Legon Journal of the Humanitie_2.cdr On the origins of locative for in West African Pidgin English: A componential approach 1 Micah Corum Abstract This paper discusses possible origins of locative for in West African Pidgin Englishes. The development of for is framed componentially, that is, in terms of deriving its constructional meaning from different components that sustained linguistic and cultural contact along the Upper and Lower Guinea Coasts during the Early Modern Period (1500-1800) among West African, Portuguese, Afro-Portuguese, British and Afro-British populations and the sailing populations of both official naval vessels of the British and Portuguese Empires and privateering ships. 1. Introduction This paper investigates traces of contact in the emergence of locative for in West African Pidgin English. We take a “componential approach” to the formation of the construction in various West African Pidgin Englishes (Hancock 1986, 1993, 1994) and discuss how the construction would have developed in the Early Modern Period from prolonged linguistic contact with Upper and Lower Guinea Coast languages, the Portuguese Creoles and the regional varieties of English that were spoken on board both the official naval vessels of the British Empire and privateering ships. Thus, following Hancock, a division of the 1 Micah Corum is a research fellow in Linguistic Diversity Management in Urban Areas, Universität Hamburg. linguistic components is made as follows: the creole component, the Guinea Coast component, and the English component. Convergence among the three components led to the formation of for as the prototypical locative construction in West African Pidgin Englishes. 1.1 The general locative construction In all West African pidgin and Creole languages, there exists a polysemous locative construction that conveys a general concept of space. Today the construction is realized as na in Portuguese-lexifier Creole languages and certain English-lexifier creole languages spoken from Guinea Bissau to Equatorial Guinea, and as for in West African English-lexifier pidgincreoles in Ghana, Nigeria and Cameroon (Bakker, 2008). Apart from pidgins and creoles, the general locative construction appears in Niger-Congo languages. Table 1 lists locative constructions in some Lower Guinea coastal languages and West African Pidgins and Creoles. Corum | 44Special Edition , 2012 | 43Legon Journal of the HUMANITIES Table 1. West African Coastal Languages and Their Locative Constructions Language General Locative Locative Copula Construction Construction English-lexifier Pidgin/Creole Ghanaian Pidgin English for dé Nigerian Pidgin for dé Kamtok for dé Pichi na dé Kwa Akan w? w? Ewe le le Fon dò dò West Benue-Congo Yoruba ní ~ n’, l’ wà Igbo na ibì, ?n? East Benue-Congo Efik ké ~ k’ du Ijoid Kolokuma ? tími, emí Bantu Duala o bé Mpongwe go ni~na~ne ne, (a)re As Table 1 shows, the general locative construction derives its form from an identical locative copula in Kwa languages. In other languages, including the pidgin and Creole languages, the origin of the general locative construction is not as transparent. It can be stated with confidence, though, that the general locative construction in West African Pidgin and Creole languages is not derived from the locative copula construction. In addition to the obvious differences in form between for and the locative copula de, the historical uses of the two constructions are incompatible. The general locative construction appeared in Pidgin English more than two centuries before the first attestation of the locative copula de in 1897 (Huber 2011: slide 16). The emergence of for should not be argued for in terms of internal linguistic motivation then. The general locative construction is assumed in this paper to have been modelled on uses of similar constructions in languages that contributed to the development of West African Pidgin Englishes. Both Portuguese and Guinea Coast languages contributed to the formation of na in Portuguese- lexifier Creoles. This paper implies that na in the Portuguese Creoles would have served as a springboard for a general locative construction to emerge in the English-based pidgins of the Lower Guinea Coast. The orthography and abbreviations used in the linguistic examples below appear as they do in the works from which they were taken. The authors’ glosses are retained so that the meanings that had been assigned to a morpheme, lexical item or other construction would not be lost. Portuguese and Guinea Coast languages contributed to the formation of na in Portuguese- lexifier Creoles. This paper implies that na in the Portuguese Creoles would have served as a springboard for a general locative Corum | 46Special Edition , 2012 | 45Legon Journal of the HUMANITIES construction to emerge in the English-based pidgins of the Lower Guinea Coast. The orthography and abbreviations used in the linguistic examples below appear as they do in the works from which they were taken. The authors’ glosses are retained so that the meanings that had been assigned to a morpheme, lexical item or other construction would not be lost. 2.1 Portuguese-lexifier Creole component Portuguese-lexifier Creoles could be found along the Lower Guinea Coast throughout the Early Modern Period (Barbot 1732; Atkins 1735; Protten 1764). In reference to the linguistic situation on the Gold Coast in the mid 18th century, Christian Protten remarked on: Fante-eller Aming-Sprog, som strax sagt er, fra Axim lige indtil Riovolta, Crepe eller Popo, naest det Neger- Portugisisk, forstaaet, talt og af alle Sorte som et General- Sprog. Fante or Amina language which, quickly said, is understood and spoken from Axim even to River Volta, Crepe or Popo, next to the Negro- Portuguese [pidgin, i.e.] by all sorts [of people] as a general language (Trutenau, 1971: 5-6, original brackets). Merchants on the Lower Guinea Coast would have carried over remnants of the Portuguese varieties when they interacted on a more regular basis with English and Dutch traders at the beginning of the 18th century. Even the Asante traders from inland would have used elements of the Portuguese variety when dealing in English, as Huber surmises: After about 1700, the hinterland Akanists must also have started to use this English-lexicon variety when coming to the coast, perhaps by relexification of their traditional Portuguese jargon. (1999: 45) It cannot be said with certainty that na was the general locative construction par excellence in 17th century Portuguese Creole–there is no documentation of the Creole during that time; th the first documents of Kriyol, for example, are from the 19 century (Bocandé 1849; Schuchardt 1888). However, the construction occurs in many of the Atlantic creoles that are believed to have an Upper Guinea origin (Hancock 1986; Martinus 1997; Jacobs 2009). If features of the core grammar of Papiamentu can be linked to language (Jacobs 2009: 27-31), then one can assume those grammatical features were in place