— 6 — Pidgin and Creole Languages Fig . 6.1. An advertisement for Sun Flower tinned fish in Tok Pisin. The slogan says 'It's my very foundation' In his speech to the English-Speaking Union Conference in Ottawa, the Duke of Edinburgh made reference to one of the best-known pidgin and Creole languages in observing that he was 'referred to in that splendid language as "Fella belong Mrs Queen"'. Although the Duke was right to consider Tok Pisin Pidgin and Creole Languages 163 ('talk pidgin') spoken in Papua New Guinea as a language rather than a dialect of English, he was wrong about his designation. He would, be called 'man bilong kwin'. Contrary to what many Europeans think about Tok Pisin, fella cannot be used in this way at all to mean 'man' or 'husband', so the Duke's statement is ungrammatical. Fella can be used only as a suffix in Tok Pisin and has a number of grammatical functions, e.g. to mark adjectives and numerals, as in tupěla blakpela pik 'two black pigs', and to mark the second person plural form of 'you', as in yupela i no ken go 'you (plural) cannot go'. Although there is no agreement on how to define pidgins and Creoles in precise linguistic terms, or on where they came from, all linguists recognize that there is such a group of languages. Their distinctiveness lies not so much in terms of a common historical origin, but in shared circumstances of socio-historical development and use. At present there is no way of deciding whether a language constitutes a pidgin or Creole unless reference is made to three criteria: linguistic, social, and historical. The term 'creolepis generally applied to Eidgins wfn^hJmyjJie££irr^_naJLYJzed, aTffiough not.juJU!hTgjnsjtójy£ree_^ stage. j^idgins^^rc^tras^t^arenobody's first language^They_eome into^ existejic^jiT_conJactsituations, where they are used by speakers with different language backgrounds to fulfil certain restricted communicative purposes, typjcally trade. Structurally speaking, they are simplifiedTärígHages characterized by a minimal lexicon, little or no morphology, and limited syntax. In this chapter I will say sSřleThTřlglÍDOuTIl^ structure, and use of pidgins and Creoles in society. Distribution Pidgin and Creole languages are spoken mainly in Third World countries. Their role there is intimately connected with a variety of political and social questions. There are probably more than a hundred pidgin and_creole languages in_ daily use around the" world, and more speakers of these languages than there are of Swedish. The exact number of languages is difficult to establish because it depends on how we define the terms 'pidgin' and 'creole'. Most pidgins and Creoles are basedjoji_£uropjejin_languages, 164 Pidgin and Creole Languages in particular Sganishj^Portuguese, French, English, and Dutch. However, those based on English are more numerous than those based on any other languagej'liUešTingTôl'Ré greater spread of English than any other metropolitan language. The next largest group is based on French, and a much smaller numiHer based on non-European languages, such as Sango spoken in the Central African Republic. The term '-based' means that the bulk of the lexicon_js drawn frojrL4hat_language, while the grammatical structure typically shows influence from other (usualtyTion-European) languages. Tfíese^ÔtherTäríg^uageš äT^~~reíeíŕrednEo~as~*tKe~l^^^^^^^^'. Thus, when scholars speak of English-based Creoles, they are referring to all those Creoles which have taken most of their vocabulary from English. Terms such as 'English-lexicon' or 'English-lexifier pidgin/creole' are also used andjhe lexifier^ language is sometimes called the 'supérštrafb'ľ'While it has often been the case that scholars have treated English-based pidgins/creoles as dialects of English and French pidgins/creoles as Romance dialects, etc., most now recognize that Creoles are languages in their own right with an indepjiident,strjigtur£.. They are not parasitic systems or corrupted versions of the languages to which they are most closely related at the lexical level. It is customary practice to label pidgins and creoles with a formula which includes their location and their principal lexifier language, e.g. Chinese Pidgin English, Berbice Creole Dutch, Rabaul Creole German, etc. Since some pidgins and Creoles may change their lexical affiliation through a process known as relexification, such as Berbice Dutch did when Dutch was superseded by English as the superstrate language, these labels are not entirely satisfactory. Other Creoles such as Tok Pisin and Sranan, an English-based variety spoken in Surinam, have been affected by this kind of change too. However, no one has suggested replacing the name Berbice Creole Dutch with Berbice'Creole English. These labels are unsatisfactory for other reasons too. One is that they imply a separation between lexicon and syntax and givemore weight to the lexicon in deciding relationships among languages. Another difficulty is that the first term in such labels may be ambiguous as to whether it specifies a language, a group of speakers, or a geographical location. Thus, in the name Hawaiian Creole English the adjective Hawaiian is ambiguous Pidgin and Creole Languages 165 because it could refer to the geographical location of Hawai'i, to people of Hawaiian descent, or to the Hawaiian language. Linguists' names for pidgins and Creoles are not always widely used by the speakers of the language themselves. Thus, Tok Pisin has sometimes been referred to by linguists as Neomelanesian or New Guinea Pidgin English, while its speakers call it Tok Pisin, or simply pidgin. Similarly, speakers of Torres Strait Creole English call their language Broken (i.e. broken English), while speakers of Australian Creole English call their language Kriol and Hawai'i Creole English speakers call their variety 'pidgin', etc. Sranan is sometimes also called Sranan Tongo 'Surinam tongue' or Taki-Taki'talk-talk'. Two major groups of languages, the Atlantic and the Pacific, are re^ogňlžečTáccording to historicál7^ěo^řapTiic7^ň^~lingíIistic factors. The^Atlantic, group was established primarily during the se^nteentlTjjng_ejghteerith centuries in the Caribbean and West Africa, while the {Pacify group originated plTníäríhTin the nin^teelrtHrThe^Aflanti.c"creoles were largely products of the.slave trade irPWest Africa which dispersed large numbers of West Africans to the Caribbean. Varieties of Caribbean Creoles have also been transplanted to the United Kingdom by ,We_stJndian immigrants. The languages~5f the Atlantic share_a„cqrnmon substrate ar|c^hj>pjay many common features.. In t^e^Pacific-'diffeTent languages formed the substratum and socio-cültufal conditions were somewhat different than in the Atlantic. Although |the plantation_setting was crucial for the emergencejDLpidgiijkJoJboth areas, in the Pacificjabqrers were recruited and indentured rather than slaves. Apart from Hawai'i, a histor^pf gradual creolization has distinguished the Pacific from the Atlanticl^artjčúliííy^Caribbean)_creoles, whose transition has beenjnore abrupt. While this traditional grouping is geographically convenient, it may obscure a far more complex picture of interrelationships. It also does not take into account that there was pjxDpjibJy_j>ome interaction between Atlantic and Pacific, and that judgements about the degree to which creolizatiorTwas abrupt depend on historical evidence which is usually lacking. Linguistic evidence to support the view that sailonT^píayedarT important role in spreading linguistic features across vast areas can be found inTfie~Täct~tHat there is some lexical sharing among distant pidgins such as Hawai'i Pidgin English, Chinook Jargon (a 166 Pidgin and Creole Languages contact language used for trade in the Pacific Northwest), and Eskimo Jargon. The term kanaka (< Hawaiian 'human being', 'person', 'man') occurs in both English- and French-based pidgins and Creoles throughout the Pacific, although its meaning and, in particular, connotations vary somewhat from language to language. In Tok Pisin, for instance, it was originally a term used by Europeans to refer to the indigenous population, often pejoratively, and it has these negative overtones in Tok Pisin today. To refer to someone as a 'kanaka', especially a 'bus kanaka' (bush kanaka), is insulting because it implies they are backward and unsophisticated. In the French overseas territory of New Caledonia, however, the term is used as a symbol of Melanesian unity and pride in the face of the continuing French colonial administration. In 1984 Melanesian activists declared an independent state to which they gave the name Kanaky, but independence has yet to be recognized by