people," while the first federal census of 1790 established that 43% of South Carolina population were black slaves, compared to the national average of 18%. While the slave population in America declined to 13% (4,000,000) in 1860, South Carolina's slave population the same year had risen to 57% with even higher concentrations in the influential low country.85

Slave purchasers in the low country slightly preferred Mande not just for their rice farming knowledge and other factors, but once Mande came in sufficient numbers, they could communicate with the Mande slaves already working on plantations. Implying this possibility, Captain Wylly wrote a fascinating memoir detailing a training system for African slaves that is chilling for its racism and deculturization, suggesting a highly non-random process concerning the ethnic groups of slaves, at least for his grandfather, Bilali's owner. Wylly thought he provided a veritable linguistic blueprint for how the African-born slaves were gradually taught English. However, in so doing he inadvertently explains how a Mande accent might very well have entered Southern English, especially through the slave drivers, who were often African born leaders among the slaves, in charge of training the newly imported slaves.

After the African slaves were bought in the Charleston market, "the newly purchased were transferred at once to the plantation. Here always would be found a number of men and women acquired in former years who belonged to the same race, frequently of the same tribe and speaking the same dialect, or at least capable of making themselves understood." The African-born slaves were then assigned in groups of ten to a "driver" or leader "chosen for his ability to command and his fluency in speech."86 [End Page 353]

In this transitional, learning, period the men, women, and children were separated from each other and assigned leaders of their own gender and relative age. Gradually, they were taught English and the work of the plantation and rewarded for good progress with extra food. The African driver lived with them, talked and walked with them. No work was yet expected. The same method prevailed in the taming of the women, boys and girls. Meat was given out at the request of the trainers, or coaches, as I should this day call them. Fish, crabs and such stuff they caught for themselves under the eye and teaching of their constant guide and watchful guard. After a tutelage of perhaps three to five months they were assigned to work not requiring skill but only manual strength, such as the gathering shell for the burning of lime, the mixing of sand, lime and shell into concrete in the mortar beds [tabi or tabby from the Wolof word, according to Turner]—still under the eyes of their teacher—and transferring in hand-barrows of the concrete to the moulds which were slowing growing into the walls of house, stable, or barn. In twelve months they were generally, as it were termed, 'tamed,' and had acquired enough of the English language to be understood and to understand when spoken to. Then, and not until then, did their master begin to notice their personal qualities and abilities and assign them to duties which they seemed best fitted for.

The second year of the 'new' Negroe's development usually found him with a gang of thirty assigned to the regular labor in the fields. One third of this gang would be men and women of his own race [ethnic group?] who had graduated years before from the same school that he was now entering. Here commenced the imitation, and long before the expiration of a year he had learned many things, for his teacher, called locally his driver, was always near to direct, instruct and command. He had been taught to rise when the conch blew in the morning, to use his hoe as he saw others use it, to come and to go when told to do so, to stand still when a white man spoke to him, and in most cases by the end of the second year the "Jack, new negro" that had marked his place and value on the plantation books had been altered into "Jack—African born."—an immense change in his life resulting: i.e. a task each day, which when completed gave to him all the remaining hours to do with as he pleased; a house and garden lot, when he had chosen a wife, and on Sapeloe freedom to fish, hunt, oyster and crab, always with a reasonable restriction. No one now living can imagine with what freedom and lack of danger was the African of 1787 to 1806 trained into the most efficient but easiest managed laborer in the world.87 [End Page 354]

Spalding had about 400 slaves at any one time, and during his lifetime gave over 1000 slaves, and the lands they worked on, to his two surviving sons and four married daughters, disseminating the linguistic influence and west-Africanized accent of his system into the Georgia coast and the South, presumably alongside a number of similar examples from other plantations.88

Despite slavery's hodge-podge mixing of ethnic groups from Africa, evidence of a Mande preference among the Gullah finds additional support in the memoir of Sapelo Island's Gullah, or more correctly, Geeche writer Cornelia Bailey, who uses styles of basket-making, "Mende ring shout dancing," linguistic and other evidence to conclude that the Mende from Sierra Leone were a strong ethnic component of the heritage of African-Americans living on Sapelo Island. What Cornelia's people called "fanners"— shallow, flat baskets used for threshing rice—the Mende call fantas.89

XIV

More than a few of the eighteenth and early nineteenth century slave owners on the coast of South Carolina and Georgia had business and personal associations with the Caribbean. Bilali's owner Thomas Spalding visited the Caribbean, had family ties there through his father-in-law, and is thought to have purchased slaves there. One of Bilali's daughters, Magret, was said to have been a slave in the Bahamas before she was brought to Sapelo. Magret passed down some untranslatable words through her daughter Cotto to Cotto's daughter Katie Brown.90 Katie reported Magret's words as "mosojo" or "sojo" for pot, "deloe" for water, "diffy" for fire, and "saraka" for the flat rice cakes made on the same special day each year, suggesting that Magret was Muslim like her parents Bilali and Phoebe, who both "prayed on the bead" and said "ameen" as a way to punctuate and agree to each other's prayers. "Deloe" and "diffy" appear to be French de l'eau and du feu, suggesting that Magret was from the French Caribbean, and that she spoke a French Creole, perhaps a Haitian Creole.

When I raised the French Creole possibility with the Mandinka scholar David Gamble, he replied that the Wolof use the word sujer or soojer for iron pots, most likely imported from Europe. When Gamble checked an old French dictionary, he found the word chaudière for large pot or cauldron, [End Page 355] and thinks this may have been in use in the eighteenth century and was the source of Magret's word for pot.91 Gamble also points out that Magret's saraka is the Bambara word for rice cake, sada and sadaji (singular and plural) are the Gambian Firdu Fula words and are from Mandinka, sadaga is Futa Jalon Fula, sadarha is Khassonke, sarax is Wolof, and sadaa is Gambian Mandinka (as it is in Pakao). Was Magret of Bambara (Mande) descent? If not, she was at least speaking a Bambara-influenced word for rice cake. The different vocabulary from sada suggest Mandingization of a religious word that comes from the Arabic word sadaqa, for alms (most Mandinka religious words are from Arabic).

