Jamaican Creole morphology and syntax Peter L. Patrick 1 Introduction 1 1. 1 History 1 1.2 The Creole continuum 2 1.3 The Jamaican mesolect 4 1.4 The data and orthography 6 2 Tense, mood and aspect marking 6 2.1 A Creole TMA system? 6 2.2 Habitual, progressive and completive aspect 7 2.3 Anterior tense 8 3 Verb forms 10 3.1 Verb inflection 10 3.2 Person and number agreement 11 3.3 (Modal) auxiliaries and past participles 12 4 Negation 14 4.1 Sentential negation 14 4.2 Negative tags and negative imperatives 14 4.3 Negative concord and other negative forms 15 5 Word order, focus and copular structures 16 5.1 Word order 16 5.2 The copula: Functions and significance 16 5.3 The copula in progressive forms 17 5.4 The copula in equative forms 17 5.5 Focus structures: Predicate clefting 17 5.6 Focus structures: Other types of clefting 18 5.7 The copula with adjectives and locatives 19 6 Complementation and subordination 20 6.1 Nonfinite clauses 20 6.2 Finite clauses 21 6.3 Subordinating conjunctions 21 7 Serial verb constructions 22 8 Relativization 23 9 Pronouns 25 9.1 Personal pronouns 26 9.2 Possessive pronouns 27 9.3 Interrogative pronouns 28 9.4 Indefinite, reflexive and reciprocal pronouns 29 9.5 Demonstratives 29 10 Noun phrase structure 30 10.1 Possession 30 10.2 Noun classification 31 10.3 Articles 32 10.4 Number marking 34 10.5 Associative plurals and other phenomena 37 11 Conclusion 38 Peter L. Patrick University of Essex Dept. of Language and Linguistics Wivenhoe Park, Colchester CO4 3SQ United Kingdom [email protected] +44 (0) 1206 872 088 To appear 2004 in A Handbook of Varieties of English. Vol 2: Morphology and Syntax, ed. Bernd Kortmann, Edgar W Schneider, Clive Upton, Rajend Mesthrie & Kate Burridge. (Topics in English Linguistics, ed. Bernd Kortmann & Elizabeth Closs Traugott.) Berlin, New York: Mouton de Gruyter. Jamaican Creole morphology and syntax Peter L. Patrick, last rev. November 2003 1 Introduction Jamaican Creole (JamC, known to its speakers as “Patwa”) is a language of ethnic identification for roughly two and a half million people in the island of Jamaica -- and overseas for many thousands of native speakers (and non- natives; see British Creole chapters.) JamC is a canonical example of an Atlantic Creole. One of the first Caribbean English-lexicon Creoles to be described using modern linguistic methods (Loftman 1953, Cassidy 1961), it remains among the best-researched. The first generative grammar of a Creole was Loftman Bailey’s Jamaican Creole Syntax (1966). The first comprehensive etymological dictionary of a Creole was Cassidy & LePage’s Dictionary of Jamaican English (1967, hereafter DJE). 1. 1 History JamC owes little or nothing to either the indigenous Arawaks or Spanish invaders, starting with Columbus in 1494, who settled the island in 1509, bringing the first African slaves. By 1601 only a handful of Arawaks remained alive alongside 1,000 Africans. When the British arrived in 1655 with 9,000 troops, they met 6,000 inhabitants, 1,500 of African descent and the rest mostly Spanish; after 1660, a few dozen Spanish remained, while 300 Maroons fought from the mountains. The Maroons today, custodians of African culture, still preserve a distinctive speech form, Maroon Spirit Language (Bilby 1983). Their ranks were supplemented by runaways under slavery, and they maintained their independence by treaty, defeating the British in 1739 and 1795. However the origins of JamC postdate 1660, in the interaction of British colonists and African slaves. The language did not yet exist in 1658, when the 7,000 settlers and soldiers in the island from Britain, Ireland and the Americas outnumbered Africans 5 to 1; but between 1677, when there were about 9,000 each of whites and blacks, and 1703, when the white population had slightly declined but the numbers of enslaved Africans had risen to 45,000, the roots of JamC were planted. Many key features were in place before 1750, though others can only be documented from the early and mid-19th century (Lalla & D’Costa 1990). Jamaican language and its place in society reflects the brutal history of Jamaica as a British sugar colony until Independence in 1962. Creolization in the 1 Jamaican Creole grammar broadest sense led to emergence of new cultural and social institutions, including language, but the subordination of JamC to English –- the native tongue of a tiny minority -– has persisted to the present day, with consequences for education, economy, and psychological independence. The collapse of the plantation economy between the two world wars brought on mass urbanization, making Kingston the largest “English-speaking” city in the Americas south of Miami (Patrick 1999; Shields-Brodber 1997). Yet only in the 21st