France. Interestingly, in Hawai'i Creole English, the term is no longer much used, despite its presence in the Hawaiian language. The word kanaka is also found in non-European-based varieties such as Chinook Jargon, a trading language used along the Pacific Northwest coast, where it and other items spread after the development of the fur trade along the north-west coast of North America in the late eighteenth century. Similarly, Eskimo Jargon has kaukau 'food', itself a loanword in Hawaiian, introduced from Chinese Pidgin English chowchow. The term kaukau is still used in Hawai'i Creole English to mean 'food' or 'eat', although most of the younger generation use the term grind in the sense of 'eat'. During the nineteenth century contacts across the North Pacific became routine. The Hawaiian Islands became a frequent port of call'and a wintering place for ships. Recent work places a great deal of emphasis on sailors in the spread and stabilization of pidgin English in the Pacific. While some scholars have claimed a relationship between the English-based pidgins and Creoles wjth those elsewhere, evidence for a world-wide English nautical pidgin English is not clear. Although a number of characteristics have been identified which are shared by the Pacific varieties with those in the Atlantic, the Atlantic features were drawn from contemporary rather than historical records, and not enough research has been done in the Atlantic area to support a historical connection. Pidgin and Creole Languages 167 There is still some dispute over the status of the variety of English spoken on Pitcairn and Norfolk Islands. The conditions under which this variety developed were exceptional and made famous through the story of the mutiny on the Bounty. Before 1790 Pitcairn Island was uninhabited, but the arrival of twenty-eight people (nine of them mutineers from HMS Bounty, and the others Polynesians) marked the beginning of a community which existed in isolation for thirty-three years. Most people were resettled on Norfolk Island after 1856. Although there have been a few studies of Pitcairn-Norfolk speech, no consensus on its classification has emerged. Pitcaimese is clearly identifiable as having an English-based lexicon and shares many features with English-based and other Creoles, though it also shows some 'exceptional' features. For example, the presence of consonant clusters, which are not found in Tahitian, is hard to explain because most pidgins and Creoles tend to have a simple syllable structure consisting of a consonant and a vowel (see below). From v one description it appears that Pitcairn-Norfolk displays quite a large number of the characteristics typical of Creoles more generally. However, it does not show any distinctly Pacific characteristics. This is of course not surprising since it has no historical link with earlier varieties of Pacific Pidgin English and developed in isolation from them. Origins Traditional approaches to historical change have relied on the family tree model, which is based on the assumption that over time languages gradually diverge from a common ancestor. This model has been widely applied to explain the historical origins of pidgin and Creole languages and has been referred to as the 'monogenetic hypothesis', i.e. that pidgins and Creoles are to be derived from a single common ancestor. Many espoused the view that all the European-based pidgins and Creoles were originally descended from a fifteenth-century Portuguese pidgin first used along the African coast and later carried to India and the Far East. This pidgin may have been a relic of Sabir, the medieval lingua franca believed to have been- the language of the Crusaders and a common Mediterranean trading language. While a common Portuguese origin would account well for certain lexical similarities 168 Pidgin and Creole Languages found across the Atlantic and Pacific pidgins and Creoles, one has to invoke the notion of 'relexification' to account for the many differences which exist between the Romance and Germanic Creoles. Of course, the monogenetic theory would have nothing to say about the origins of non-European-based pidgins and Creoles. Most linguists reject the monogenetic view because pidgins and Creoles are typically formed through a convergence of linguistic structures from more than one genetic stock. It was the Romance linguist Hugo Schuchardt (1842-1927), often called the father of Creole studies, who used data from pidgin and Creole languages to argue against prevailing nineteenth-century views on the regularity of sound change. Conversely, polygenetic theories stress the idea that pidgins and Creoles arose independently but developed in parallel ways because they used common linguistic material and were formed in similar socio-historical terms. Stated in this way, there is a universalist element to the polygenetic view (see further below). Proponents of substratum explanations for the origins of pidgins and Creoles claim that models for many of the structures common to these languages can be found in the substratum; particularly, the languages of West Africa are believed to have had a great influence on the Atlantic Creoles. Similar arguments about the importance of the Oceanic substrate have been made more recently for the Pacific pidgins and Creoles. Typologically, it could be argued that Creoles show more similarities to one another regardless of their base language than they do to their lexifier or base language. Some linguists find the concept of a 'Creole syntax' uncontroversial, even though there might be disagreements about exactly which features are included. Others, however, are sceptical about the extent to which Creoles can be regarded as a distinct structural type. This has led many scholars to explain the origin of pidgins and Creoles in universalist terms. At one level, one can appeal to universal pressures of a functional type which exist in any communicative situation where speakers do not share each other's language. There will be pressure towards simplification, greater reliance on context, slower rate of speaking, etc. The language used to foreigners, or 'foreigner talk', e.g. I no speak English, is a good example of these strategies and may result in an incipient pidgin. Some of these same features can be found in the speech addressed to young Pidgin and Creole Languages 169 children, or 'baby talk', e.g. Daddy go bye-bye. In situations where the communicative partners are not social equals, the access of the subordinate group to the language of the dominant group may be only partial. Therefore, imperfect learning may lead to the emergence of a reduced version of the upper group's language. A more far-reaching type of universalist explanation is based on an alleged innate 'bioprogram', which contains a blueprint for the features the Creole must have. The bioprogram hypothesis links the emergence of Creoles with first-language acquisition as well as with the evolution of language in the human species more generally. It claims that the features which children learn early and effortlessly are among those prominent in Creole languages. The existence of a bioprogram is argued for by appeal to the alleged lack of adequate input which children receive in a pidgin-speaking community. Under such circumstances children fall back on the bioprogram to produce rules for which they have not had models in the input from the older generation. This theory has provoked a great deal of controversy, particularly since its claims cannot be tested directly. No children are currently acquiring a Creole language under the relevant circumstances required to validate the operation of the bioprogram. Many linguists prefer not to invoke only one source, but appeal instead to a convergence of forces to account for the development of creole features. Some examples will be given below. Thus, while some characteristics of particular Creoles obviously have their origin in particular substratum languages, others require a universalist explanation. Structure The boundary between pidgins and Creoles cannot be defined in purely linguistic terms. Thus, some languages, such as Tok Pisin and West African Pidgin English spoken widely in West Africa, may exist in both pidgin and Creole forms, which display different degrees of structural expansion and stability depending on whether they are used by first- or second-language speakers. Creolization can take place at any point during the pidgin's life cycle, ranging from a jargon to an expanded pidgin. The term 'jargon' refers to a speech variety with a minimal linguistic system and great individual variation used for communicating in limited situations between