Interestingly, Magret's daughter is named Cotto; Koto, meaning older sister, is a widely-used Mandinka woman's name. If Bilali was an African-born Fula from Timbo, he might have been able to speak some Mandinka, since many Mande live around Timbo, and gave the Mandinka name Koto to his grand-daughter. Bilali gave the name Bentoo, suggestive of the popular Mandinka name Binta, to another daughter, and two Muslim names, Fatima and Medina, also possibly Mandinka names, to two other daughters.

The explorer Mungo Park provides a haunting portrait of a significant Mandinka presence in Caribbean slavery, and how a slave ship from the Gambia River funneled slaves into the Americas—slaves who might have ended up on the plantation Bryan Edwards wrote about or on Thomas Spalding's Sapelo Island. Trying to get back to England after having discovered the Niger River, Park waited around for weeks in 1796, hoping for a ship to return him directly. Finding none, he finally was obliged to book passage on the most likely ship to take him out of Banjul—a slave ship bound not for England but Charleston, and named for this city, at a time when significant trans-Atlantic slave shipments from the Gambia region took place. As the voyage turned out, the ship leaked so badly it almost sank following prevailing wind patterns into the Caribbean, and had to unload its human cargo in Antigua long before it reached Charleston.

Park had learned to speak Mandinka during his travels upstream along the Gambia toward the Niger River, and conversed with the slaves during the crossing, compelled by their suffering to serve as their doctor. He estimated about 25 Muslim slaves in a cargo of 130 that included a great many Mandinko and at least a few from a failed jihad against a Wolof ruler. Coming up the Atlantic coast from the Gambia River, this ship docked at Gorée to take on supplies, and then headed across the Atlantic. 22 of the slaves died before reaching Antigua, several before even reaching [End Page 356] Gorée. Impressed by the great numbers of Mandinka slaves arriving in the Americas, Park chose to include a Mandinka wordlist in his famous book as a guide for people needing to converse with the great number of African-born slaves found at that time in the Caribbean. "The following questions and answers may be useful in the West Indies," he entitled his word list, as if it were common knowledge that large numbers of Mandinka slaves were shipped there. The vocabulary list is composed entirely of Mandinka words and phrases, several pertaining to medical issues.92

A growing body of field work and anthropological studies suggest, as Park implied, that a major influx of Mandinka or "Mandingo" slaves poured into the Caribbean during the slave era, and, aided sometimes by their Muslim beliefs, could attain positions of leadership in slave society, making them forces to be reckoned with by white settlers. Nishida reports that in Trinidad an "urban Mandingo community, whose members were Muslims, used part of its considerable economic assets to function as an emancipation society. As in the case of nineteenth-century Salvador in Brazil, some Mandingoes in Trinidad became slave owners. (Carl Campbell's implicit assumption is that the Mandingoes, who showed strong ethnic identity, owned non-Mandingo slaves and some of them traded in non-Mandingoes for their freedom)."93 The Free Mandingo Society on Trinidad helped convert a whole regiment of West Indian blacks to Islam.94 Nishida elsewhere reports that the jailing of an important, African-born Muslim slave leader named Pacifico, or Bilali by his Muslim name, helped precipitate the slave revolt in Salvador in 1835 by a largely Yoruba group of Muslim slaves who had already tried unsuccessfully twice to rescue him.95 Pacifico's ethnic identity is not given, but "Bilali" (from Bilal, Muhammad's slave advisor and first muezzin) appears to have been a highly esteemed slave Muslim honorific name among the Mandinka. One slave Bilali mentioned in the oral traditions of Pakao during my fieldwork was identified with the additional honorific samanung, "hard-working" (literally elephant head: samanung Bilali).96

XV

Strong African ethnic identities in Cuba, wrote George Brandon, rightly or wrongly became the basis for stereotypes used by slave-owners in selecting [End Page 357] and purchasing their slaves. The Mandinka were "excellent workers." The Carobali were "proud," the Gangars "thieves and runaways," the Fanti "also runaways" and "revengeful," Ebos "less black . . . and lighter wool," Congos "short," and Lucumi "[i]ndustrious workmen." Slaves also used such classifications to guide relationships between subgroups on the same plantation or simply to help identify themselves. Montejo, a Cuban slave born in 1860, used such concepts to describe relationships among various ethnic stereotypes among his fellow slaves, who, depending on their ethnic group, could be hard-working, cowardly, or prone to run away. Lucumi (Yoruba descendants) and Congolese did not get along, for example. "The Mandingoes were reddish-skinned, tall and very strong. I swear by my mother, they were a bunch of crooks, too."97 Brown reports that neighborhood (cabilo) processions occurred in Cuban slave society, where the various ethnic groups could be distinguished by their appearance, movements, and sounds. The Mandinka and other groups such as Congo and Lucumi could immediately be singled out by their clothing and markings. "The Mandinka stood out for their sartorial luxury; wide silk pants, short jackets and turbans, all bordered with marabout (feather boa)."98

The Santeria priest Nicolas Angarica wrote in Cuba at some length that Ozain, the Santeria orisha or god of herbs and medicines, "comes from the Mandingas." George Brandon speculates that perhaps a Mandinka "with particularly impressive knowledge of Ozain's lore arrived in Cuba and was able to plant anew Ozain's worship . . . Ozain priests in Nigeria are simply not good herbalists."99 I found several herbalists among the Pakao Mandinka, often hunters but sometimes marabouts, who could each name scores of plant remedies, and that a few remedies, such as one for snakebite, were common knowledge and the subject of their own oral traditions.100

In another example where informants knew the ethnic identify of a cultural tradition, Scott Mahler, an editor with Smithsonian Press, told me he had heard directly in Cuba that the Mambo is from the Mumbo Jumbo secret mask society of the Mandinka. The Mambo was said to be a special section in Cuban music where in a transcendental moment mumbo jumbo is spoken. Mumbo Jumbo, a secret mask society closely related to the kangurao, and probably the word jumbo as well, were also [End Page 358] introduced into English by Francis Moore's 1738 work about the Mandinka, but the terms may already have been in use among African slaves in the Americas before then; it certainly appears so in Cuba.101

Masked figures or dancers show up in a number of disparate places in the New World as part of the broad slave legacy, and in some cases are traceable to the Mandinka. Judith Carney told me how a few years ago on New Year's eve, she witnessed the kangkurao mask and dance in Triunfo (or Trujillo), Honduras, and was astonished that local people did not know its origins.