much earlier in the Guinea Bissau and Cape Verde varieties. Taking this as a point of departure, we infer that persons on the Gold Coast who were familiar with the “Negro- Portuguese” language that Protten refers to, would have used na in locative constructions when referring to space in a general way. In fact, by Protten’s time at least, it had already emerged as the prototypical marker for locative relations in the Surinamese Creole varieties. The construction was recorded in a court deposition in 1745 and glossed as a preposition meaning “in, up, at,” in early Sranan (Van den Berg 2000: 92). No Creolist would suggest that the Portuguese-lexifier Creoles inherited na from the importation of creole languages from the Americas to West Africa, an issue that seems to be certain for Krio (Huber 1999a). On the contrary, the locative construction na was carried across the Atlantic to the Caribbean and across the Indian Ocean to Asia, where it survives in various forms in Afro-Caribbean creole languages, Philippine Creole Spanish, and Indo Portuguese creoles (Forman 1988). The issue at hand, then, is to identify possible local sources for the general locative construction na in the Portuguese-lexifier Corum | 48Special Edition , 2012 | 47Legon Journal of the HUMANITIES Creoles. These same sources could have potentially influenced the formation of locative for in Pidgin English. Kihm believes that na in the Portuguese-based Creoles is derived from an item in the lexifier language (1994: 69): Na, from Portuguese na 'in the (feminine)', appears as the locative preposition par excellence, encoding as it does the spatial relations BE-INSIDE (i sta na kwartu 'S/he's in the room'), BE-ON-THE-SURFACE (i sta na mesa 'It's on the table'), BE-AT-A-LOCATION (i sta na fera 'She's at the market'. Concerning the first of these relations, na has a broader meaning than dentru di as it is not limited to clear-cut containers or locales. For instance, one may say dentru di kwartu or dentru di kasoti 'inside the box' as well as na kwartu orna kasoti. But ?i na dita dentru di kama 'She's lying inside the bed' is not a proper alternative to i na dita na kama 'S/he's lying in bed.' Many Atlantic and Mande languages, however, have lexical items that are similar in meaning and sometimes identical in form (e.g., Sérèr, Bambara, and Susu) to the locative construction na in the Portuguese-lexifier creoles. Even more, there are polysemous items in languages of the Lower Guinea Coast and one of the functions of those items is to convey a general notion of spatial relations between entities. Igbo, for example, uses ná in this general locative construction, which is “one of the few words in Igbo that can be called prepositions…translatable as ‘at, on, in, to’” (Lord 1993: 14). The example in (1) shows the particle ná as a preposition in Igbo (ibid), which is similar to the locative item in example (2) taken from Kriyol (Kihm, 1994: 69): 1) ó bì ná oká he live at Awka ‘He lives at Awka.’ 2) Omi sinti kusa ngata i na garganti man feel thing attach him at neck ‘The man felt that something was attached at his neck.’ Taylor (1971) has argued that na in Creole varieties holds a broader semantic range than the preposition in Portuguese. In addition to Igbo, though, there are numerous locative structures from Guinea Coast languages that have functions resembling the locative construction in the Portuguese-lexifier creoles. We begin with those that are identical in form and then move on to items that are different in form, but mirror the construction in the creoles in terms of function. 2.1.1 Guinea Coast contributions to locative na in West African creole languages Lord suggests that a Proto-Niger-Congo verb na meaning be at is the source for many of the conjunctions, comitative prepositions, locative verbs/prepositions, incompletive aspect particles, and structures used in possession constructions that are found in Benue- Kwa languages today (1993: 30, 50; see also Welmers 1973: 312). Traces of the Proto-Niger-Congo verb can be found in locative verbs/prepositions, possessive verbs, and aspect particles in West African languages spoken along the Guinea Coasts as far as Gabon. Mpongwe, a coastal language of Gabon, uses both na and go as two polysemous locative particles (Wilson 1847). In Douala, a coastal Bantu language of Cameroon, there is a lexical Corum | 50Special Edition , 2012 | 49Legon Journal of the HUMANITIES construction similar in form and meaning to the general locativeconstruction na in the Portuguese creoles. Saker calls na in Douala a dative case marker (1855: 37). He glosses the dative case marker na as according to, concerning, with, from, for, and towards (ibid). Saker also notes that adverbs in Douala can be constructed by prefixing na to an abstract noun. Saker provides the following example, “Na janwa, ‘Wisely,’ lit. with wisdom. He speaks truly, A makwala na mbadi” (1855: 35). Yoruba has a similar construction with ni and West African Pidgin Englishes use for in this way. Moving west along the coast, the same construction can be found in ??an?, an Ijoid language of Nigeria. Blench lists only two items in the dictionary of ??an? that he compiled: ??, which means on and in, and na, a conjunction meaning and and a preposition meaning with (2005: 70). In Ga, na functions as an adposition. Zimmerman (1858: 70) described it in one entry as mouth, which serves as “a postposition and adverb expressing the relation of place and by tropic use also that of time and manner (§ 29) as: at, to, near, according, according to; at the point of; along etc.” In Susu, an Atlantic language spoken on the coast of Sierra Leone, na is the copular verb to be (Houis 1963: 114). Duport recorded instances of the language almost a century earlier. He glossed a postposition ra meaning “with, or for,” which becomes na in certain linguistic environments (1865: 15). As an adverb, the na construction means there in Susu (Duport 1865: 24). Similarly, certain Kru languages employ na as an affix that roughly means “place” (Koelle 1854). The fact that there are remnants of various forms of the Proto- Niger-Congo verb na in Guinea Coast languages and those structures resemble the functions of na in the Portuguese Creoles does not prove substrate influence on the emergence of the general locative construction in Portuguese creoles, though it does not hurt the argument either, which had been made early on by Taylor (1960: 157). This paper does not argue that those identical structures were the sources from which the Portuguese developed t h e i r g e n e r a l l o c a t i v e c o n s t r u c t i o n . T h e f a c t t h a t thoseconstructions exist in the languages that came into contact with the “Negro Portuguese” that Protten refers to provided the motivation for the use of a multiple meaning-function construction in the Portuguese varieties that were used by speakers of Douala, Igbo, Ga, Kru languages, Susu, Bambara, and Sérèr. Aside from identical forms of the locative construction in Portuguese creoles and Guinea Coast languages, there are also numerous semantic similarities in the use of locative structures in the Upper and Lower Guinea Coast languages. Those locative structures are used as general locative constructions and can be considered an areal feature of the languages of the Guinea Coast. 