speakers of different languages, e.g. trade, while a pidgin 170 Pidgin and Creole Languages Type 3 jargon í stabilized pidgin expanded pidgin Creole e.g. New Guinea Tok Pisin has a certain degree of stability. Three possible routes to creoliza-tion are depicted in Fig. 6.2. Once creohzation has occurred, the evolutive changes that take place thereafter may make it impossible to identify a prior Creole or pidgin stage, as in the case of Black English in the United States, which many linguists believe to be a variety in the late stages of decreolization. This term is used to refer to changes which bring a Creole closer to its superstrate language, and there is some dissatisfaction with it because some changes may not be motivated by influence from the lexifier language. In other words, not all changes are unidirectional. The terms 'basilecť, 'mesolecť, and 'acrolecť are used to describe the range of varieties often found after creolization, as shown in Fig. 6.3 illustrating the Guyanese Creole continuum. They are arranged along a post-creole continuum which has acrolectal (those varieties closest to the superstrate) at one end and basilectal (those varieties furthest from it) at the other. Mesolectal varieties are intermediate. This can be illustrated with this set of eighteen variants from the Guyanese post-creole continuum. The ordering of the varieties in the model presented here makes no claim about their diachronic development since it seems likely that in many cases mesolectal and acrolectal varieties were present even in the earliest phases of Creole formation and do not always develop after creolization in adaptation to the superstrate. Thus, acrolectal varieties cannot be Type 1 jargon Creole e.g. West Indian English Creole Type 2 jargon * stabilized pidgin creole e.g. Torres Straits Creole Fig. 6.2. Three types of creo Pidgin and Creole Languages Guyanese Creole -«^-------------------------------------------- basilect mesolect 171 English acrolect Guyanese Creole (basilect) mesolect English (acrolect) mi gii am mi bin gii am mi bin gii ii mi bin gi ii mi di gii ii mi di gi hii a di gii ii a di gi ii a did gi ii a did giv ii a did giv hii a giv ii a giv im a giv him a geev ii a geev im a geev him I gave him Fig. 6.3. The Guyanese Creole continuum regarded as necessarily 'later' than basilectal ones. Creolization and decreolization can be coexistent, as they are for instance in Papua New Guinea. Tok Pisin is in the process of being nativized by the younger generation at the same time as increased accessibility to English schooling is creating a range of varieties which are neither Tcfc Pisin nor English. Because creolization can occur at any stage in the development continuum from jargon to expanded pidgin, different kinds and degrees of structural repair may be necessary to make the pidgin fully adequate to meet the demands placed on it for use as a primary language. Sources for new structures vary too. Some whose primary interest is in the first type of creolization, where a pidgin is abruptly transformed into a Creole within one generation, believe that bioprogram universals provide the most important input into the newly emerging creole. There is relatively little to be said about morphology in pidgins and Creoles since lack of it is one of the defining characteristics of the pidginization process. The absence of highly developed inflectional morphology was generally equated with lack of 172 Pidgin and Creole Languages grammar and thought to reflect the primitiveness of both the language and its speakers. Pidginization can entail loss of all bound morphology, many free grammatical morphemes, and even a large part of the vocabulary. We can compare Yimas, a language spoken in the Sepik region of Papua New Guinea, with Pidgin Yimas, a variety used for inter-tribal trade. Yimas proper is ;i highly complex language morphologically while Pidgin Yimas is vastly simplified. For example, while Yimas proper makes seven tense distinctions, Yimas Pidgin makes only two. Yimas proper distinguishes four numbers in its pronominal paradigm (singular, dual, paucal, and plural) while Yimas Pidgin has only three. Duiil is used when only two things are referred to and paucal when there are more than two but not many. This economy appears to represent convergence towards Arafundi (the language spoken by the main trading partners of the Yimas), which has only three. Most of the grammatical features commonly found in European languages are usually lost, e.g. gender and case. Loss of gender, which in European languages only vaguely corresponds with the semantic concept of sex, will not involve a conceptual loss, but it will result in a substantial simplification. Lexicalization of gender (expression of gender by lexical means) results in irregularity, as can be seen in the English sets like sow/boar, cow/bull, marc/ stallion as compared to Tok Pisin pik merilpik man, bulmakau merilbulmakau man, hos manlhos men (where men means 'woman' and man means 'man'). Compare also Jamaican Creole man hagl uman hag 'boar/sow'. In Rabaul Creole German there is an invariant definite article, de, instead of the three-way gender distinction found in standard German (cf. der Baum 'tree', das Mädchen 'the girl', and die Katze 'the cat'). This article does not take any case marking and the infinitive forms of verbs are often used instead of the standard inflected forms, e.g. Wenn de Baby weinen, de Mama muss aufpicken 'If/when the baby cries, the mother must pick (it) up'. Pidgins and Creoles also have few prepositions. In Tok Pisin, for instance, two forms, long and bilong, indicate all grammatical relations, as in Mi go long taun 'I went to town'; Em givim tupěla pik long mi 'He gave me two pigs"; and Haus bilong papa bilong mi i stap long hap 'My father's house is over there'. Others such as Bislama have up to five prepositions, but none appear to have a range anything like their superstratc languages. Pidgin and Creole Languages 173 Other grammatical distinctions lacking in pidgins appear to be more essential to the adequate functioning of language and will be reconstituted as part of the process of creolization. Lexical forms will usually be recruited from the superstrate language to perform these functions. There are many similarities in the lexical resources used by Creoles to mark these functions. Thus, for English-based Creoles the indefinite article is usually derived from the numeral one, as in Hawai'i Creole English / got one dog 'I have a dog'; a future marker from a verb meaning 'go', e.g. I go leave om outside for you 'I will leave/it outside for you'; the completive marker from a verb meaning 'finish', as in Tok Pisin mi toktok pinis 'I have finished speaking', etc. Question words are usually composed of two elements, e.g. Guyanese Creole wisaid (< which side) 'where', Cameroons Creole wetin (< what thing) 'what', and Tok Pisin wanem (< what name) 'what'. It has also been claimed that no Creole shows any difference in syntactic structure between questions and statements. If a Creole has special question particles, they are sentence-final and optional. In Guyanese Creole an utterance such as 7 bai di eg dem 'He bought the eggs' is not formally distinguishable as an interrogative (question) or declarative (statement). The difference is marked by intonation. It is in the area of syntax that the boldest claims have been made for the distinctiveness of Creoles. In fact, some time ago, scholars noted in connection with Jamaican Creole that the most striking differences between £>asilectal and acrolectal varieties lay not so much in phonology and vocabulary as in grammar. Although the reason offered by many was that Creole grammar had African origins, the conclusion was that basilectal Jamaican Creole could not be regarded simply as a dialect of English, but was instead a new and different language. Among the grammatical features widely shared by both the Atlantic and Pacific groups are the following: (a) lack of forms of the verb 'to be', or use of so-called 'zero copula', as in: De pikni sik (Jamaican Creole English) 'The child is sick' and Pikinini sik (Tok Pisin) 'The child is sick'. (b) Use of the same word to indicate possession and existence, e.g. get in most English-based pidgins and Creoles, as in: Get wan uman we get gyal pikni (Guyanese Creole English) 'There is a woman who has a daughter' and Get wan wahine shi get wan 174 Pidgin and Creole Languages data (Hawai'i Creole English) 'There is a woman who has a daughter'. Compare also Malaccan Creole Portuguese, where forms of the verb tem 'to have' are used for both possession and existence: irmang machu teng na rua (brother have in street) 