NSF Grant #BCS-0113826 RF Project # 741792

Final Report. December 2004.

The influence of West African languages on the TMA systems of two Surinamese creoles.

Project description.

The main objective of this research project is to investigate the role of influence from West African (specifically Gbe) languages in shaping the TMA system of the early creole that emerged on the plantations of Suriname roughly between 1680 and 1720. For this we investigate and compare two modern conservative descendants of the plantation creole and several of the modern conservative descendants of the main West African inputs to its formation. With respect to the former we focus on the mother tongue of the coastal Afro-Surinamese population, Sranan (SN), and its more conservative sister language, the Eastern Maroon Creole (EMC) spoken by the Aluku-Boni, Ndjuka and Paamaka ethnic groups. With regard to the West African input, we focus primarily on the Gbe group of languages and secondarily also on Kikongo and Akan. The linguistic analysis is based primarily on tape-recorded and elicited data collected in Suriname (SN and EMC) and Benin (Gbe). The data on the secondary input comes from the published literature and formal elicitations where that was possible. The study is intended as a contribution to a theory of creole formation, the principles and constraints which regulate it, and its relationship to other outcomes of language contact.


PI. Donald Winford, The Ohio State University.

Co-PI: Bettina Migge. University of Frankfurt, Germany.


In Benin: Professor Hounkpati Capo, University of Benin. Director, Labo Gbe (Int.).

Daniel Gagnon (fieldwork coordinator). Labo Gbe (Int.).

In Suriname: Dr. Robby Morroy, Teacher Training Institute, University of Suriname.

In Holland: Dr. Enoch Aboh, Dept. of Linguistics, the University of Amsterdam.

Dr. James Essegby, Dept. of Linguistics, Leiden University.

Other organizations involved:

The University of Frankfurt.

Labo Gbe (Int.). (Laboratory for research on Gbe languages, Come, Mono, Benin).

IRD (Institut de recherche pour le développement, laboratoire linguistique Cayenne, French Guiana, France)

Objectives and Significance.

In accordance with the methodological frameworks outlined by scholars such as Weinreich (1953), Thomason and Kaufman (1988) and others, the empirical investigation of the origins of the Surinamese creole TMA systems involves the following stages:

·  An account of the historical and social aspects of the contact setting in order to determine (a) the linguistic inputs and (b) how factors such as the community settings, demographics and patterns of social interaction affected the outcomes of the contact.

·  Analysis and comparison of the TMA systems of the relevant input languages so as to identify the possible sources of the functional categories.

·  A principled explanation of the emergence of the TMA system of the early SPC which accounts for the “selection” of the categories and their means of expression. Such an explanation would be based on sociohistorical as well as linguistic evidence.

·  An account of the linguistic mechanisms involved, in terms of more general processes and principles operating n contact situations generally.

·  Some explanation of the emergence of a stable and relatively uniform creole grammar out of the different contact varieties that are likely to have emerged under the influence of different inputs.

4.1. Inputs and social contexts of the contact.

Our account of the sociohistorical background to the emergence of the (varieties of the) early Surinamese plantation creole is given below, in Section 5.1. Our account has implications for current debate among creolists concerning the nature of the linguistic inputs to creole formation, and the relative roles of adults vs children in this process (DeGraff 1999). The evidence we present will support the view that the creators of the early SPC were adult Africans who drew heavily on the resources of their L1’s to fashion a new medium of inter-group communication. We assume that in the initial stages, there were second language varieties of English which provided the lexical input for a pidginized variety that served as a lingua franca for inter-ethnic communication. In all probability, this early pidginized variety was a language without functional heads (TMA markers, complementizers, etc.) or other grammatical devices such as movement rules, We further assume that this pidgin became a target for learning by new arrivals from Africa who elaborated it, creating new varieties that in turn served as targets of learning and restructuring for successive waves of later imported slaves. This scenario created the conditions for the kinds of creative L1 retention and innovation typical of contact situations generally, but especially of creole formation.

Research Methodology

Our research plan is organized as follows:

·  Identification of the primary linguistic inputs to the contact setting that produced the early (varieties of the) Surinamese plantation creole.