3. The general locative construction as an areal feature of Guinea Coast component The languages of the Guinea Coast contain items that can be considered general locative constructions. The following sections explore how the constructions are similar to locative na and for in West African Pidgin and Creole languages. Corum | 52Special Edition , 2012 | 51Legon Journal of the HUMANITIES 3.1 The Upper Guinea Languages Hancock (1986) provides a map that illustrates the location of early Lançado settlements along the Upper Guinea Coast in the 17th and 18th centuries where an early creole English emerged. The language called Guinea Coast Creole English was used alongside Atlantic and Mande coastal languages, as seen in Figure 1 below.Table 2 provides the general locative constructions that appear in the languages that Hancock lists as pertinent to the areas where the early Lançado settlements were located. Interestingly, a polysemous locative item exists in the majority of the languages that Hancock includes in his map. Table 2. General locative constructions in Upper and Lower Guinea Coast languages. Figure 1. Distribution of languages along the Upper and Lower th Guinea Coasts. 17th & 18 Century settlements Indicated with ? (Taken from Hancock 1986: 89) Corum | 54Special Edition , 2012 | 53Legon Journal of the HUMANITIES Two basic assumptions are made concerning the linguistic situation in and around the Lançado communities. The first one is confirmed by sociohistorical research, but is common knowledge about West African communities in general: the Lançado communities that Hancock identifies would have had multilingual speakers of Portuguese-lexifier Creoles who would have spoken one or more of the languages that surrounded their communities (Kihm 1994: 4-5). The second assumption relies on outcomes of language mixing in multilingual settings: the Atlantic and Mande coastal languages spoken around Lançado communities would have influenced aspects of the emerging grammar in Portuguese-lexifier Creole languages and Guinea Coast Creole English. Since Mande, Atlantic, and Kru languages contained items that conveyed a general notion of location, it is not surprising that the Portuguese varieties and Guinea Coast Creole English inherited similar constructions as well. The Upper Guinea coastal languages were not the only linguistic source for the Portuguese varieties. The contribution of the Lower Guinea Coast languages on the Upper Guinea Creoles must not be underestimated. General spatial grams in Lower Guinea Coast languages could have influenced locative na in Portuguese creole varieties and Guinea Coast Creole English too, as briefly mentioned in section 2.1 above. Portuguese Creole varieties were in use along the Upper and Lower Guinea Coasts by the end of the 17th century. Barbot, a 1685 French Huguenot refugee, remarks on language use on the Gold Coast: The Fetu [Afutu] language being most generally understood at the Gold Coast amongst the Blacks, as I have said before; I have made a collection of some familiar words and phrases, which shall be found in the supplement.... besides the other languages, in which we can talk to them: for many of the coast Blacks speak a little English, or Dutch; and for the most part speak to us in a sort of Lingua Franca, or broken Portuguese and French. (1732: 249) The Lower Guinea coastal languages indigenous to the areas where Europeans had recorded a Portuguese Creole language should be included as an ingredient of the Guinea Coast component as well. In addition to their influence on locative na in the early Portuguese varieties, those languages would have motivated locative for in Pidgin Englishes after Portuguese Creole had lost linguistic strength in those areas at some point in the 18th century. 3.2 The Lower Guinea languages The attention is now turned away from the semantic and formal similarities in the use of locative particles in Atlantic and Mande languages and the locative construction na in creole varieties of the Upper Guinea Coast, to focus on locative structures in the Lower Guinea coastal languages that could have influenced the use of locative for in West African Pidgin English. 3.2.1 Certain Bantu and East Benue-Congo languages In addition to na, which provides a wide range of meanings, Douala has a spatial gram o. Kolokuma, an Ijoid language of Nigeria, contains the same grammatical morpheme (Williamson 1965: 38). The semantics of o in Douala and Kolokuma matches the general meanings that for expresses in West African Pidgin English. Saker translates it as about, around, in, into, to, and unto (1855: 37). The locative compound is also used in more complex constructions in Douala. For example, the deictic adverb there is a compound word une or one, which means at yonder place (Saker 1855: 35). The o + ne construction in Saker’s grammar looks similar to the for + deictic de constructions Corum | 56Special Edition , 2012 | 55Legon Journal of the HUMANITIES that are often used in Nigerian Pidgin and Ghanaian Pidgin English. Ibibio and Efik have a lexical item ke that serves a multipurpose locative function (Anyanwu 2010; Una 1900), though Efik also has a construction ye that has a general locative meaning as well (Una 1900: 29). With regard to ke, Una states, “‘ke’ seems to be infinite in its capacity for producing prepositional meanings. It is sometimes combined with nouns or verbs to form prepositions” (1900: 29). The same combination occurs with na in the Portuguese-lexifier Creole languages and in Saramaccan and Sranan. Similarly, for combines with nouns to form “complex prepositions” in the pidgincreoles, for example for on top, for inside, and for im side (Faraclas 1996; Mann 1993). The constructions discussed in this section support a case for the domestic origin of locative for that were used in the Pidgin English in the Calabar area or in Cameroon by the end of the Early Modern Period (Fayer, 1982). 3.2.2 Some West Benue-Congo languages Yoruba contains a locative-relational marker that is similar to the associative function of for in West African Pidgin Englishes when the construction co-occurs with the locative copula de. In Yoruba, as in many Benue-Kwa languages, possessive constructions share semantic and syntactic characteristics with locative structures. Lord provides the following example (1993: 13): 3) ó ní owó he have money. ‘He has money.’ Historically, this could be derived from a construction that meant Money is at him. That would be compatible with Lord’s argument concerning the development of the Yoruba locative structure ní, which she believes is “historically derived from a former locative verb ní, related to the homophonous verb of possession” (1993: l13). Crowther also provides an example of possession in his grammar of Yoruba: “‘to have money’, lówo; ‘to have or possess heaven’, l?run’” (1852: 8). In Yoruba, /l/ and /n/ are variants of the same locative construction. Crowther and Lord’s examples demonstrate the close relationship that exists for constructions of possession and location in Yoruba. Similar to the Douala example above, the spatial gram ní in Yoruba can also be used before adverbs, as shown in the following example (Lord 1993: 20, originally in Awobuluyi 1978: 77): 4) Ayé ñ l? ní m.èl?m.èl? Life PROG go at smoothness ‘Life went on smoothly.” A similar construction is found in Nigerian Pidgin (Faraclas 1996: 145): 5) A byud haws fòr smol-smol Isp buildF house p be small/nR ‘I built the house slowly.’ Temne can form the same construction using locative ka, as seen in the more than 100 examples that Schlenker lists in his grammar (1864: 48-60). 3.2.3 Gbe influence In Ewe, pé serves as a relational noun for place or area. It has grammaticalized into a possessive marker, according to Heine (1997: 93): 6) Kofí pé x? Kofi of house ‘Kofi’s house’ (Historically: ‘The house at Kofi[’s place]’) Corum | 58Special Edition , 2012 | 57Legon Journal of the HUMANITIES Welmers (1973: 308) found that in Ewe the expression for possession was realized as a construction in which the possessed was located at the possessor’s hand. Heine claims that these types of possessive constructions are based on a Location Schema, which is composed of two sub-schemas (1997: 92): Y is at X’s home, and Y is at X’s body-part. According to Heine, the Location Schema “forms one of the most frequently employed templates for expressing attributive possession: The possessor is conceptualized as the place where the possessee is located” (1997: 93). We combine the two schemas in the chart below and list the glosses in English from coastal Guinea languages that could have influenced locative for in West African Pidgin Englishes. The schema is represented as a general conceptualization of LOCATION for POSSESSION. Table 3: Locative Structures Used in Possessive Functions This kind of possessive construction occurs in Nigerian Pidgin as well. It fits Heine’s ‘Y is located at X’s body-part’ Location Schema (Faraclas 1990: 108): 7) Mòni dé yù fòr hánd Money copula you preposition hand ‘You have money.’ And again in Rotini and Faraclas’ grammar of Nigerian Pidgin (n.d.: 42): 8) Dì búk dé mì f?r hánd DEF book COP PRON1ps LOC hand. ‘The book is in my hand.’ The Nigerian Pidgin example can be interpreted as, I have the book, which looks strikingly similar to the Mande example below, except that Nigerian Pidgin for is prenominal and the construction in Kpelle is postnominal (Heine 1997: 93): 9) S?ŋkau a n yee- I money PL be my hand- LOC ‘I have money.’ (Lit.: ‘Money is in my hand.’) Languages of the Akan dialect cluster also have a locative verb that functions as an existential or possessive verb. Riis gives an example of 19th century Twi (1854: 113): 10) mogya vo yeñ ãnom blood be_located 3PP-POSS mouth_inside ‘There is blood in our mouth.’ The sections above showed that the general locative construction exists as an areal feature of Guinea Coast languages. The locative structures of those constructions are used for possessive functions in many cases. The origin of na in Upper Guinea Portuguese lexifier Creoles should not be traced solely back to a preposition in Portuguese. Similarly, for in West African English-lexifier pidgincreoles should not simply be considered a borrowing from Metropolitan English. Corum | 60Special Edition , 2012 | 59Legon Journal of the HUMANITIES Language Possessed Y Possessor XLocated at/ location of English Gloss Mande Ewe Yoruba money house money located at location of located at my hand Kofi him I have money Kofi’s house He has money 4. Where did for come from? Na did not survive in the general locative construction in the Lower Guinea English varieties that Europeans recorded in their travel accounts in the 18th century. Instead, for took its place. Since for subsumed the functions that na held in the early Englishes of the Lower Guinea Coast, it is less useful to ask, “Where did na go?”, and more pertinent to ask, “Where did for originate from?” The following sections will attempt to answer this question. 4.1 Early uses of for The earliest uses of for are found in the second half of the 17th century. The following quote was recorded by Ralph Hassell in 1686 and, according to Huber, it is the earliest attestation of restructured English from the Gold Coast that has been discovered so far: butt hee no savee for wt hee noo love mee (1999: 42). In the next century, John Conny, the caboceer at Axim, was heard saying, “by G--- me King here, not only for my water, but the Trouble has been given me in collecting it” (Atkins 1735: 75; also in Huber, 1999b: 43). For in that instance would mean in, on, at, around, etc., which makes it the earliest attestation of for as a locative structure. This is much earlier than Huber’s date for the earliest attestation of the “Locative/Temporal Preposition” for in West African Pidgin English, which he identifies as 1773 in Nigerian Pidgin (1999b: 82). Sometime in the 1790s, Adams reported hearing four uses for by an Anomabo man: You be old man for dis country….you hab black man net too. Agar man, name Quacoo, lose net; some man teef him; you hab for ship, Brinny omo….Suppose he be teef for Aberrikirry (England), he no be teef for dis water…(Adams 1823: 22-3, taken from Huber 1999b: 43) The polysemy of for in the previous quote is similar to the uses of for that Antera Duke uses in his diary, which dates from 1786- 1789 (Forde 1956). The use of for as a locative structure emerged on the Lower Guinea Coast at the beginning of the 18th century. It took over the role of na, which was already an entrenched feature of Portuguese-lexifier and Portuguese-derived creole varieties. One factor that contributed to locative for–the factor that most scholars implicitly accept as the source, in fact–was the use of the preposition for in Early Modern English. The construction had a much wider range of use than it does today in British or American varieties of English, although not as wide a semantic range as the Guinea Coast languages discussed above. 