'my brother is in the street' and yo teng irmang machu (I have brother) 'I have a brother'. (c) Preverbal negation, as in: Hongri man no de set dan won ples (Kru Pidgin English) 'A hungry man doesn't sit down in one place' and Melabat no kaan go garram yumob (Australian Kriol) 'We can't go with you'. These and other features are believed to be shared by all pidgins and Creoles, whatever their lexical base. Their common existence thus suggests either a much broader genetic relationship or universal principles of development. Nevertheless, there are some Creoles which do not share these features. Compare, for instance, ori wa fiki (Berbice Creole Dutch) 'he was ill', where the particle wa performs some of the functions of a copula, and hernia se wanem ial (Bislama) 'What is that?', where se (probably from French c'est) acts like a copula in some contexts. And some features are clearly due to substratum. A case in point is the inclusive/exclusive distinction in Melanesian Pidgin first person plural and dual pronouns, e.g. yumi 'we [inclusive of speaker and hearer]' versus mipela 'we [exclusive of the hearer]'. Thus, an utterance such as mipela go long taun 'we [exc] are going to town' would exclude the hearer from those going to town. Pidgin grammars tend to be shallow with no syntactic devices for subordination or embedding. They generally use no formal marking to indicate that one part of an utterance is subordinate to another. Distinctive marking of structures such as relative clauses comes later in the stabilization or expansion phase of the pidgin life cycle, or arises in the process of creolization. While most Creoles lack passives, most have relative clauses. Most pidgins and Creoles follow SVO (subject-verb-object) word order. The most-discussed feature of Creole syntax is the system of verbal markers. As early as the nineteenth century scholars commented on markers of tense, mood, and aspect which were shared across Creoles with different lexical. bases. Many have claimed that although these markers are lexically related to the superstrate language, they behave syntactically and semantically like the preverbal tense, mood, and aspect markers in the substrate languages. Pidgin and Creole Languages \ 175 The following similarities can be observed: \ (a) The simple form of the verb without any markers refers to whatever time is in focus. In this example a speaker of Nicaraguan Miskito Coast Creole English is explaining how each jungle spirit guides the animals under his protection to hide them from hunters: Him a di uona. Him tek dem an put dem an dis wie . . . die kom an him liiv dem all hiia an guo de 'He is their owner. He takes them and puts them on the right path . . . they come and he leaves them all in that place and goes off.' All the verbs refer to a permanent state of affairs. In this next one, however, unmarked verb forms refer to the past: Wi liiv from der an kom doun hiirfo stodi 'We left there and came down here so I could study.' (b) Each Creole language tends to have three markers: one to mark anterior tense (simple past for states and past before past for actions), one to mark irrealis mood (future and conditional), and one to mark non-punctual aspect (progressive and habitual). In the case of Hawai'i Creole English, for instance, the markers are bin (anterior), go (irrealis), and ire (non-punctual), as in these examples: A bin go si Toni abaut go spansa da kidz, ae, da baesketbawl tim, da wan ai ste koch f o 'I went to see Tony about sponsoring the kids, eh, the basketball team, the one I am coaching'; and Bambai til tumoro he go teli telifon 'Later, by tomorrow, he'll call'. (c) Where there is more than one particle accompanying a verb, the particles always have a fixed order before the verb: tense-mood-aspect. Althoirgh all combinations are not found in all Creoles, examples off'this pattern can be seen in conservative Creoles such as Saramaccan and Sranan, as in this example from Sranan: me ben sa e go 'I would have been going', and this one from Jamaican Creole: mi en a go sing 'I was going to be singing'. The best examples of the full three-term system of tense-mood-aspect marking tend to be found in the Atlantic rather than the Pacific, although there has been significant development in Tok Pisin, for instance, in recent decades. As recently as the late 1960s and early 1970s there appears to have been a constraint in Tok Pisin against having two particles in preverbal position. This has since been clearly abandoned by some speakers, as is evident in these examples, where combinations of various particles occur: ok, yu ken bai kisim 'OK, you can take it' and Yumi mas bai helpim 'We [inc.] will have to help'. The reasons for these 176 Pidgin and Creole Languages differences between the Atlantic and Pacific Creoles are not yet clear, but they may have to do with time-depth since the Pacific Creoles are all relatively younger. The contribution of decreoliza-tion to the development of tense-mood-aspect systems is also unclear. While most Creoles have preverbal particles rather than inflections, Berbice Dutch is unique among basilectal Creoles of the Caribbean in its use of a mixture of preverbal particles and suffixes in its tense-mood-aspect system. {<£) Another common feature of Creole syntax is serialization of verbs. Serial verb constructions are chains of two or more verbs which have the same subject, e.g. im tek imfut kik me (Jamaican Creole) [he take foot kick me] 'He kicked me'. In Creoles they are used for marking functions such as direction and instrumental, which in other languages may be marked either by inflectional case or prepositions, as in these examples: a waka go a wosu (Sranan) [he walk go to house] 'he walked home' (direction), and a teke nefi koti a meti (Ndjuka) [he take knife cut meat] 'he cut the meat with a knife' (instrumental). This is yet another instance where syntax makes up for losses in other areas. In many cases the verbs involved in serialization can be translated into English with a single verb, and the first verb in the series is often come, go, take, etc. The length of such constructions is apparently not constrained in some languages and thus in Nigerian Pidgin English three or more verbs may be concatenated together across a multiple clause structure as in this example: Dem come take night carry di wife, go give di man [they come take night carry the wife go give the man] 'they came at night, got the wife and gave her to her husband'. It is widely believed that serialization in the Atlantic pidgins and Creoles is to be attributed to substratum from the African languages, particularly the Kwa group. The parallels can be seen in examples such as: akoroma no kyeree akoko no wee (Twi) and di haak kets di tskikin iit it (Jamaican Creole) [the hawk caught the chicken ate] 'the hawk caught the chicken and ate it'. Compare also Koku so äsón wá äxi (Fon) and Jan prán krab ale nan mache (Haitian Creole French) [Koku/John take crab go to market] 'Koku/John brought a crab to market'. Similarly, in the Pacific, it has been suggested that there may be Austronesian substratum influence in the serial constructions found in Tok Pisin. Compare, Pidgin and Creole Languages 177 for instance, boro di-rau-mate-i (Manam) [pig they hit die it] and ol i kilim indai pik (Tok Pisin) [they hit die pig] 'they killed the pig'. However, serialization has a much more limited occurrence in the Pacific Creoles and is arguably of a different character. On the whole, the phonology of Creoles has been less well investigated than their syntax, and within the domain of phonology, there is scant information on suprasegmental phenomena such as tone, stress, and intonation. Many have commented that the suprasegmental phonology of the Atlantic English-based Creoles has been influenced by the tonal systems of the African substrate languages. In Jamaican Creole, tone is lexical in a few minimal pairs. Thus, /at/ with a high level tone means 'hat' or 'hurt', while with a high falling tone means 'heart'. Another contrastive set is /bit/, which with a high level tone means 'biť and with a high falling tone means 'beat' or 'beet'. As in syntax, the relative contribution of universals versus substratum influence needs further investigation in particular cases before more general conclusions can be stated. It is not always possible to trace the origin of a particular Creole feature to a unique source. For example, I noted earlier the tendency towards a simple syllable structure. The preference for open syllables which end in vowels rather than consonants may, however, derive from universal developmental tendencies as well as from substratum influence. Compare, for