XVI

Mandinka masked figures like the Mumbo Jumbo or its close cousin the kangkurao also possibly show up in Haitian Vodou. Sidney Mintz cautions me not to push this too far, but it does seem more than a mere coincidence that an important Vodou divinity, Gran Bwa, when represented as a masked dancing figure covered in leaves, looks very much like some representations of the Mandinka kangkurao along the Gambia River, where dancers in the latter secret society also cover themselves with a combination of bark and leaves. In Vodou Gran Bwa is associated with medicinal healing, the forest, and initiation.102 Also, like the Haitian Gran Bwa, the Mandinka kangkurao is associated with initiation and the forest, although only tangentially to medicinal healing.

In yet another potential cultural convergence, the Haitian word for this divinity, Gran Bwa, might derive either from the French bois for wood or forest or from the Mandinka bwa or bua, reported since the 1730s as the commonly-used word for both witch and owl. I translate Mandinka bwa as cannibal-witch, a widely-believed fundamental of local witchcraft belief in Pakao, and include several mentions or a description of it in Djinns, Stars and Warriors.103 Bwa, often human witches transforming [End Page 359] into animal shapes, is such a common Mandinka word and widespread belief that it is hard not to see some linguistic hint of it within the Haitian Gran Bwa. This Vodou divinity might thus be seen as an amalgam of the Mandinka name for cannibal-witch, bwa; the French word bois; the leafy appearance of a Gambian type of kangkurao, and the protectiveness toward initiates of either a kangkurao or fangbondi variant.104

There is another striking parallel in the pervasive Mandinka belief in jinno, djinns or spirit doubles, and the Haitian Vodou term ginen, described as a place identified with "spirits," as well as a sort of idealized Africa or Guinea.105 Ginen in Haitian Vodou is also a place beneath the sea, a kind of spirit-world watched over by the sea-god Agwe.106

Along with Mandinka cultural influence come additional hints of their political leadership. David Geggus notes that there are both Mandinka and Kongo-Petro interpretations (cultural convergence?) of the legendary Bois Caiman ceremony that paved the way for the Haitian revolution, when dissident slaves gathered in a secret forest ritual, sacrificed a pig, and drank its blood. The Mandinka interpretation (by Diouf) asserts that the leader Boukman and the high priestess Cecile Fatiman were Mandinka Muslims. To Geggus this assertion might be contradicted by the Muslim proscription against hogs.107 However, through much of the last millennium of their history, Mandinka non-Muslims lived side by side with Mandinka Muslims, sometimes in neighboring hamlets or villages, and the Muslims incorporated certain non-Muslim beliefs such as purgatory. The Mandinka hunter Baba Sagnan demonstrated for me his prowess as a hunter in 1974 by shooting a boar pig simply because it was a worthless beast, and then showed me how he could track it by following tiny—to me nearly invisible—flecks of blood on the leaf bed of the forest floor. Finally, if Cecile Fatiman were Mandinka, she would have experienced, and perhaps studied, the commanding and essentially non-Islamic powers of her village circumcision queen in West Africa. [End Page 360]

XVII

There are also tantalizing hints of an older Mandinka influence in Brazil—a country whose southern latitude linked it more closely with large ethnic groups from Nigeria, the Congo region, and Angola.

Similarity to the kangkurao can be seen in one of the best-known gods or Orixas from Brazil's popular, Africanized Candomble religion in a figure called Omalu or Obalouaie. This Orixa is represented by a dancer inside of a haystack being twirled on a pole. The kangkura, also very frightening, is sometimes represented in a similar fashion—for example, by the Senegalese national dance troupe. However, there are a number of other African ethnic groups, at least as far south as Zaire, that have a twirling haystack masked dancer, thus invoking the idea of a cultural convergence to create Omalu. This most terrifying of Orixas is sometimes called "grandfather" or "the old one," as if a reference is being made to the god's origins from one of the larger and earlier west African ethnic groups whose members were sold into Brazil.108

A book on Brazilian folklore called Brasil, Histórias, costumes e lendos by Aleceu Araújo and José Lanzellotti shows other possibilities for Mandinka influence, including a popular dance to drumming, called Jongo or Jongo Africano, which started among African descendants in Rio, clearly invoking the Mandinka word for slave jungo or jongo. The book includes a picture of Omalu with the haystack headdress and wearing cowry shells used among some Mande groups. This book also includes several additional male and female Orixas portrayed as wearing white (Iemanjá, Oxalufam, Oxumaré, Oxalufá, Oxodiã, Iaõ and Oxum), all invoking the color white and style of Mandinka Muslim robes and other Muslims from West Africa.109

Another hint of Mandinka influence is in the word samba, the name of the well-known Brazilian dance so vividly on display at carnival. Wafer [End Page 361] explains that traditionally samba dancing took place in a circular style, the circle samba or samba-de-roda, and had an important ritual role in Candomble, leaving little doubt of this dance style's African origin.110 In this Candomble format, which seems identical to the most popular dance style among the Mandinka of Pakao, one or two people dance in a circle of jubilant peers, and then summon one or two more dancers, who repeat the process. The Pakao Mandinka also widely use the verb samba (meaning send, sent, bring or brought) in conjunction with references to the trans-Atlantic slave trade. I often heard them teasingly threaten a misbehaving child with a warning: "M ba samba tubabodu" (I'm going to send you to the land of the white man). One can easily imagine how this Mandinka verb could have been used in Brazil's colonial slave era to refer to a dance brought by African slaves. Reinforcing this usage, in another potential cultural convergence, samba is also a first and last name among the Wolof and Tukolor. Michael Coolen tells me that his halam teacher was Abdulai Samba, and one of the most famous halam performers was Samba Jebere Samba. Also, the article mentioned above on the Jola akonting, notes that Samba is a common Jola family name. Finally, Pollitzer points out that samba means "to jump about" in the Tshiluba language. In Bobangi samba means "to dance the divination dance;" in other Bantu languages, it's meaning is related to worship.111