·  Justification of our selection of contemporary communities in Suriname and Benin for purposes of data collection.

·  Methods of data collection.

·  Analysis and comparison of the TMA systems of the West African inputs and the Surinamese creoles.

·  Explanation of the emergence of the creole TMA systems in the light of the sociohistorical, demographic and linguistic evidence.

·  Implications of our findings for a theory of creole formation and theories of language contact outcomes in general.

5.1. Justification of choice of linguistic inputs.

Both the available sociohistorical documentation and the linguistic evidence from recent studies provide evidence that varieties of Gbe in particular played a major role in the formation of (varieties of) the early Surinamese plantation creole (Arends 1989; Bruyn 1994, 1995, 1996; Migge 1998a & b, 2000). The other West African languages which were also present in the formative contact situation, Kikongo and Akan, appear to have contributed much less to the grammar of the SPC, though Kikongo, and to a small extent, Akan, contributed to the vocabulary (Arends 1994, Huttar 1985). It is possible that Kikongo and Akan may also have influenced the emergence of the TMA system of the SPC . If (some) Kikongo varieties brought by slaves to Suriname in the 17th to 18th centuries employed periphrastic systems of TMA marking (Mufwene 1988, 1990b), they may well have reinforced the Gbe (and Akan) influence on the verb complex of the SPC, since Gbe, like the SPC, employs a periphrastic system. Kikongo influence may also have affected the semantics/pragmatics of particular TMA categories in the SPC. We plan to investigate these possibilities using the available literature on TMA in Kikongo (e.g., Mufwene 1988, 1990, Laman 1912).

The sociohistorical evidence indicates that the plantation creole was created roughly between 1680 and 1720 - the period during which the plantation economy replaced the earlier homestead-based economy established by English planters and continued under Dutch rule (1651 - 1680). The contact setting involved three broad groups of people - the Europeans, the early or "elite" slaves, and the new or "field" slaves. The three groups differed from one another with respect to their relative size, and social cohesion, and their members' social status, linguistic background, work tasks and patterns of interaction. The Europeans were the smallest group, making up less than 5% of the entire population. They made up the top of the social hierarchy and were responsible for organizing and delegating the different tasks on the plantation. They were mainly speakers of different European languages but also used L2-like varieties of English in communication with the slaves. They mainly interacted among themselves and with the elite slaves primarily for work-related issues. In the initial period, they remained on the plantation for a relatively long period of time. The elite slaves made up about 15% to 20% of the entire population and an intermediate social position on the plantation. Some of the early or elite slaves worked as domestic servants for the Europeans, others performed more skilled tasks on the plantation, and yet others were in charge of the main work force, the new slaves. They supervised and organized their work on the fields and in the sugar mills and they introduced the new arrivals to life on the plantation. They were speakers of different African languages but their main means of (interethnic) communication among themselves and with the Europeans and slaves were the L2-like varieties of English that they had acquired in Suriname or in other colonies prior to the onset of the plantation economy. This group had a relatively high degree of cohesion since its members had either already spent a relatively long period of time on a plantation or homestead, or were born there. They interacted primarily among themselves and for work-related issues they also interacted with the Europeans and the new slaves. Finally, the largest group on the plantation were the new slaves who were arriving during this period. They constituted between 75% to 80% of the entire population but had the lowest social status. They were responsible for the planting, harvesting, transporting etc. of the plantation's main cash crop(s). They had a low degree of social cohesion since the majority of them were not familiar with life on the plantations, were constantly changing, and did not share common social and linguistic conventions. In the initial plantation period (1684-1695), the new slaves consisted of equal numbers of Kikongo and Gbe speakers and in the second part more than 70% of the new slaves were Gbe speakers. The new slaves interacted mainly among themselves and to a small degree also with the elite slaves mainly about work-related matters. These facts suggest, first, that the plantation creole was created in the interaction among the new slaves and between them and the elite slaves. Second, the inputs to the formation of the plantation creole were the native languages of the new slaves, particularly Gbe, and the L2 varieties spoken by the elite slaves. However, the former played a more important role, since the interactions among the slaves were more frequent and communicationally more complex. Third, in the interactions among themselves, the new slaves accommodated to each other linguistically since they were socially on a par while in the latter settings the new slaves accommodated to the elite slaves since they were socially in a subordinate position. Fourth, the new slaves were not able to acquire the L2 varieties of English and shift to them. They only learned and adopted a limited range of highly salient items from them in the interactions with the elite slaves. As these features gradually became part of their linguistic repertoire, they also employed them in their interactions with their fellow field slaves by combining them with material from their native languages.