example, Tok Pisin /giraun/ from English ground, Negerhollands /filis/ from Dutch vleis 'meat', West African Pidgin English /sikin/ from English skin, and Jamaican Creole /takf from English talk and /habi/ from English have. Many scholars have commented on the small size or reduction of pidgin and Creole phonological inventories when compared to both their lexifier and substratum languages. Although it has been sometimes argued that Creoles represent the lowest common denominator of the source languages, this is not always true. Pidgins and Creoles may also have segments which are not found in the superstrate. The Tok Pisin consonantal system used by Usarufa speakers actually has a much greater range of diversity than the Usarufa system in terms of both distribution and phonetic variety. Many speakers of Tok Pisin, for example, particularly those from the Highlands, have prenasalized stops, e.g. ngut 'gut'. In Saramaccan, a Creole English spoken by some 20,000 'Bush 178 Pidgin and Creole Languages Negroes' or Maroons, whose ancestors escaped into the interior part of Surinam from plantations nearer the coast during the late seventeenth and early eighteenth centuries, the double stops /kp/ and /gb/ have been retained from the African substratum languages. The same is true for some varieties of West African Pidgin English. Sounds that are typically absent or infrequent in Creole phonologies are often those that are absent or vary in phonological status in the substrate languages. Both Jamaican Creole and Tok Pisin, for instance, do not regularly contrast /p/ and HI. In Tok Pisin as well as Sranan Irl and III are not distinct. This has been attributed to the fact that in the substratum languages for Jamaican Creole and Sranan and many of the languages of Papua New Guinea there is no systematic contrast between /p/ and HI and Irl and III. In Tok Pisin, for example, Ipl is used in words deriving from English /p/ such as /pik/ 'pig', /planti/ 'many, plenty', etc. Tok Pisin Ipl derived from English HI is pronounced in one of three ways: as English /pf/, as English If I, or as Ipl. This means that some Tok Pisin speakers use Ipl for every use of English Ipl and lil, e.g. /pik/ 'pig',./prut/ 'fruit'. Other speakers use English lil or Ipl for words deriving from English HI, e.g. /pik/, /frut/. Most speakers use HI for words written with /, e.g. /fut/ 'foot'. Nevertheless, at the same time as Creole phonology reflects pressures from the substratum, there are also universal tendencies of first- and second-language acquisition which will tend to eliminate marked (i.e. highly unusual, infrequently occurring) sounds and to reduce the overall number of phonological contrasts. We can note, for instance, the general avoidance of fricatives (sounds like/) and affricates (sounds like the ch in church) in pidgin phonological inventories. From a markedness perspective, fricatives are more marked than stops such as p. Their presence in a language presupposes the existence of stops. This is a good example of how substratum influence and universal pressures may converge on a common solution. However, other sounds commonly not found may be due to superstratum influence. For example, the fact that /h/ is generally absent in Jamaican Creole and Tok Pisin is probably due to the variable omission of initial /h/in non-standard varieties of English to which Creole speakers were exposed. In other respects Creoles have been conservative in preserving phonological patterns found in the superstrate during the time of Creole Pidgin and Creole Languages 179 formation. For example, in Jamaican Creole the pronunciation of /k/ and /g/ as in /kyar/ 'car' and /gwain/ 'going' were customary in some of the varieties of British English transported to Jamaica in the eighteenth century. Pidgins and Creoles do not have a single phonology and phonology remains the least stable system in otherwise stabilized pidgins. As I have shown, phonological variability may be due to substratum and/or superstrate influences as well as universal developmental tendencies, but it is also affected by external factors such as education, sex, age, acquisitional history, etc. This means that variant pronunciations may be found for the same items even within a small group of speakers. Expansion of a pidgin usually involves a steady increase in phonological distinctions. During decreolization, when the superstrate exerts pressure on the creole, the Creole phonological system may merge partially with that of the superstrate language. Thus, the phonology of individual speakers of Tok Pisin varies from heavily Anglicized to what some have called a 'core phonology'. The core phonology is shared by all speakers of the language, while the Anglicized phonology makes the most of the consonant and vowel distinctions in English. In principle, the fluent bilingual speaker of Tok Pisin and English has the whole phonological inventory of English to draw on when speaking Tok Pisin. By virtue of increasing exposure to English, where contrasts such as that between Ipl and HI and hi and IV are significant, speakers are beginning to remodel Tok Pisin forms on their English cognates. This is the phonological dimension of decreolization. New phonological distinctions are being introduced, particularly in the Tok Pisin spoken by the younger generation of speakers in urban areas. There is an interesting phenomenon going on at the same time in rural areas, which is also motivated by exposure to English, but it has the opposite effect. Some young rural speakers now say/fik/instead of/pik/and /fikinini/ instead of /pikinini/. This is the result of a process known as 'hypercorrection', discussed in Chapter 3 in connection with a sociolinguistic pattern in which young lower-middle-class women, in particular, produce more standard forms than the middle class. The Tok Pisin case is somewhat different. These speakers know that where they have forms beginning with Ipl many English speakers have HI, but they do not know English well enough to know how English speakers make use of this contrast between Ipl 180 Pidgin and Creole Languages and/f/. So in attempting to correct their use of /p/ to /f/, they overcorrect and produce forms which are not English at all. This kind of hypercorrection occurs in English too, and probably in most cases where two varieties of differing prestige are in contact. In Brooklyn, for instance, hypercorrect forms such as terlet for 'toilet' are as stereotypical as pronunciations such as 'toity toid street' for 'Thirty-Third Street'. They arise when speakers substitute the sound /er/ wherever they would normally say /oi/. Lexicon As far as the content of the lexical items in pidgin and creole languages is concerned, at least two general points can be made. One is the common existence of a nautical element, which is not surprising given the fact that most pidgins and Creoles tend to be located near a marine expanse. Another is the presence of a common core of items shared across unrelated pidgins and Creoles. Among these are words such as pikinini 'child', 'baby' (< Portuguese pequeňo or pequeniňo 'small'), as found in Jamaican Creole, Tok Pisin, and many others, and save (< Spanish/Portuguese saber/ sabir 'know'), as found again in Tok Pisin and most of the Atlantic Creoles. The latter has since passed into more general English usage, e.g. he has a lot of business savvy. Pidgins and Creoles generally take at least 80 per cent or more of their lexicon from the superstrate language. Consider, for instance, Negerhollands, which has at least 84 per cent of its vocabulary derived from Dutch, 7.5 per cent from English, 2 per cent from Iberian languages, 1.4 per cent from Danish, and 5 per cent from African languages. This estimate is based on 1,300 words recorded in the 1920s. The African element seems to be largely shared with other languages in that region, e.g. bukra 'white man' (< Twi, Efik, Ibo mbakara) (cf. also Jamaican Creole) and funtji 'cornmeal dish'. By comparison, Tok Pisin is estimated to have derived from 11 per cent to 20 per cent of its lexicon from indigenous languages such as Tolai, spoken in New Britain. There are also words from Malay, such as binatang 'insect', and at least 5 per cent from German, though much of the latter is now archaic and has been replaced by English. It is not always possible to trace Creole lexical items to a unique source, e.g. Tok Pisin gaden may be equally Pidgin and Creole Languages 181 from German Garten or English garden. Similarly, bel 'stomach' may be derived from English belly or Tolai bala 'stomach'. The residue of faulty analyses is particularly evident in the lexicons of pidgins and Creoles. Tok