African dances certainly showed up during the slave era in this hemisphere, as noted in Johnson's Drums and Shadows. Lydia Parrish writes about Gullah/Geeche shout dances, including "The Buzzard Lope," in her Slave Songs of the Georgia Sea Islands. Patience Pennington writes that her young black women workers "with bare feet and skirts well tied up danced and shuffled the rice about with their feet . . . singing, joking, displaying their graceful activity." An accompanying illustration shows one of the young women with her arms straight out, as if moving them up and down, while pounding her left foot up and down, in a classic Mandinka dance form.112

XVIII

I have probably only scratched the surface in identifying Mande and Mandinka influences in the Americas. The presence of the Mandinka and their cultural legacy has been documented in Surinam and Mexico. Evoking [End Page 362] this legacy, a popular beach in the Mexican City of Vera Cruz (a former colonial center built up after Cortés landed there) is called Mandinka. Robinson A. Herrera told me how his sources on Guatamala assert that this country's national instrument, the marimba, came from the traditional balafong, the xylophone-type instrument constructed by Mandinka slaves. Paul Lovejoy told me he is working on an Arabic script document left by a Mandinka slave in Jamaica. Svend Holsoe told me that his research on St. Croix into some 16,000 ethnic identities written down in the slave-era church records for baptism notes the presence of a significant number of Mandinka slaves. Haitian Revolutionary Studies notes the presence of groups of Mandinka slaves among those loyalist blacks shipped out of Haiti by their French sponsors to other locations in the Caribbean.

Mandinka cultural survivals help us see the rich history of this particular ethnic group in a more ancient and geopolitical way, through the trade linkages of Manding across the Sahara in medieval times to the Islamization of west Africa, and through the horrors of slavery to the Americas. We must see the Mandinka from west Africa in a greater Atlantic Rim context, in which traces from their culture show up all over the Americas, especially in the United States, from jazz and rap to the diction of Southerners and the drawl of Southern presidents. The breadth and reach of Mandinka influence perhaps raises more questions than it answers. Will some Mandinka influence ever be found in tango, the name of Argentina's famed dance that supposedly originated in part from African slaves? The Mandinka word for the palm tree, under which the Jola danced and played their banjo, is tengo. Tantango is the Mandinka word for drum.

Clearly, "tango" was an important and widely used sound in the Mandinka language during the time of my research in Pakao. Beyond that, Argentine historical works on the tango, such as that by Benedetti, suggest an origin from West Africa somewhere between "Cape Verde and Dahomey."113 He points out that some early, nineteenth-century, singers of tango lyrics had one name, suggesting they were of west African slave origin, and calls for more linguistic research on the tango's potential African origin. More research is needed, for example, on whether cultural convergence ideas can apply to the tango. Does the "tango" sound exist in other West African languages besides Mandinka? What does the word mean, and how many slaves speaking those languages were sold into Argentina during the nineteenth century and before? Is the tango a purely [End Page 363] African invention. or more of a west African name based on a form of exuberant African dancing melded into forms of Spanish colonial dancing in Argentina?

In this paper I suggest the possibility of a Mandification not just of Southern English, but of Southern culture, both of which offer a compelling laboratory for linguistic and structural analysis. At least seven principles may seem to emerge from this analysis. First, cultural convergence increases the chances that words related in sound and meaning, used by a critical mass of people, win out and become absorbed in the cultural free-for-all environment of the South and the Americas generally. Such words as master/massa/mansa, y'all/ al, for/fo, jazz/jams/jong, and OK/OK, OK kuta and tom-tom/tantango help to open up this possibility for consideration.

Second, Mandification was already happening in Africa during the trans-Atlantic slave era, as the Mande began to gain the upper hand vis-à-vis neighboring ethnic groups through trade and war, perhaps amplifying their impact in the South and the Americas. The widespread use of Malê in nineteenth-century Brazil to describe Muslims is an example of this process, and so perhaps are the word combinations master/massa and mansa and jazz/jams/jong. We are told Malê (from Mali, Muslim) is derived from imale among the Yoruba, who received the word from the Mande in Africa. Massa (meaning mansa or king) is used by the Cassanga and Banol ethnic groups, who also have a word and mask closely related to the Mandinka Mumbo Jumbo. Mande slaves came to the New World with some ability to intercommunicate through some similarities in the various languages within the overall Mande linguistic group.

Third, cultural convergence applies not just to language, but also to cultural features, such as the Haitian Gran Bwa and similarities between the baptizing of Gullah initiates in white robes in the river, as compared to the white Mandinka circumcision costumes and the "riverwash" purification ceremony in their circumcision seclusion. Structural parallels between Christianity and Islam (e.g., heaven and hell, mercy, penitence, charity, devotion, piety, regular and visible practice, appropriateness of the color white) offered a coherent structure for the rapid and unique Christianization of African-born slaves or their descendants. The common values mentioned above for Islam and Christianity of course may have figured in animist religions too. As well, animist values seem to have already influenced Mandinka Islam before slaves took it to the Americas, with the Pakao Mandinka believing in everything having a spirit double, the sanctity of historical trees, and the second chance of purgatory—all different from mainstream Islam. The point here is that west African [End Page 364] Islam, with its coherent values derived from one book, the Qur'an, already had converted many (but not all) Mande people during the slave era, as well as certain non-Mande people. This commonality between west African Islam and Christianity perhaps amplified the impact of the Mande slaves in the Americas. What an irony that "the Amen corner" in the African-American church may descend from the "ameen" used to punctuate and emotionally ratify a Mandinka or Mande Muslim blessing or prayer.

Fourth, possible examples of Mandinka influence showing up outside the United States in places like Brazil, Haiti and Cuba, despite the major importation of slaves from other African ethnic groups, tend to give weight to, or even confirm, the potential for a Mandification of Southern English and culture. The pieces of the argument supporting significant Mandinka influence tend to reinforce each other and the whole.