4.2 The English component Some of the uses of for in Early Modern English look similar to the kinds of for instances that one finds in West African Pidgin Englishes. The sailors and merchants that visited the Gold Coast during that era employed for in ways that sound anachronistic to speakers of regional Englishes today. Literature of the Early Modern Period can shed light on acceptable uses of for in the 17th and 18th centuries. The following examples are taken from the first 10 chapters of the Gospel of John in the King James Bible, which was standardized in 1760 by Francis Sawyer Parris. Each excerpt appears as it does in the King James Version along with the translation from the English Standard Version from 2001: 11) The same came for a witness (John 1:7) He came as a witness 12) grace for grace (John 1:16) gracupon grace Corum | 62Special Edition , 2012 | 61Legon Journal of the HUMANITIES 13) Jesus would go forth into Galilee (John 1: 43) Jesus decided to go [ø] to Galilee 14) Samaritans of that city believed on him for the saying of the woman (John 4: 39) Samaritans believed in him because of the woman’s testimony 15) The thief cometh not, but for to steal, and to kill, and to destroy(John 10:10) The thief comes only [ø] to steal and kill and destroy The five examples above show extended uses of for as a preposition, a conjunction, and an infinitive marker. There are hundreds of examples throughout the New and Old Testaments of these uses of for. Popular 17th-century literature depicts similar polysemous uses of for. George Chapman’s Monsieur’s D'Olive contains an instance in which for functions as a lative preposition: “I leave your Highness to deal with Atropos…I am for France” (1606: iii.1.75-76). In Shakespeare’s All’s Well That Ends Well, a derived construction of the spatial gram functions as a preposition of place, which today would be said using in front of or before: “Fore whose throne ‘tis needful…to kneel” (1623: iv.4.3-4). For had a wider range of use in 17th and 18th-century English than it does today. If the construction had extended grammatical functions in literary language, one can imagine the range of use it retained in dialects, jargons and secret languages. Sailors, for example, were known to speak a particular vernacular of ship English (Rediker 1987: 162). Their uses of for either equaled the polysemous functions of for that can be found in literature of the Early Modern Period, or possibly even exceeded it. James Parris, an Englishman who visited the factory Sekondi on the Gold Coast, wrote in 1683, “Addoomes came on board with two small canoes for to have us come on shoare againe with our flgg…” (Law 1997: 2). James’ English is typical of the linguistic variety of the times in which for preceded infinitive constructions. African persons who dealt with sailors, traders and merchants speaking Early Modern English would have heard for used frequently and in extended ways. In addition to the British, the Dutch were also an important historical and linguistic component in Afro-European contact on the Gold Coast. The Dutch took the São Jorge castle at Elmina from the Portuguese in 1637. The Elminans who were Portuguese interpreters remained in the community and were involved in dealings with Africans and Europeans well after the period when the Portuguese had lost political ground on the Gold Coast. Huber says of the linguistic situation: “In the Dutch possessions previously held by the Portuguese these interpreters spoke (jargonized) Portuguese, which, towards the end of the 18th century, was possibly relexified and later supplanted by (jargonized) Dutch” (1999a: 87). It is likely that na was used by those interpreters who spoke “jargonized Portuguese.” Spatial grams in Dutch, possibly voor, could have supplanted na in the general locative construction. Although this is conjecture, it is the best hypothesis since voor is the closest lexical item to for in West African English-lexifier pidgincreoles and it conveys other meanings than just the benefactive, for example, before and because. However, the Dutch componentshould only be taken so far: “There is no clear evidence of a Dutch-derived domestic creole having developed on the coast, although a Dutch pidgin may have been used at some time in Calabar (Smith, Robertson, & Williamson 1986); it is probably for this reason that the Dutch employed Creole Portuguese as a medium of communication” (Hancock 1986: 88). This section will not focus on linguistic Corum | 64Special Edition , 2012 | 63Legon Journal of the HUMANITIES aspects of voor in Middle Dutch or the dialects that emerged from it. Future research should investigate aspects of this issue by looking into polysemous uses of high-frequency spatial expressions in the varieties of Dutch that were spoken in areas where sailors lived who were recruited by the West India Company in the 17th and 18th centuries. The majority of those young men were from lower class backgrounds and were not only of Dutch origins, as Feinberg explains (1989:86): Many of the Company's recruits, especially the military, came from the orphanages and workhouses of such cities as Rotterdam. Zielverkoopers (literally,"soul sellers"), labor recruiters and seedy characters marginally cared for poor and homeless men until a ship sought a crew or a company hired new men for its forts and factories overseas. Boxer [1977, p. 51] suggests that many of the men who served with the overseas companies were the ‘dregs of the Dutch nation’.... though the West India Company was of Dutch origin, the Company's employees included not only Netherlanders, but also men from other parts of the continent, including what are today known as Belgium, France, and Germany. Thus, the designation Europeans is more accurate than Dutchmen, especially for those serving the Company as soldiers. The European component does not provide a complete account for the appearance of locative for in West African Pidgin Englishes. Aspects of the local languages where Pidgin English varieties were spoken would have contributed to the formation of an all-purpose construction that expressed spatial relations in a general way. We turn to the Guinea Coast component again to address this side of the issue. 5. The Gold Coast component in the emergence of for Both Barbot and Protten stated that “Negro Portuguese” and varieties of English were spoken on the Gold Coast in the 18th century. The speakers of those languages resided on the coast where they could trade with Europeans. Inland persons also travelled to the coast to trade with Europeans. After they traded, they often needed help returning inland with their goods. Barbot says of hiring slaves to travel inland: Those inland Blacks, who come down without slaves of their own to carry back the commodities purchased, hire either free-men or slaves, who commonly live under the forts, at such rates as they can agree upon, according to the distance of the places the goods are to be carried to. (Barbot 1732: 261) Free men and slaves who needed to communicate with the inland traders would have spoken one or more of the coastal languages, most likely a dialect of the Akan cluster or Ga. However, if they lived around the European forts, they would have also been familiar with some form of the Portuguese-lexifier Creole or an emerging English-lexifier Pidgin or Creole that Europeans kept referring to in their early accounts of the Guinea Coasts. This means they would have used a locative structure in one of the Kwa languages that they spoke, or, quite possibly, they used the existing na structure from Portuguese creoles. Since we have already explored na above, we turn now to