Pisin, for example, has incorporated bow and arrow as one lexical item, bunara. (Compare also trausel from tortoiseshell.) These fused forms persist partly through lack of access to correct target models. The number of items in the lexicon of a pidgin is highly restricted. Estimates of size vary from about 300 to 1,500 words, depending on the language. In the case of the younger generation of Tok Pisin speakers, for example, it ranges from about 800 in rural areas to about 2,500 in urban areas, where a greater number of English words have been incorporated. Compare the number ^-^T^ of lexical items which a speaker of an ordinary language like/ English has, i.e. 25,000-30,000, with the number of lexical items in Tok Pisin, i.e. 1,500. However, these 1,500 words can in principle be combined into phrases so as to say anything that can be said in English. The implication is that there is no reduction in the overall semantic domains covered by a pidgin, but merely in the number of items used to map them. Thus, reduction of number of items does not in itself simplify a language, though it entails certain adjustments in lexical structure which may make a language more regular, as illustrated above in my discussion of gender. Lexical items in pidgin languages cover a wider semantic domain than in the base language. Tok Pisin has the term pisin (< English pigeon) for 'bird'. Historically, this represents a case of extension of reference from the English term pigeon, which refers to a specific kind of bird in English, to a term with more general reference. Due. to their small inventory of items pidgins associate different grammatical information with the same semantic and phonological items. This is generally referred to as multifunctionality, i.e. use of the same lexical item in more than one grammatical function. For example, in Tok Pisin askim can function as both noun and verb: Mi gat wanpela askim T have a question', Mi laik askim em 'I want to ask him/her/if. The use of the same lexical item in a number of grammatical functions constitutes a gain in simplicity of the lexical entry but it also violates the principle of 'one form equals one meaning', which has often been cited as characteristic of pidgins and Creoles. The existence of multifunctionality also 182 Pidgin and Creole Languages calls into doubt the extent to which traditional grammatical categories such as noun, verb, etc. are applicable to Creoles. As a consequence of phonological reduction pidgins also have widespread homophony, where one form stands for a number of meanings. Homophony rarely leads to ambiguity since the items concerned are usually quite disparate in meaning, e.g. Tok Pisin sip may be 'sheep', 'ship', or 'jeep', so that context usually supports only one interpretation. It is obvious that there will be gaps in the pidgin lexicon, particularly in the early stages of its development. These may be filled by borrowing or circumlocution. Only at a later stage in its development does the pidgin develop productive internal resources for expanding its lexicon. Circumlocution is a strategy which involves letting the syntax make up for the lack of productive morphological processes which would be used to form words in the lexifier language. Europeans have been eager in the past to cite mythical circumlocutions to demonstrate the inadequacy of pidgin, one of the most notable being that for 'piano' in Melanesian Pidgin, e.g. big fellow bokkes, suppose missis he fight him, he cry too much 'the big box, if the European woman hits it, it cries a lot'. Although it is possible that such a description was used in the first encounter with the piano, it is highly unlikely that it persisted. Nonce circumlocutions such as the following in Bislama for 'burp': oli pulum win afta sakem bakegen 'they breathe air in and then throw it out again', probably occur quite frequently. However, once an innovation has caught on and is used, it will become conventionalized and shortened. Initially these expressions served as descriptions. Later, however, in response to the demands of efficient communication, speakers economized. Once these forms had been used several times they were reduced and stabilized. As the lexicon expands, the clumsy but motivated compounds and periphrastic expressions disappear, e.g. Tok Pisin kot bilong ren 'raincoat' is now kotren or renkot and bel bilong mi hat [stomach of me hot] is now mi belhat 'I am angry'. This is motivated by the desire to give common concepts a reduced expression. It is a direct consequence of their small vocabulary that pidgins and Creoles exhibit a high degree of motivation and transparency in compounding. Many aspects of the lexicon and grammar of Tok Pisin reflect a semantic orientation which is basically non-European, Pidgin and Creole Languages ' 183 Table 6.1. The conceptual field expressed by Tok. Pisin gras, compared with Warapu Tok Pisin English Warapu gras 'hair' pei gras bilong fes 'beard' mausgras 'moustache' gras antap long ai 'eyebrow' gras bilong pisin 'bird's feather' ndru pei gras bilong dog 'dog's fur' naki pei gras nogut 'weed' while other aspects draw on universal principles which determine maximum regularity of encoding of grammatical categories^and lexical distinctions. In some areas of the lexicon we see a con^ spiracy between these factors. As an example, Table 6.1 illustrates the set of concepts which are expressed by terms involving the Tok Pisin word gras (< English grass), which can be used to refer to 'head hair', 'body hair', 'fur', 'feathers', 'moustache', and makes a partial comparison with Warapu, a non-Austronesian language spoken in north-west Papua New Guinea. The fact that meanings such as 'grass', 'beard', 'feather', and 'weed' are all expressed by separate, unrelated words in English is an indication of its greater degree of lexicalization, i.e. to have completely different words for different things. Note too that we would not even consider this to be a coherent conceptual field in English, containing item's which English speakers do not normally see as bearing any similarity to one another. In Tok Pisin, however, there is a kind of iconic relation between these items, which is expressed by the fact that they are all encoded by means of constructions incorporating the word gras. In other words, the terms in Tok Pisin are motivated, whereas in English they are arbitrary. This represents a good illustration of what is called diagrammatic iconicity, i.e. a systematic arrangement of signs, none of which necessarily resembles its referent, but whose relationships to each other mirror the relationships to their referents. Thus, one could say that grass has the same relationship to the ground or earth that feathers have to a bird, a beard to a face, etc. They are all covering on different surfaces. In this particular semantic domain the optimal solution to the problem of encoding 184 Pidgin and Creole Languages is reinforced by the existence of a similar system in the indigenous languages. Many of the languages of Papua New Guinea (both Austronesian and non-Austronesian) have the same lexical item for 'feather' and 'hair'. Often 'fur' and 'leaf are included within the scope of reference of this same term. In Warapu, a single lexical item is used for 'head hair' and 'leaf. The same item in Warapu also means 'body hair', 'fur', and 'feathers', as seen in the table. This is also true of Karam, a New Guinea Highlands language. This distribution of domains is not uncommon in Austronesian languages of New Guinea, but a more normal set for non-Austronesian languages is to have one word for all kinds of hair and a separate word for 'leaf. We can now see how Tok Pisin serves to channel an alien conceptual system into a native conceptual system, while constituting at the same time a simplified means of expression which creates pressure for the grammatical complexities often found in other languages to be eliminated or sharply reduced. Nevertheless, there is an inverse correlation between the lexical expansion of a language and the iconicity of its grammar. Now English terms like feda are increasingly used, especially by the younger generation of urban speakers, who are rapidly expanding the vocabulary of Tok Pisin. In such cases borrowing from English is making the structure of Tok Pisin more irregular and complex. Names for body parts such as ai 'eye' and maus 'mouth' are used as metaphors in both Tok Pisin and Cameroon Pidgin. For example, Tok Pisin has aipas 'blind' (< 'eye' + 'close'). Compare also Cameroon Pidgin lokai and also Tok Pisin mauspas 'silent' and yaupas 'deaf (< 'ear' + 'close'). Tok Pisin also has ai bilong sua 'head of a sore', ai bilong