Fifth, non-US examples of Mandification in the Americas also suggest this process may have taken some unique forms, just as if did in the U.S., and could prove a fruitful area for future analysis. For example, Reis tells us that when the rebelling Malê Islamists poured into downtown Salvador, in 1835, it was the first time Brazilians saw large numbers of people dressed in white in the streets, allowing us to infer at least one Islamic derivation for the whiteness of clothes worn by Brazilians during carnival, alongside traditional African sources for white clothes, including Mandinka circumcision costumes. In another example, Brazilians widely use the adjective inho (masculine, pronounced eenyo) or inha (feminine, pronounced eenya) at the end of nouns to modify them into "little" or "opposite," in words such as cafezinho ("little coffee" or espresso), pezinho (little foot), Ronaldinho (a personal name), camisinha ("little shirt," slang for condom), and so on. Is it coincidence or direct influence that the Mandinka use dingo (child or little) and ringo (opposite) to modify countless nouns, for example, baringdingo (mother's brother's daughter) to mamaringo (grandson or daughter). The Brazilians amplify this idea with a marvelous saying about pervasive African genealogy in their society: "Todo Brasileiro tem um pezinho in Africa" meaning "All Brazilians have a little foot in Africa."

While the sounds inho and inha existed in Portuguese, the use of these sounds is far more extensive in Brazil. Brazilian anthropologist Roberto da Matta suggested to me that an infusion of inho by early Portuguese explorers gave rise to the Mandinka ingo sound. However, the Mandinka word dingo is used so widely as an adjective for several words ranging from child to fruit that dingo seems more like an indigenous creation. If we accept a possible African and Mandinka influence for the wide usage [End Page 365] of inho/inha in Brazil, perhaps we must ask why ito/ita is widely used in Mexico, but curiously, not in the mother country Spain. Does this Mexican idiom represent slave and Mandinka influence?

Sixth, if analysis singles out Mandinka in the Americas, the same kind of analysis ought to be possible for tracing other African ethnic origins in this hemisphere. Betty Kuyk singled out Kongo influence among the Gullah. Paul Lovejoy focused more broadly on Yoruba influence in the trans-Atlantic slave trade, noting in particular the influence of jihad directly through captives and indirectly through refugees of jihads streaming toward the coast.114 The identification of particular ethnic influence also gains support in Cosentino, during a series of articles about Haitian Vodou. Sidney Mintz and Michel-Ralph Trouillot dissent by cautioning that "Vodou was created by individuals from many different cultures."115 But others, such as Robert Ferris Thompson for the Fon, Eve, and Aja, and Suzanne Preston Blier for the Eve and Fon, suggest that western African ethnic groups who influenced Vodou can be identified.116

Finally, where does cultural convergence in its broadest sense come from, as a commonality between European and African language, if not from some basic and ancient source of language? While learning Mandinka in the field, I was struck by this possibility. "Na si (Come sit)," Sanjiba Drame used to say when asking me to come talk with her beneath the low, smoky roof of her cooking house. "Na si ka cha (Come sit and chat),"she would add; na sirango ("come sit down on the stool,") she said, while pointing to a carved sirango (stool). The basic Mandinka herabe or ibe herato("how are you?") followed by heradro ("I'm fine") came to mind repeatedly as I rode a bicycle in 2003 on Cumberland Island in Georgia, and every Southern tourist I met said, "How ya doing? or "Hey, how are ya?" Is the structure of this Southern greeting a coincidence with Mandinka or direct influence, or commonality with some ancient proto-language? Already the field of Nostratics has arisen as a theoretical super-family of languages in which Indo-European is only one of six branches of a much larger language family. In the last fifteen years, linguists postulate an even more ancient language, the first language, Proto-Human, Proto-World, or Mother Tongue, probably arising in Africa, the continent where the earliest hominid skeletons have been found, and geneticists tell us that a precursor female hominid, the mother "Eve" of all humans, once lived.117 The commonality master/massa/mansa and or y'all/al may ultimately [End Page 366] prove to be part of this pattern, suggesting that deep and ancient linguistic sounds and structure may lie behind Creolization in the Americas and the Mandification of Southern English.

Bibliography

Angarica, Nicolas Valentin. 1955? [n.d.] Manuel de Oriate, Religion Lucumi. Cuba

Alexis, G. 1970. Lecture en anthropologie haïtienne. Port-au-Prince.

Araújo, Alceu Maynard and José Lanzelloti. n.d. Brasil Histórias, costumes e lendas. São Paulo.

Austin, Allen D. 1984. African Muslims in Antebellum America, A Sourcebook. New York.

Bailey, Cornelia Walker. 2000. God, Dr. Buzzard and the Bolito Man. New York.

Ball, Edward. 1998. Slaves in the Family. New York.

Bell, Malcolm. 1987. Major Butler's Legacy, Five Generations of a Slave-Holding Family. Athens.

Benedetti, Hector Angel. 1997. Antología de Tangos. Buenos Aires.

Brandon, George. 1997. The Dead Sell Memories—Santaria from Africa to the New World. Bloomington.

Brown, David H. 2003. Santeria Enthroned: Art, Ritual, and Innovation in Afro-Cuban Religion. Chicago.

Bühnen, Stephan. 1993. Ethnic Origins of Peruvian Slaves (1548-1650): Figures for Upper Guinea. Paideuma 39:57-110.

Campbell, Carl. 1975. John Mohammed Bath and the Free Mandingoes in Trinidad: the Question of their Repatriation to Africa 1831-38. Journal of African Studies 4:467-95.

Carney, Judith. 2001. Black Rice, The African Origins of Rice Cultivation in the Americas. Cambridge, MA.

—. 2003. The African Antecedents of Uncle Ben in U.S. Rice History, Journal of Historical Geography, 29:1-21.

Cassidy, Frederic G. n.d. Sources of the African Element in Gullah. Studies in Caribbean Language, ed. Lawrence D. Carrington et al. St. Augustine, Trinidad, 75-81.

Coolen, Michael T. 1991. Senegambian Influences on Afro-American Musical Culture. Black Music Research Journal 11:1-18.

Cosentino, Donald J. 1995. Sacred Arts of the Haitian Vodou. Los Angeles.

Curtin, Philip. 1967. Africa Remembered. Madison.