century has the Jamaican government seriously begun to explore language planning and recognition of JamC as a national language. Jamaican Creole’s dramatic genesis in British slavery, imperialism and the African diaspora to the Americas has focused creolist research on language contact, especially the influence of African languages (Akan and Kwa families, along with Bantu), and to a lesser extent British English dialects (West of England, Irish and Scots), as well as universals of language acquisition and creation. Over 90% of Jamaica’s population are of African origin. Other groups claim Indian, Chinese, Syrian and European heritage; of these, only Europeans were present before 1845 and contributed to the formation of JamC. For all these Jamaicans, JamC is a shared marker of ethnic and national identity which serves to distinguish them from other peoples, and to unite them in possession of a rich, diverse set of discursive resources. 1.2 The Creole continuum Social stratification in Jamaica is crucial to understanding the extreme variability of contemporary Jamaican speech. The complex linguistic situation may be related to an equally intricate web of social relations, using the model of the creole continuum. This is opposed to discrete multilingual or multidialectal descriptions such as community bilingualism, standard-plus-dialects, and diglossia. The inapplicability of classic diglossia to Jamaica (Ferguson 1959, 1991) motivated DeCamp to invent the (post-)creole continuum model: “There is no sharp cleavage between Creole and standard... [but] a linguistic continuum, a continuous spectrum of speech varieties, ranging from... ‘broken language’... to the educated standard” (DeCamp 1971: 350), i.e. from basilect to acrolect. JamC is natively available to nearly all Jamaicans, but Standard Jamaican English (StJamE), the acrolect, is not -– it is a home language for a small minority, and learned as a second language of school, literacy, mass media and work by others. This is the direct result of the colonial distribution of power in earlier centuries, which worked to create and maximize the norms that still 2 Peter L Patrick, July 2003 devalue JamC and elevate StJamE. Many Jamaicans, and even many linguists (Creole-speaking and other), still maintain this contrast in prestige as a base component of their attitudes towards Jamaican language, and it surfaces in many linguistic descriptions. In truth, both poles of the continuum are idealized abstractions, a collection of features most like standard Englishes (the acrolect) or most distant from them (basilect). Yet between these poles lies the continuum of everyday speech: a series of minimally differentiated grammars with extensive variation – - an apparently seamless web connecting two idealized varieties, which arose in the same place and time-frame and share distinctive features, yet cannot be genetically related. The descriptive problem is thus to reconcile genetic descent and nongenetic, contact-induced language change within a finely-graded continuum. While StJamE is recognized as an English dialect, descended by normal transmission from 17th and 18th-century British input dialects, creolists agree that the grammar of basilectal JamC differs radically from native English dialects, due to extensive language contact resulting in structural mixing. There is less agreement on whether this process took the form of abrupt creolization, whether a pidgin developed in the island first, or whether a prior pidgin existed –- e.g. on the African coast -– and was relexified (Cassidy 1971; Alleyne 1980; Singler 1984; Hancock 1986; Lalla & D’Costa 1990). The prevailing opinion is that this sharp contrast makes it impossible to relate JamC genetically to English –- or indeed to its African input languages, with which there is also a radical structural break (Thomason & Kaufman 1988, Thomason 2001) -– though it bears obvious historical links to both. 1.3 The Jamaican mesolect As linguists since Bailey have preferred to focus on these extremes, most research concentrated on basilectal JamC, until the recent emergence of studies on StJamE (Christie 1982; Shields 1989, Shields-Brodber 1997; Mair 2003, Mair & Sand 1998, Sand 1999). (Patrick 1999 is the only study of the mesolect.) Yet in purely social and demographic terms, the most important variety in Jamaica is the intermediate one known as the mesolect; its broad limits include the speech uttered by most Jamaicans, in most situations. Although empirical data for language description of JamC are nearly always drawn from points within the continuum (i.e. the mesolect), it remains undertheorized and underdescribed. This may be because most linguistic treatments of JamC adopt a categorical