botol 'lid of a bottle', and ai bilong kokonas 'hole of a coconut'. Of course, English has metaphorical extensions of body parts too, such as 'eye of a storm' and 'ear of corn', but the fact that it has many more arbitrary terms like 'blind', 'deaf, and 'dumb' to enclode meanings which are all formally linked in Tok Pisin is an indication of the greater degree of lexicalization of English. Another more general process of grammaticalization is probably at work here too. Body parts provide the most important source domain for spatial concepts in language. Thus, prepositions and postpositions expressing location in many languages are cognate with or derived from body parts. Thus, mouth and eye frequently Pidgin and Creole Languages 185 provide the source for the spatial dimension of 'front'. In Tok Pisin sanap long ai bilong ol means 'to stand up in front of everyone'. Similarly, the word for 'head' in many languages comes to encode the dimension of 'on' or 'front'; 'back', the location of 'behind'; the buttocks or anus, the location of 'under'. In Tok Pisin, for example, we have as bilong diwailflaua 'the base of the tree/flower' and asples 'one's place of origin' as well as as bilong kros 'the reason/cause for anger'. In the latter case we have a metaphorical extension from the concrete domain of body parts to a more abstract concept which does not have an inherent spatial dimension of 'bottomness' or 'under'. In pidgins and Creoles these metaphorical uses are an important means of extending a restricted vocabulary with limited syntactic means. Borrowing from English, however, is threatening to disrupt the unity of a great many conceptual fields which are linked by these metaphors. While the borrowing of lid represents an economy at one level over periphrastic, maus or ai bilong pot 'mouth/eye of the poť, at another level it leads to complexity. The price that has to be paid when a concept is fully lexicalized is that it has to be learned as a totally new item. Pidgins and Creoles in social context Although they are often widely used by the majority of the population, throughout their history most pidgins and Creoles have not had any official status in the countries where they are spoken. In the Pacific, for instance, only Tok Pisin and Bislama have received some official recognition. Tok Pisin is a de facto official language in Papua New Guinea spoken by more than half of the population of 3i million; however, English is the official medium of education. There is also another pidgin language, Hiri Motu ('trade Motu'), based on the indigenous language, Motu, which shares the same de facto official status as Tok Pisin. In practice, all this means is that Hiri Motu and Tok Pisin may be used in the House of Assembly, the country's main legislative body. In fact, most business is conducted in Tok Pisin, which is the most widely shared language among the members. While Bislama is recognized by the constitution of Vanuatu as the national language of the country, paradoxically it is forbidden in the schools. Vanuatu may be the only country in the world 186 Pidgin and Creole Languages which forbids the use of its national language! English and French, the languages of the former colonial powers, are still the official languages of education. Although Australian Kriol does not have any official status, it has been used in bilingual education programs in parts of Australia. While many linguists in both the Caribbean and Pacific have argued for an increase in status and standardization of the pidgins and Creoles spoken in these regions so that they can be used in education, the governments concerned have generally ignored these issues and preferred instead simply to continue the colonial legacy of using the metropolitan European language already in place. French-based Haitian Creole is the only one which appears to have been given serious attention by government planners. The low status of these languages is generally a consequence of the fact that they have not been regarded as full-fledged languages, but as corrupt and bastardized versions of some other language. As recently as 1986 the Times Higher Educational Supplement (17 Jan. 1986) carried a report from a newspaper in Ghana complaining about the use of Pidgin English on Ghanaian campuses and recommending that stern measures be taken against it. The report notes that in no other case do the future leaders of a country talk 'a mixture in which all tenses are thrown to the wind, and words are picked from far and wide, making no sense to the listener'. Most pidgins and Creoles are not written languages, and therefore not standardized. This also fuels popular ideas that they are not 'real' languages. The few such as Tok Pisin which have been given writing systems and have undergone some degree of standardization have been used primarily by missionaries for proselytization. Many of those who tried to develop orthographies for Creoles assumed they were dealing with a version of English or other European language. As a result of these perceived similarities, writing systems based on European languages often did the Creoles a disservice in suggesting that they were inferior and amusing versions of European languages. Often literacy in the Creole was promoted as an explicit bridge to the acquisition of literacy in the related European language. This underlined the importance of using an orthography which emphasized the similarities between the Creole and its lexifier and did not distort etymological connections. For example, in one of Pidgin and Creole Languages 187 the orthographies proposed for the French-based Creole spoken in the Dominican Republic, 'today I am sick' would be written as: zordi mwe malad, while in French this would be: aujourd'hui je suis malade. The first person singular pronoun in the Creole is from the French stressed pronoun moi 'me', and many people object to spelling it in creole in such a way as to obscure its French origin. To get an idea of how widely varied the uses of pidgin and creole languages can be, we can compare two Pacific pidgins/creoles which differ quite dramatically in terms of their functions: Tok Pisin and Hawai'i Creole English. Hawai'i Creole English is the first language of probably the majority of children in Hawai'i, but it has no official recognition. It arose on sugar plantations, where workers of many different nationalities—in particular Japanese, Chinese, and Filipinos—and hence language backgrounds were imported. It is far more decreolized than Tok Pisin, due to the greater and more rapid metropolitanization of the Hawaiian Islands, particularly through the influence of the United States, which annexed the Islands not too long after white planter interests succeeded in overthrowing the Hawaiian monarchy in 1893. Creole was never used as a language of education and the local Department of Education has actively campaigned against it for many years in an effort to eradicate it completely. There is no official orthography for Hawai'i Creole English and therefore no standard. There are, however, some writers who have attempted to use Creole as the medium for poetry, short stories, and drama. AH these genres exi^t in Tok Pisin too. In addition, however, there are Tok Pisin versions of the Old and New Testaments plus a weekly newspaper with comics, advertising, etc. and a wide range of other secular material. The advertisement in Fig. 6.1 is a good example of how Tok Pisin is used creatively in the media. The slogan says literally 'It bone belong me straight', a colloquial expression which means that it is just the thing to serve as the foundation of a good diet. In its literal sense bun means 'bone' or 'skeleton', and one who is bun noting (literally 'bone nothing'), for example, is a skinny person. Tok Pisin can also be heard on the radio and was for many years used as a language of education by missionaries, who still make use of it in some church-supported schools. Although most of the classics of the emerging literature of Papua New Guinea have been written in English, there is some creative writing in Tok Pisin in the form of poems and plays. 