—. 1969. The Atlantic Slave Trade. Madison. [End Page 367]

Dalby, David. 1970. Black through White, Patterns of Communication. Bloomington.

Dent, Ophelia Troup. Memoirs of Ophelia Troup Dent. Archives of Hofwyl-Broadfield Plantation. 6 August 1902.

Diouf, Sylvane. 1998. Servants of Allah: African Muslims Enslaved in the Americas. New York.

Donnan, Elizabeth. 1930-35. Documents Illustrative of the History of the Slave Trade to America. 4 vols. Washington.

DuBois, W. E. B. 1946/47. The World and Africa. New York.

Edwards, Bryan. 1983[1793]. The History, Civil and Commercial, of the British Colonies in the West Indes, as excerpted in After Africa, ed. Roger D. Abraham and John F. Szwed. New Haven.

Fernandes, Valentim. 1951. Description de la Côte Occidentale d'Afrique (Sénégal au Cap de Monte Archipel). ed T. Monod, A. Texeira da Mota, and R. Mauny. Bissau.

Fletcher, Andria, 1977. The Gambia. Covina.

Geggus, David P. 2002. Haitian Revolutionary Studies. Bloomington.

Hecquard, Hyacinthe. 1855. Voyage sur la Côte et dans l'intérieur de l'Afrique Occidentale. Paris.

Jagfors, Ulf. Winter 2003-04. The African Akonting and the Origin of the Banjo. The Old-Time Herald, A Magazine Dedicated to Old-Time Music 9/2:26-33.

Jobson, Richard. 1968[1623]. The Golden Trade. London.

Johnson, Guy, Mary Granger et al. 1986[1940]. Drums and Shadows. Athens, GA.

Kuyk, Betty M. 2003. African Voices in the African American Heritage. Bloomington.

Lovejoy, Paul. 2003. "The Yoruba Factor in the Trans-Atlantic Slave Trade." Paper presented at the African Studies Association Annual Meeting.

Meireles, Cecília. 1983/2003. Batuque, samba, and Macumba Drawings of Gestures and Rhythms. São Paulo.

Mollien, Gaspard. 1967[1820]. Travels in the Interior of Africa. London.

Moore, Francis. 1738. Travels into Inland Parts of Africa. London.

Najman, C. 1995. Haiti, Dieu seul me voit. Paris.

Nishida, Mieko. 2003. Slavery and Identity, Ethnicity, Gender and Race in Salvador, Brazil, 1808-1888. Bloomington.

Park, Mungo. 1971[1799/1815]. Travels in the Interior Districts of Africa. New York.

Parrish, Lydia. 1942. Slave Songs of the Georgia Sea Islands. New York.

Pennington, Patience. 1913. A Woman Rice-Planter. New York. [End Page 368]

Pollitzer, William S. 1999. The Gullah People and Their African Heritage. Athens, GA.

Reis, João José. 1993. Slave Rebellion in Brazil, the Muslim Uprising in 1835 in Bahia. Baltimore.

Rout, Leslie B. 1976. The African Experience in Spanish America. New York.

Schaffer, Matt. 1975. Pakao Book. African Languages/Langues Africaines 1:96-123.

— and Christine Cooper. 1987. Mandinko, The Ethnography of a West African Holy Land. Prospect Heights, IL.

—. 2003. Djinns, Stars and Warriors, Mandinka Legends from Pakao, Senegal. Leiden.

Shevoroshkin, Vitaly. 1990. The Mother Tongue: How Linguists have Reconstructed The Ancestor of All Living Languages. The Sciences May/June:20-27.

Turner, Lorenzo. 1973. Africanisms in the Gullah Dialect. Ann Arbor.

—. 2002. Africanisms in the Gullah Dialect. Columbia, SC.

Turner, Richard. 1997. Islam in the African American Experience. Bloomington.

Vydrine, Valentin. 1998. Sur l'écriture mandingue et mande en caractères arabes (mandinka, bambara, soussou, mogofin). Mandenkan 33:1-87.

— and T.G. Bergman, Cartographer Matthew Benjamin. 2001, updated 2003. Mande Language Family of West Africa: Location and Genetic Classification. website of MANSA, the Mande Studies Association.

Wafer, Jim. 1991. The Taste of Blood, Spirit Possession in Brazilian Candomble. Philadelphia.

Washington, John. 1838. Some Account of Mohammedu Sisei, a Mandingo, of Niani-Mara in the Gambia. Journal of the Royal Geographcal Society 8:449-54.

Wepman, Denis, Ronald B. Newman, and Murray B. Binderman. 1976. The Life: The Lore and Folk Poetry of the Black Hustler. Philadelphia.

Wylly, Charles S. 1915. The Story of Sapeloe. Typescript in the University Libraries of Georgia, Marine Institute, Sapelo Island.

Footnotes

1. See Djinns, Stars and Warriors, Mandinka Legends from Pakao, Senegal, published by Brill Press in 2003, containing oral traditions I collected in 1972 and 1974 in the Pakao region of middle Casamance in southern Senegal. This volume is a companion book to my basic ethnography of the Mandinka first published in 1980 and kept in print since 1987. Of the many people who helped me with this article, I want to single out Michael Coolen and Judith Carney for special thanks. I'm also grateful to National Geographic and the Rhodes Trust for funding my fieldwork.

2. Curtin, 1969:156-57. His numerous sources include the work of Elizabeth Donnan.

3. Ibid.

4. Vydrine/Bergman 2001.

5. Pollitzer 1999:116-17.

6. Ibid., 148.

7. Ibid., 41.

8. Ibid., 41-42.

9. Ibid., 44-45, table 6. Pollitzer states (ibid., 37) further that, roughly speaking, Senegambia means Senegal and Gambia of today. "Sierra Leone" in eighteen-century English sources refers not just to that country of today, but Guinea-Bissau and Guinea, a small part of northern Liberia, and the Casamance River region of southern Senegal. Pakao Mandinka, though closely connected to Gambian Mandinka, may have been considered more from "Sierra Leone" than Gambia, although the frequency of travel between Pakao and Gambia suggest Mandinka slaves from Pakao may certainly have been shipped from the Gambia as well. The "Windward Coast" includes roughly Ghana and Ivory Coast, but the usage varied.