possible influence from Akan languages on the formation of for. 5.1 Locative w? in the Akan dialect cluster Speakers of the dialect cluster referred to today as Akan can be found throughout central and southern Ghana. By the end of the Early Modern Period, Fante was the language that had the most widespread use in areas where slave forts and castles were Corum | 66Special Edition , 2012 | 65Legon Journal of the HUMANITIES built along the Gold Coast. Fante is one variety of the Akan dialect cluster. Another important variety is Twi, which was spoken inland. The Akuapem dialect of Twi was the first to be described in detail (Riis 1853). Akuapem Twi is spoken north of Accra. The main language spoken in Accra, and hence around Fort Christiansborg, is Ga. Grammars of Ga, Fante and Twi date from the mid 18th and 19th centuries. The present study uses those grammars and dictionaries in order to get an idea about the structure and lexicon of the indigenous languages that were being used alongside restructured European varieties on the Gold Coast at that time. There is a locative/possessive verb w? in all dialects of Akan. The verb takes the form ni in the negative, which is close to the Proto-Niger Congo verb na that Lord hypothesizes influenced locatives and possessive verbs in Benue-Kwa languages (1993: 13). Wo can also serve as a general spatial gram in certain linguistic environments. Christaller provides the following description for w? in the section on prepositions and postpositions in his 1875 grammar of Tshi: Rest in a place: (in answer to the question where?) is indicated by wo (section 102,3). This verb introduces a place where the subject or object is, or an action goes on; the place itself is indicated by nouns of place, parts of which, when referring to certain objects, are also called postpositions (section 118-127). Together with such postpositions, wo may stand for the prepositions in, on, by, with, upon, over, above, under, below, amongst, between, before, behind, about, near. (74) Riis (1854), another Basel missionary who lived in Ghana in the mid-19th century, produced an early description of the Akuapem dialect. The description of the locative/possessive verb w? is similar to what Christaller included in his dictionary (Christaller admits that he drew on Riis’ work when compiling his own grammar and dictionary 20 years later (1881: VI)). However, the entry that Riis uses for the form of w? in his description of Akuapem is peculiar. He transcribed w? as vo. His section on the pronunciation of consonants in Akuapem might provide an idea why he chose to use /v/ instead of /w/ (1854: 6, original italics): The pronunciation of v and w it is difficult to describe [sic]. The former is not so very different from the English v, though not quite the same. W has nothing of the bread sound of the English w. It is the softest and least articulated consonantal sound, so that it in some words easily escapes a European ear altogether. In the German version of the same grammar and dictionary, Riis described the /v/ sound (1853: 5): “In ‘v’ herrscht das labiale Element vor, und es steht daher dem deutschen w oder dem französischen ‘v’ näher. In V the labial is dominant, and so it is closer to the German W or the French V” (translation credited to L. Pietsch, personal communication, May 23, 2012). Riis lists two different lexical items, vo and wo in his index (1854: 244-45), so we can assume that [w] was qualitatively different from [v]. When Riis refers to w?, he uses vo, which would have been pronounced va. If the locative/possessive verb was va and that item appeared as a general locative construction in particular linguistic environments, then that sound could have been interpreted as [f] in some cases by persons not familiar with the Akan dialects, for example, West India Company slaves who came to Elmina and also Akan speaking boatmen who travelled to the Slave Coast and the Niger Delta region (Feinberg, 1989: 81). This would have rendered a locative particle [va ~ f?]. Almost a century earlier in 1764, Protten describes Ga and Fante, and uses /v/ to represent sounds that start with /w/ today (Trutenau 1971). Corum | 68Special Edition , 2012 | 67Legon Journal of the HUMANITIES He includes w? in his index of Ga, but provides readers with two options for /w/ pronunciation: either omitted altogether, or pronounced as [v]. In Wilhelm Müller’s 1667 list of Efutu vocabulary, words that begin with /w/ that are followed by a back vowel in modern Fante are either omitted or changed into a cluster of consonants. There is one entry in Müller’s work that matches the same form used in the general locative construction in Akan languages: O fa, which means mother’s brother and is listed in 19th- and 20th-century Akan grammars as w?`fa (Jones 1983: 286). Zimmermann remarks that in Ga, /w/ is “somewhat different from the English and from the German /w/; though nearer to the latter, it is an /u/ slightly modified by the lips (1858: 5). According to 18th- and 19th-century accounts of Kwa languages on the Gold Coast, the pronunciation of word initial [w] followed by a back vowel was different from what pertains in varieties today. In sum, the Gold Coast languages in the 18th century could have contained a lexical item vo that was frequently used as a pronoun, a locative verb, or a general locative construction. Riis provides a number of examples in his grammar in which the construction co- occurs with the locative verb it is derived from (1854: 244): 16) o-vo mosia vo ne kotoku-m 3psSUB-wo pebbles LOC 3psPOSS bag-inside ‘He has pebbles in his bag.’ 17) mivo abusuafo vo ha ni Akam 1ps-wo relatives LOC here and Akam ‘I have relatives here and in Akim.’ 18) mivo nnamfo bebri vo kra yim 1ps-wo friends many LOC place this ‘I have many friends in this town.’ Riis describes vo as locative marker as well (1854:244): It is used to connect an object of locality with the predicate if the object is merely qualifying or accessary (not integrant), and the action expressed by the predicate is conceived as a motion or rest in a place (not as motion to or from a place). Examples of vo as a general locative construction include (Riis 1854: 244): 19) mi-hû no vo Mamfe 1psSUB-see 3psOBJ LOC Mamfe ‘I saw him at Mamfe.’ 20) o-digwa vo b?bi 3psSUB-trades LOC somewhere ‘He is trading somewhere.’ Riis lists a number of examples in which the spatial gram occurs with the verb fi, ‘to come forth’, as in “Sarre fi vo asâse so (more frequently fifi), grass grows on the ground” (Riis 1854: 182). This could be rendered today in Ghanaian Pidgin English or Nigerian PidgiN as roughly grɑs de gro f? groŋ (Bankole, personal communication, November 4, 2010). The w? construction as a general or as sole preposition as some linguists have called it (Osam 1997), behaves syntactically similar to for in the West African Pidgin Englishes. The following Twi example from Riis (1854: 183) resembles the Nigerian Pidgin example from Faraclas (1996: 64): Corum | 70Special Edition , 2012 | 69Legon Journal of the HUMANITIES 21) mi-ko fy? no vo fiase 1psSUB-go visit 3psOBJ LOC prison ‘I went to visit him in the prison.’ 