188 Pidgin and Creole Languages The following poem in Hawai'i Creole English was composed by Joseph Balaz, a local writer who is the author of Ramrod,-a. literary publication. Balaz's 'Da History of Pigeon' takes its motivation from what he calls the 'phonic' association between the linguistic term 'pidgin' and the bird 'pigeon'. The poem was written for oral presentation at a colloquium on pidgin and Creole languages at the University of Hawai'i at Manoa in Honolulu in 1986. Da History of Pigeon (in phonic association to pidgin) Like different kind words, da world was full of different kind birds: yellow birds, blue birds, red birds, love birds—and den came da pigeon. Da history of da word pigeon is li' dis—Wen da French-speaking Normans wen conquer England in da year ten-six-six, dey wen bring along wit dem da word pigeon, for da type of bird it was. Da resident Anglo-Saxons used da word dove, or D-U-F-E, as dey used to spell 'um, to mean da same bird. It just so happened dat terms in Norman-French wen blend wit Old English sentence structure, to form what we know as Middle English. In da process, da French word became da one dat referred to da pigeon as food. Today in England, if you look for dem, you can find recipes for pigeon pie. Food for taught, eh—Even back den, da word pigeon wen blend with pigeon for get some moa pigeon. So now days get pigeon by da zoo—get pigeon on da beach—get pigeon in town—get pigeon in coups—and no madda wat anybody try do, dey cannot get rid of pigeon—I guess wit such a wide blue sky, everyting deserves to fly. A number of the phonological features specific to Hawai'i Creole English have been represented in Balaz's spelling; for example, da 'the',-dis 'this', den 'then', dey 'they', dem 'them', dat 'that', wit 'with', taught 'thought', everyting 'everything' illustrate the tendency for the English interdental fricatives to become stops in the creole. Spellings such as moa and madda indicate the absence of post-vocalic hi in creole. Hawai'i Creole English does not mark case, gender, or number, so '«m (probably from a phonetically reduced form of him or them) can mean 'him/her/ them/iť. There are also some syntactic features specific to the creole, such as the use of wen (from English went) to mark the simple past, as in wen conquer. The construction for get moa pigeon is also characteristic: other English-based Creoles use for Pidgin and Creole Languages 189 where English would use to or in order to. Another feature is the use of get in existential constructions, e.g. get pigeon by da zoo 'there are pigeons at the zoo'. This phrase also illustrates the lack of plural marking. This means that the plural form of pigeons and the term for the language, pidgin, are homophonous. This allows Balaz to make the humorous point that because both pidgins and pigeons are everywhere, they are hard to get rid of. Elsewhere writers have faced similar problems to those of Balaz, e.g. in Sierra Leone. In 1964 Thomas Decker translated Shakespeare's Julius Caesar into Krio in order to create propaganda for the serious use of the language. The performance of the play by the National Theatre League heralded the birth of drama in Krio. Subsequent plays took the development of Krio as a dramatic medium even further since they were not translations, but original works with local themes. The lack of an agreed orthography (writing system), however, still hampers both playwrights and actors. Annotated bibliography There are a number of good recent surveys of the field, e.g. John Holm's two-volume Pidgin and Creole Languages (Cambridge: Cambridge University Press, 1989), the first volume of which is an overview of theory and structure while the second includes sample texts and short descriptions of these languages. A general bibliography, now somewhat outdated, can be found in John Reinecke, David DeCamp, Ian Hancock, and Richard E. Wood's A Bibliography of Pidgin and Creole Languages (Honolulu: University fbf Hawaii Press, 1975). Other surveys include Peter Mühlhäusler's Pidgin and Creole Linguistics (Oxford: Blackwell, 1986) and Suzanne Rqmajne's Pidgin and Creole Languages (London: Longman, 1988). These latter two focus more on the Pacific, while Holm's volumes treat the Atlantic. There are also some useful collections of articles, such as Dell Hymes's Pidginization and Creolization of Languages (Cambridge: Cambridge University Press, 1971), Albert Valdman's Pidgin and Creole Language (Bloomington: Indiana University Press, 1977), Kenneth C. Hill's The Genesis of Language (Ann Arbor: Karoma, 1979), Albert Valdman and Arthur Highfield's Theoretical Orientations in Creole Studies (New York: Academic Press, 1980), Arthur Highfield and Albert Valdman's Historicity and Variation in Creole Studies (Ann Arbor: Karoma, 1981), Ellen Woolford and William Washabaugh's The Social Context of Creolization (Ann Arbor: Karoma, 1983), and Pieter Muysken and Norval Smith's Substrata vs. Uniyersals in Creole Genesis (Amsterdam: John Benjamins, 1986). 190 Pidgin and Creole Languages The bioprogram hypothesis was first put forward in detail in Derek Bickerton's Roots of Language (Ann Arbor: Karoma, 1981) and subsequently modified. A good overview of the issues can be found in Romaine's Pidgin and Creole Languages, ch. 7, and Bickerton's article 'Creole Languages and the Bioprogram', in Frederick Newmeyer (ed.), Linguistics: The Cambridge Survey, 1. Linguistic Theory: Extensions and Implications (Cambridge: Cambridge University Press, 1988), 268-84. The example from Pidgin Yimas is taken from William Foley's article 'Language Birth: The Processes of Pidginization and Creolization', in Frederick Newmeyer (ed.), Linguistics: The Cambridge Survey, 4. Language: The Socio-Cultural Context (Cambridge: Cambridge University Press, 1988), 162-83. The story of the mutiny on the Bounty can be found in James Norman Hall arid Charles Nordhoff s trilogy Mutiny on the Bounty. The Guyanese Creole example is taken from William O'Donnell and Loreto Todd, Variety in Contemporary English (London: George Allen and Unwin, 1980). Further information on the Guyanese situation is contained in John Rickford's book Dimensions of a Creole Continuum (Stanford: Stanford University Press, 1986). A discussion of the problems surrounding the use of Hawai'i Creole English can be found in Charlene Sato's article 'Linguistic Inequality in Hawaii: The Post Creole Dilemma', in Joan Manes and Nessa Wolfson (eds.), Language and Inequality (Berlin: Mouton, 1985), while further information on the social status and uses of Tok Pisin can be found in Suzanne Romaine's articles 'The Status of Tok Pisin in Papua New Guinea: The Colonial Predicament', in Ulrich Ammon and Marlis Hellinger (eds.), Status Change of Languages (Berlin: Mouton de Gruyter, 1991), and 'Pidgin English Advertising', in Christopher Ricks and Leonard Michaels (eds.), The State of the Language (Berkeley: University of California Press, 1990), 189-203. For information on socio-political issues in the Caribbean, Hubert Devonish's Language and Liberation: Creole Language Politics in the Caribbean (London: Karia Press, 1986) is useful. Sarah G. Thomason and Terence Kaufmann's book Language Contact, Creolization and Genetic Linguistics (Berkeley." University of California Press, 1988) attempts to inform the methodology of traditional historical linguistics with findings from studies of language contact. — 7 — Linguistic Problems as Societal Problems A senior professor of education visited a London comprehensive school and discussed with one class the languages they spoke at home. One boy put up his hand and said that his family spoke a French Creole. In an unguarded moment the professor replied, 'That's nice.' 'What's nice about it?', asked the boy. Sociolinguistic research, in particular on social dialects and minority languages, has had many practical implications since it is concerned with fundamental inequalities in language use. There are many areas of public life where language has an impact, such as the medical and legal professions, but particularly in the school. Sociolinguists have been actively engaged in studying the problems which arise from language use in these contexts, and especially what happens when there is a mismatch or difference in language involved between the participants, such as doctor and patient, lawyer and client, judge and jury, etc. In this chapter I will focus on some of the types of problems arising in school which are language-related. J; Language and educational failure Language has often been cited as the main cause for the greater rate of school failure among minority children. As one of society's main socializing instruments, the school plays a powerful role in exerting control over its pupils. It endorses mainstream, and largely middle-class, values and language. Children who do not come to school with the kind of cultural and linguistic background supported in the schools are likely to experience conflict. This is true even of working-class children belonging to the dominant culture, but even more so for children of ethnic minority background. In Britain, for example, there is a hierarchy of educational success or failure. Indigenous middle-class children do best, while