10. Ibid., 47.

11. Ibid., 108.

12. Ibid., 47.

13. Ibid., 57, quoting various sources including Ball 1998.

14. Pollitzer 1999:88-89.

15. Ibid., 37.

16. Ibid., 58, 56, 61, 55, 65, 113; Cable 1886:517-32.

17. Carney 2003:1-21; Carney 2001:41, map.

18. See the quotes from his grandson's journal cited below.

19. Kuyk 2003. See her broad analysis of Kongo influence on the Gullah. Pollitzer 1999 for Kongo, Angolan, and Yoruba influence, as well as Mandinka.

20. Kuyk, 2003:2.

21. Ibid., (2003:xxii.

22. Curtin 1969:96-99.

23. Coolen 1991:1-18. A book devoted to west African Mande musical style is Charry 2000. I'm grateful to Michael Coolen for his helpful comments in reviewing this article, and to Eric Charry for introducing us.

24. Bühnen 1993:90. Table 3b, 69, and 86, 91, 100, and map 102. I can only speculate that the very low Fula percentage of .2% from 1548 to 1650 was because of their distance from the coast (away from Portuguese traders) and military strength that included significant mounted troops.

25. See Schaffer/Cooper 1987:101-04, Schaffer 2003:cover photo, 104-07.

26. Bühnen 1993:90.

27. Pollitzer 1999:129; see ibid., 124-29 on creolization and how an English-based Creole might have developed on the coast of West Africa and spread to coastal South Carolina and Georgia, developing variants from one plantation to the next.

28. Coolen 1991:1-18.

29. Jagfors 2003/04:26-33.

30. Ibid., 30.

31. Ibid., 30-31.

32. Painting shown in Pollitzer 1999:106.

33. Charry 2000:224 (for early twentieth-century Mande jembe players wearing turbans); 216 (for jembe and tama drums, the latter typically played with a single drumstick).

34. Coolen 1991. In Pakao, the word for slave was a very similar jong or jongo/jungo.

35. Pollitzer 1999:255n196. Pollitzer (1999:125) also sees jam, from the Wolof jaam for slave, jive from jev, to talk disparagingly; hip from hipi, to open one's eyes; and juke from Gullah joog for disorderly, ultimately from Bambara (Mande) dzugu, meaing wicked.

36. Ibid., 115; Turner 2002:194, for hudu, 202, for tabi.

37. Johnson 1986:149; first published in 1940.

38. See, for example, Wepman/Newman/Binderman 1976.

39. Charry 2000:54.

40. Edwards 1793: 2:56-57.

41. Ibid.

42. See for example, Austin 1984:265-307, "Bilali: African Patriarch in Georgia," mentioning Bilali's 13-page manuscript in Arabic, or Curtin 1967, photo section, for the letter written from a Virginia plantation by a Fulani named Job Ben Solomon, who was captured and sold into slavery by Mandinka along the Gambia River.

43. Reis 1973:93-97, 101 (for a photo of the grigri).

44. Austin 1984.

45. Local oral traditions widely point out how Pakao's Syllaba, during a several-year stay, forged a key alliance with the Almamy of Futa Jalon in Timbo to garner troops for his 1843 battle which destroyed the infidel king of Manduari. A proverb developed in Pakao, where tardy children were teased, "You take as long as Sylla stays in Futa."

46. Edwards 1983:65.

47. Turner 1973: words lists and 32-35.

48. Ibid., 292.

49. Ibid., 256-57.

50. Ibid., 32-35.

51. Ibid., 12.

52. Moreover, the word 'toting" is specifically used for carrying rice on the head, as done by the Mandinka. (see illustrations in Pennington 1913:34-35). Pennington also noted (ibid., 78-79) that the long-handled rice hoe is considered strictly a female tool, while the plow is a man's implement—as the Mandinka do today. The men's farming experience in Africa, and women's rice-farming in particular, might have been among reasons for the preference by Charleston purchasers for Mandinka or Mande slaves.

53. Turner 1973:19l, for bubu and bidi bidi; 194, for gumbo and gola/gula; 197, for kunu; 203, for tote; and 204, for yam.

54. Dalby 1970. Joseph Hill, an anthropology graduate student at Yale, told me an equally interesting, possible Wolof explanation for OK, and I wonder if they passed it on to the Mandinka, or vice versa. Two principal Wolof words for roughly "yes" are waaw and kay: waaw used at the beginning of a sentence and kay, for a bit more emphasis, at the end. Sometimes these two words are used together as waaw kay for "OK" or "all right then," to communicate overtones of respect and acceptance.

55. This Mandinka kuta is not to be confused with kuta as turtle, absorbed into the South as cooter. Mandinka is a tonal language; e.g,. jato can mean lion, oinion, or human body, depending on the tone.

56. Schaffer 2003:196-97. Patience Pennington (1914:447), a rice planter on the South Carolina coast, says that "unna" is a Gullah word for "you all," but "y'all" is close enough to English that she might not have considered it influenced by Gullah.

57. Schaffer 2003:202-03.

58. Dent 1902:5, of a modern typescript. She was also a distant cousin of Bilali's owner Thomas Spalding, and her own father, the physician James Troup Dent, traveled at least once from Broadfield on the mainland out to Sapelo Island to treat Spalding's family during the War of 1812. Ophelia's grandfather William Brailsford came down from Charleston with numerous slaves when he bought Broadfield on the Georgia coast in 1806, and Brailsford's father Samuel was one of Charleston's slave traders.

59. Schaffer 2003:116-17.

60. DuBois 1946, 1947:208. For matrilateral kinship in Pakao see Schaffer/Cooper 1987:87-90.

61. Johnson 1986:46, 149. Drumming also occurs for funerals, which Johnson, in the appendix, says Francis Moore reported in the 1730s, but today's Muslim Pakao Mandinka have stopped doing this. Ibid., 64, also relates a foot-wide drum covered with goatskin. Johnson himself compares Gullah drums to Mandinka drums described by Francis Moore (ibid., 215).