22) a gò bit yù fòr haws. lsP- R beat 2oP p farm ‘I will beat you at the house.’ The languages of the Akan dialect cluster often use the general locative construction in combination with the deictic adverb (Riis 1854: 92): 23) mi-h?u no vo ho empriñpriñ 1psSUB-see 3psOBJ LOC Deictic temporal adverb ‘I saw him there just now.’ Welmers gives a similar case (1973: 374): 24) wë-tën èdìbán wë hë they-sell food be-at there ‘They sell food there.’ An excerpt of educated Ghanaian Pidgin English reveals similar uses of for (Huber1999b: 278): 25) d? koti kam, bat ì no si nob? di f? di DET [police] come but 3SB NEG see nobody for there ‘The policeman came, but he did not see anybody there.’ Yourba has the same kind of constructions. The spatial gram ní appears as an adverbial there construction in nìb? and a here construction in nihiŋyi (Bowen 1858: 51). Akuapem Twi also has a similar construction to the for + NP + spatial gram constructions that appear in Ghanaian Student Pidgin, a variety of Ghanaian Pidgin English: “Micia no vo kwanm, I met him on the road” (Riis 1854: 86, bold added). Other examples include (Riis 1854: 244): 26) o-y? adyuma vo afu so 3psSUB-is worker LOC plantation upper_surface ‘He is working in the plantation.’ 27) mu-gyaw me poma vo ne dañ-m 1psSUB-leave 1ps-POSS stick LOC 3psPOSS house- inside ‘I have left my stick in his house.’ 6. Conclusion This paper argues against the widely accepted claim that locative for in Nigerian Pidgin and Ghanaian Pidgin English is derived solely from the preposition for in Metropolitan English. As section 5.1 showed, the locative/possessive construction in Akan languages more than likely motivated the use of a general locative construction that emerged in Pidgin English spoken on the Gold Coast in the Early Modern Period. At the same time, locative na should be considered a template for locative for, since it was in place as a general locative construction in the Portuguese varieties that were established as trade languages since the mid- 17th century. The research in creolistics on the origins of locative for in West African Pidgin Englishes is not conclusive. Metropolitan English and sailors’ jargon from the 17th and 18th centuries contributed to the use of a polysemous locative structure in Pidgin English, but the general spatial gram cannot fully be attributed to the English preposition for. Corum | 72Special Edition , 2012 | 71Legon Journal of the HUMANITIES Niger-Congo-speaking persons were part of the speech networks in the Gold Coast communities that saw the emergence of contact language varieties of European languages (Feinberg 1989: 42): The towns were developed by the Akan and the Ga, not by the Europeans. European forts were constructed where people already lived, not on uninhabited stretches of the coast. The people came first, then the forts. However, the presence of a fort may have contributed to the growth of the population of an area, definitely so if a headquarters fort was located there. The scale of the town, but not the origin, and the population mix of the people, but not their existence, were affected by the European presence. The meanings associated with the general locative construction in West African Pidgin Englishes resemble locative particles and verbs in Upper and Lower Guinea coastal languages too much to discount the Niger-Congo role in the emergence of locative for. Convergence among the three components discussed in this chapter led to the emergence of a general locative construction that Europeans would eventually record as for in their accounts of the Pidgin English spoken on the Lower Guinea Coast. References Abiven, O. 1906. Dictionaire Francias-Malinké et Malinké- Francais: Précédé d’un abrégé de grammaire Malinkée Conakry: Mission des PP. du Saint-Esprit. Adams, J. 1823. Remarks on the Country Extending from Cape Palmas to the River Congo […]. London: G. & W. B. Whittaker. Anyanwu, O. 2010. The Ibibio locative copular constructions. Journal of Language and Linguistic Studies. 6(2): 69-85. Atkins, J. 1735. A Voyage to Guinea, Brazil, and the West Indies. London: C. Ward & R.Chandler. Awobuluyi, O. 1978. Essentials of Yoruba grammar. Ibadan: Oxford University Press Nigeria. Bakker, P. 2008. Pidgins vs. Creoles and Pidgincreoles. In J. V. Singler & S.Kouwenberg (Eds.), The Handbook of Pidgin and Creole Studies .London:Blackwell. pp. 130-157 Blench, R. (2005). Ethiopia Inferior, Vulgarly Angola; Being a New and Accurate Account of the WesternMaritime Countries of Africa. In Six Books. London: Lintont & Osborn. A dictionary of ??an?: An Ijoid language of the Niger Delta. Retrieved from http://www.rogerblench.info/Language% 2 0 d a t a / N i g e r - C o n g o / I j o i d / I b a n i e n g l i s h % 2 0 dictionary.pdf Corum | 74Special Edition , 2012 | 73Legon Journal of the HUMANITIES Boucandé, B. 1849. De la langue creole de la Guinée portugaise (notes sur la Guinée portugaise or Sénégambie méridionale). Bulletin de la Société de Géographie. 12:73-77. Chapman, G. 1603. Monsier D’Olive. The plays of George Chapman. Elizabethian Authors. Retrieved from Childs, G. T. 1995. A grammar of Kisi: A Southern Atlantic Language. Berlin: Mouton de Gruyter. Christaller, J. G. 1875. A Grammar of the Asante and Fante Language called Tshi. Basel: Basel Evangelical Missionary Society. Christaller, J. G. 1881. Dictionary of the Asante and Fante Language called Tshi. Basel: Basel Evangelical Missionary Society. Christaller, J. G., Locher, W., & Zimmerman, J. 1874. A Dictionary, English, Tshi (Asante), Akra. Basel: Basel Evangelical Missionary Society. Duport, Rev. J. H. 1865. Outlines of the grammar of the Susu Language. London: Society for Promoting Christian Knowledge. Esizimetor, D. O. 2010. Historical developments of Naijá. NAIJA: Proceedings of the Conference on Nigerian Pidgin, University of Ibadan, Nigeria. 