5.2. Selection of communities for data collection.

5.2.1. Benin.

The Dutch slave coast that supplied most of the slaves to Suriname from 1680 - 1720 corresponds roughly to the coastal areas of modern Benin and Togo, hence our decision to investigate Gbe varieties in these areas. For purposes of this study, we follow Capo's (1988,) classification of the five major subgroups of Gbe, as shown in Table 2.

Table 2: Major subgroups of Gbe


Subgroups: Vhe (Ewe) Gen Ajá Fon Phla-Phera

Languages: Wací Glijí Dogbó Maxí Alada

Awlan Anéxo Stádó Gun Phla (Xwla)

Vo Agoi Hwe Wéme Phelá (Xwelá)

etc. etc. etc. etc. etc.

The historical records do not allow us to accurately determine exactly which Gbe groups were present in what numbers in Suriname from 1680 - 1720. However, research by Pazzi (1979) suggests that the Gbe peoples who were at the time living on the coast between the present-day towns of Cotonou and Lome and roughly 100 miles inland were primarily involved in the slave trade as either slaves or traders. They were the ancestors of the present-day Gen, Phla-Phera and various Vhe groups, among others. The present investigation will employ representatives of each major subcluster of Gbe for the linguistic analysis. We chose varieties that are spoken by groups directly descended from those that were involved in the slave trade. These groups are still functioning ethnolinguistic communities in the same area today, and have not been subject to any significant new external linguistic influence since the period of the slave trade (Migge 1998a & b, 2000).

5.2.2. Suriname.

The Surinamese communities we selected include Afro-Surinamese working-class groups in the capital, Paramaribo, and rural groups in Coronie, all of whom speak Sranan as their first language. The varieties of Sranan that are used as a first language by working-class urban and rural Afro-Surinamese tend to be more conservative and less subject to influence from Dutch than those used by middle-class persons, or by other ethnic groups for whom Sranan is often a second language (Winford 1997).

The EMC data come mainly from monolingual subsistence farmers who were raised and live in the remote ancestral villages of the Ndyuka and Paamaka communities in eastern Suriname. The EMC varieties as spoken in the villages are not subject to much external influence, and represent the most conservative descendants of the early plantation creole. The escaped slaves who created these communities in the mid 18th century employed (varieties of) the plantation creole as their community language. They remained in close contact with the plantation slaves, replenishing their communities with (mostly female) slaves from that group (Hoogbergen 1990a, b). These communities did not have contact with other language groups, and have remained in relative isolation until recently. The comparison of the EMC with Sranan will therefore allow us to identify those shared aspects of grammar that are likely to have descended directly from the SPC.

5.2.3. On the use of contemporary data - a caveat.

One of the major problems facing the student of creole genesis is the unavailabilty of data from the period of creole formation. This makes it difficult to determine the structure of the inputs at the time of creole formation some 300 years ago. However, as Thomason (1993) reminds us, 300 years is not a very long time in the history of a language. We can assume (unless there is strong evidence to the contrary) that the contemporary grammars of the relevant input languages are quite similar to what they were in the 17th to 18th centuries.

The choice of more conservative (rural) varieties of SN and the EMC was based on the assumption that such varieties preserve older features of the language more faithfully than urban ones. Moreover, Ndjuka itself is an excellent point of comparison and contrast with Sranan. The former has remained quite isolated from external (e.g. Dutch and Sranan) influence over the centuries, and arguably provides a better window on the earlier plantation creole than its sister. For example, Sranan has borrowed a great deal of vocabulary as well as some grammatical formatives from Dutch in the recent past, whereas Ndjuka shows none of this influence, at least in its TMA system.