62. Ibid. 1986:67, 143; ibid., 181 describes a drum 18" wide and 15" deep, like a tabala. Ibid., 181, says "drums" from hog (pre-Islamic Mandinka) while "base drums" from cow, distinguishing the deeper sounding, summoning tabala from the tam-tam or jembe used for music and dancing.

63. Ibid., 118, 137.

64. Schaffer 2003:116-17.

65. Ibid., 56-59.

66. Johnson 1986:186-87, figures IIIa and IVd. For the sake of clarity, Gullah words are in quotes, to differentiate them from italicized Mandinka.

67. Ibid., 112-13, 143. For pictures of the white-robed Mandinka circumcision novices and riverwash, see Schaffer/Cooper 1987: xviii, xix, 97, 98.

68. Ibid., 161. Bilali's Muslim wife Phoebe, says "Ameen, Ameen." Katie Brown, Bilali's great-grand-daughter, also reports that her Ibo grandmother Hannah and Ibo uncle Calina were Muslim, probably converts through Bilali, and say "Ameela;" Hannah also says "Haka bara [Allah Akbar]" (ibid., 163-65).

69. Ibid., 80-82, 169, among several references to flying back to Africa.

70. Ibid., 110, 170.

71. Ibid., 34, 59, 79. 246; see Schaffer 2003:216-19, for Kadri Drame on "the fengkoto or hag."

72. Ibid., 101, 75, 99. Among the Gullah, "Duh owl is a true messenger of death." See Moore 1738, Appendix, reporting bua as owl or witch.

73. Ibid., 18, 65, 67, 102, 109-10, 124, for anti-conjuh charms, including hair, nails, and graveyard dirt, among numerous other references to conjuh and anti-conjuh.

74. Schaffer 2003:220-21, for Kadri Drame's use of boro as "poison" in "the Seer of Sunakarantaba."

75. Schaffer 2003:181-221.

76. Ibid., 2003:5, 188-89, 214-15.

77. Cassidy n.d., 5-81, cited in a new edition of Turner 2002:xxix.

78. Kuyk 2003:xxi.

79. Ibid. 2003:15, 42, 43, 57, 83, 88, 91, 115-18, 128, 141, 154.

80. Ibid. 2003:154-55.

81. Johnson 1986:70, 120, 145-46, 184.

82. Turner 1973:4.

83. Wylly 1915:39.

84. Parrish 1942:237.

85. Pollitzer 1999:51, 63.

86. Wylly 1915:39.

87. Ibid., 1915:40-41.

88. Ibid., 1915:42.

89. Bailey 2000:301-07.

90. Johnson 1986:162.

91. Personal communication, David Gamble, October, 1988.

92. Park 1971:360-63, first published in 1799.

93. Nishida 2003:83, quoting Campbell 1975:472.

94. Turner 1997:24, quoting Washington 1838:449-54.

95. Nishida 2003:96-98.

96. Schaffer 2003:34-35.

97. Brandon 1997:56-57; Montejo quoted in Rout 1976:32.

98. Brown 2003:49.

99. Brandon 1997:137, quoting Angarica 1955?

100. Arfanba Sagnan told how the Mandinka received snakebite medecine Katirao; see Schaffer 2003:166-169.

101. Writing at the end of the nineteenth century, England's Helen Bannerman incorporated Mumbo Jumbo terminology in her children's story Little Black Sambo, infamous for its racist stereotypes, further popularizing Mandinka terminology in the English-speaking world. Sambo's mother was called was called Black Mumbo, and his father was called Black Jumbo. These African Mandinka names and characters were grafted onto a largely Indian tale involving tigers. Sambo is a Jola name.

102. Cosentino 1995:430, for a description of the Haitian Gran Bwa. See the photographs of the leaf-covered Gambian kangkurao in Fletcher 1977:28-29, and compare this with a similar-looking, leaf-covered Haitian Gran Bwa in Cosentino 1995:179. In Pakao the kangkurao is entirely of bark.

103. Schaffer 2003:184-89, 202-03, including accounts both by Fodali Cisse and Kadri Drame. Francis Moore 1738:Appendix, word list first reported bwa, as mentioned, to mean witch and owl. In Moore (40, 116, 117, 133) the Mumbo Jumbo was a secret society of men centered on a masked figure, and the society spoke a secret language to maintain a certain mystery and power over women and girls, and uncircumcised boys.

104. Schaffer 2003:202-05.

105. Cosentino 1995:58.

106. Ibid., 32-33, in "Imagine Heaven" by Cosentino.

107. Geggus 2002:254n66). He quotes Diouf 1998:152-53, 229. Geggus 2003:229nn., considers the Mandingoes of northern Haiti to be "the most striking Muslim cultural survival in the Americas," and mentions they are described in Alexis 1970:173-85, and Najman 1995:158-60.

108. Wafer 1991:126, 198, 201-02, about Omalu and Brazilian Candomble.

109. Araújo and Lanzelloti n.d.:132, 134, 146-47, 149. Meirles 1983/2003:74, who observed and painted Candomble in Bahia from 1926 to 1934, wrote that Oxalá was "probably a distortion of Alah," the word of course used by Muslim West Africans, including the Mandinka, for Allah. Oxalá seems to be the origin of Oxalufam and Oxalufa, raising the possibility that the white-robed deity Oxumare comes from Omar, the second Caliph. Interestingly, Meirles also found that Candomble was a form of Macumba or magic practice devoted to achieving "good," while canjerê was devoted to achieving "evil;" the latter, ibid., 68, evokes "conjuh," the negative or positive magic of Gullah root doctors of coastal Georgia and South Carolina. Indeed, the lilting drawl from the Bahia state in Brazil, where slave descendants form a large portion of the population, evokes the Southern accent in the United States.

110. Wafer 1991:78-80.

111. Pollitzer 1999:115.

112. Pennington 1913:11-12. See the modern photo of a Mandinka woman dancing in similar fashion, pounding her left foot, arms extended out, in Charry 2000:202.

113. Benedetti 1997:8.

114. Lovejoy 2003:14.

115. Cosentino 1995:123.

116. Ibid., 78, 87.

117. Among several sources dealing with Proto-World or Mother-Tongue are Shevoroshkin 1990.