8-9 July, 2009. http://www.elizabethanauthors.org/olive3.htm Faraclas, N. 1984. A grammar of Obolo. Indiana: Indiana University Linguistics Club. Faraclas, N. 1990. From Old Guinea to Papua New Guinea: A comparative study of Nigerian and Tok Pisin. In J. W. M. Verhaar (Ed.), Melanesian Pidgin and Tok Pisin: Proceedings of the first international conference on pidgins and creoles in Melanasia. Amsterdam: JohnBenjamins. Faraclas, N. 1996. Nigerian Pidgin. New York: Routledge. Feinberg, H. M. 1989. Africans and Europeans in West Africa: Elminans and Dutchmen on the Gold Coast during the Eighteenth century. Philadelphia: The American Philosophical Society. Forman, M. L. 1988. Several prodigal sons: a closer look at the preposition na in Zamboangueno and Asian Creoles Portuguese. In R. Hadlich & J. D. Ellsworth (Eds.), East meets West: homage to Edgar C. Knowlton, Jr. pp. 63- 71 Honolulu: University of Hawaii, Department of European Languages and Literature, College of Languages, Linguistics and Literature. Gamble, D. P. 1963. Elementary Wolof grammar. In Manesset & Sauvegeot (Eds.) Wolof and Sérèr: Etudes de Phonotique et de Grammaire Descriptive . Publications de langues et literature 12. Dakar: Université de Dakar pp. 131- 161 Corum | 76Special Edition , 2012 | 75Legon Journal of the HUMANITIES Gilman, C. 1980. The origin of Cameroon Pidgin dialects. Anthropological Linguistics.22 (9)363-372. Hancock, I. 1976. Nautical sources of Krio vocabulary. Linguistics, 14: 23-36. Hancock, I. 1986. The Domestic Hypothesis, diffusion and componentiality: An account of Atlantic Anglophone creole origins. In P. Muysken & N. Smith (Eds.), Substrata vs. Universals in Creole Genesis. Amsterdam: John Benjamins. pp. 71-10 Hancock, I. 1987. A preliminary classification of the Anglophone Atlantic creoles, with syntactic data from thirty-three respresentative districts. In G. Gilbert (Ed.), Pidgin and Creole Languages: Essays in Memory of John E. Reinecke . Honolulu: University of Hawaii Press. pp. 264- 233 Heine, B. 1997. Cognitive Foundations of Grammar. New York : Oxford University Press. Houis, M. 1963. Étude Descriptive de la Langue Susu. Mêmoires de L’Institut Français D’Afrique Noire 67. Dakar, Senegal: IFAN. Huber, M. 1999a. Atlantic English Creoles and the Lower Guinea Coast: a case against Afrogenesis. In M. Huber and M. Parkvall (Eds.) Spreading the Word: The Issue of Diffusion among the Atlantic Creoles. London: University of Westminster Press.pp. 81-110 Special Edition , 2012 | 77Legon Journal of the HUMANITIES Huber, M. 1999b. Ghanaian Pidgin English in its West African context. Amsterdam: John Benjamins. Huber, M. & Velupillai, V. 2011, August. German colonial sources and the history of West African Pidgin English: A first analysis of the holdings of the Deutsche Kolonialbibliothek. Conference presentation at the summer meeting of the Society for Pidgin and Creole Linguistics, University of Ghana, Legon, Ghana Innes, Gordon. 1966. An Introduction to Grebo. The Hague: Mouton and Co. Kamarah, S. U. 2007. A Descriptive Grammar of K?Themna. Muenchen: Lincom Europa. Kihm, A. 1994. Kriyol Syntax: The Portuguese-based Creole Language of Guinea- Bissau. Amsterdam: John Benjamins Press. Koelle, S. W. 1854. Outlines of a Grammar of the Vei Language, together with a Vei- English. Vocabulary: A n d Account of the Discovery and Nature of the Vei Mode of Syllabic Writing. London: London Church Missionary House. Lord, C. (1993). Historical Change in Serial Verb Constructions. Amsterdam: John Benjamins. Mann, C. C. 1993. Polysemic functionality of prepositions in pidgins and creoles: The case of ‘Fò’ in Anglo- Nigerian Pidgin. In F. Byrne & J. Holm (Eds.), The Atlantic meets the Pacific: A global view of pidginization and creolization (pp. 57-67). John Benjamins: Amsterdam. Corum | 78 Martinus, F. 1997. The Kiss of a Slave: Papiamentu's West- African connections. Willemstad, Curaçao: De Curaçaosche Courant. McWhorter, J. 1999. Skeletons in the closet. Anomolies in the behavior of the Sarammacan copula. In J. Rickford & S . Romaine (Eds.), Creole Genesis, Attitudes and Discourse: Studies Celebrating Charlen J.Sato Amsterdam: John Benjamins. pp. 121-142 Nyländer, Rev. G. R. 1814. A Grammar and Vocabulary of the Bullom Lanuage. London: Ellerton and Henderson. Osam, E. K. 1997. The lone preposition in Akan. Ghana Studies C o u n c i l N e w s l e t t e r , 1 0 . R e t r i e v e d f r o m http://people.tamu.edu/~yarak/gsc97osam.html Rediker, M. 1987. Between the Devil and Deep Blue Sea: Merchant Seamen, Pirates, and the Anglo-American Maritime World, 1700-1750. Cambridge: Cambridge University Press. Riis, H. N. 1853. Elemente des Akwapim-Dialects der Odschi- Sprache Enhaltend Grammatishe Grundz rige und Wo r t e s a m m l u n g n e b s t s e i n e r S a m m l u n g v o n Sprichtwoerte der Eingebornen. Basel: Bahnmaier. Riis, H. N. 1854. Grammatical Outline and Vocabulary of the Oji language: With a special Reference to the Akwapim- dialect, together with a Collection of Proverbs of the Natives. Basel: Bahnmaier. Rotini, O. and Faraclas, N. (n.d.). A Dictionary and Grammar of Nigerian Pidgin. Unpublished manuscript. Saker, A. J. S. 1855. Grammatical Elements of the Dualla Language. With a Vocabulary Compiled for the Use of Missionaries and Teachers. Cameroon: Baptist Mission Press. Sauvant, L. P. 1913. Grammaire Bambara. Algere: Maison Carrée. Schuchardt, H. 1888. Beiträge zur Kenntnis des kreolischen R o m a n i s c h I I : Z u m N e g e r p o r t u g i e s i c h e n Senegambiens. Zeitschrift für Romanische Philologie. 12: 301-322. Senghor, L. S. 1963. L’harmonie vocalique en Sérèr. In Manesset & Sauvegeot (Eds.), Wolof and Sérèr: Etudes de Phonotique et de Grammaire Descriptive.Dakar: Université de Dakar. pp. 287- 294 Shakespeare, W. 1623. All’s well that Ends well. The complete works of William Shakspeare. Retrieved from Shumway, R. 2004. Between the castle and the golden stool: Transformations in Fante society in the eighteenth century (Unpublished doctoral dissertation). Emory University, Georgia. Singler, J. V. 1987. Where did Liberian English na come from? English World-Wide. 8: 69- 95. Swift, L. B. et al. 1965. Fula Basic Course. Washington D.C.: Foreign Service Institute. Taylor, D. 1960. Language shift or changing relationship? International Journal of American Linguistics, 26(2): 155-161. http://shakespeare.mit.edu/works.html Corum | 80Special Edition , 2012 | 79Legon Journal of the HUMANITIES Taylor, D. 1971. Grammatical and lexical affinities of creoles. In D. Hymes (Ed.), Pidginization and Creolization of languages . Cambridge: Cambridge University Press. pp. 293-296 Trutenau, H. M. J. (trans.) 1971. Christian Protten's 1764 Introduction to the Fante and Accra (Gã) languages. 2 n d edition. Ga Dictionary Project: Documents and Studies 1. London: Afro Presse. Una, Rev. F. X., 1900. Efik Language. Uyo, Nigeria: Modern Business Press. van den Berg, M. 2000. Mi no sal tron trongo: Early Sranan in court records 1667- 1767 (Unpublished master’s thesis). University of Nijmegen, Netherlands. Welmers, 1973. African Language Structures. Berkeley: University of California Press. Welmers, W.M. E. 1976. A Grammar of Vai. Berkeley: University of California Press. Westermann, D. 1924. Die Kpelle-Sprache in Liberia: Grammatische Einführung, Texte, und Wörterbuch. Berlin: Reimer. Williamson, Kay. 1965. A Grammar of the Kolokuma Dialect of ?j?. London: Cambridge University Press. Wilson, J. L. 1847. A Grammar of the Mpongwe Language with Vocabulary. New York: Snowden & Prall Printers. Wilson, W. A. A. 2007. Guinea Languages of the Atlantic gGroup: Descriptions and Internal Classification. Schriften zur Afrikanistik 12. Frankfurt: Peter Lang. Zimmerman, J. 1858. A Grammatical Sketch of the Akra- or Ga- language, with some Specimens of it from the Mouth of the Natives and a Vocabulary of the Same, with an Appendix on the Ada? me-dialect. Stuttgart: Basel Missionary Society. Corum | 82Special Edition , 2012 | 81Legon Journal of the HUMANITIES