perspective (Chambers 1995), seeking to explain away inherent 3 Jamaican Creole grammar linguistic variation by attributing it to the random mixing of so-called ‘invariant grammars’, viz., the basilect and acrolect. Thus Bailey (1971: 342) tried to model mesolectal speech as “standard with incursions from the creole, or creole with incursions from the standard” through “borrowing and interference”, while Akers (1981: 4) believed it was due to a failure of acquisition by speakers who “incompletely control their code”. Both views portray Jamaicans as less than competent in their everyday language, and the mesolect as grammar-less. Such an approach fails to reach descriptive adequacy. The mesolect cannot be reduced to interference between two discrete, polar systems, and no such detailed description has ever been attempted. The existence of language ideologies and attitudes (resembling those commonly found in bilingual communities) which do not explicitly grant the mesolect autonomy, should not mislead as to its systematic internal organization (Beckford Wassink 1999; Muhleisen 2002). Although highly variable, it comprises a grammar describable via both qualitative linguistic generalizations and quantitative constraints, which has evolved over three centuries, arriving at a set of socially-evaluated patterns with their own historical and cultural ecology. Its post-creolization development is broadly similar to that of other, non-creole speech communities, to which variationist theory and descriptive methods have been profitably applied (Chambers, Trudgill and Schilling-Estes 2002). Earlier speculations that the creole continuum might be so variable as not to constitute a speech community at all proved unfounded (Patrick 2002). In the most detailed account of the mesolect, Patrick (1996, 1999) concludes that it is characterized by the systematic presence and integration of English forms and rules in a partial and variable, but non-random, manner. On this view, mesolectal grammar does not result from improvised mixing or code- switching between two polar varieties, nor are its speakers fossilized learners. Rather, the mesolect is an organized, distinctive collection of elements with a long history and its own complex norms, structures and social patterning. Many choices and variants are possible within it, but many are not. Ways of speaking are not accidental but conventionalized; borrowing occurs, but is not the sole source of variation; grammatical rules exist and interlock; and it is transmitted through normal language acquisition. Though change occurs, the mesolect contrasts with newer and less stable varieties such as BrC. Despite the defining presence of English elements, which mark it off clearly from the basilect, the mesolect shares with the latter many constraints, structures and organizing principles which are not generally characteristic of native dialects of English. Insofar as creoles are defined through such contrasts 4 Peter L Patrick, July 2003 (Mühleisen 2002), the mesolect is thus Jamaican Creole, and not Jamaican English (i.e. it cannot be genetically related to English). Indeed, it probably appeared earlier than the basilect (Alleyne 1971). English-like surface forms (some exclusive to the mesolect, e.g. did, others shared with the acrolect, e.g. neva, or even the basilect, e.g. ben –- all three tense-markers are discussed below) characteristically alternate with zero, governed by constraints shared with basilectal JamC but not with native Englishes. This pattern is found in both earlier Jamaican texts and contemporary speech. The mesolect is naturally the primary object of description here, with frequent reference also to basilectal structures. Though there is a clear dividing line between these two grammars (Patrick 1999), there is none between mesolect and acrolect, since the partial presence of English forms and constraints merges indistinguishably into the possession of full competence in StJamE. While the many structures shared with the basilect provide a firm linguistic basis for treating the mesolect as JamC, there is no such structural warrant for restricting “English” only to the high acrolect –- it is strictly the power of social convention which influences speakers, and therefore linguists, to do so. In practice, this lack of a sharp upper boundary creates difficulties in analysing some speakers or texts. The search for a single point, a linguistic and social division, where StJamE starts and JamC ends, is the misguided product of colonial language ideologies. Below, however, illustrative contrasts are drawn. This coincides with the symbolic value speakers attach to fine, or even illusory, distinctions between “proper English” and “Patwa” (a term broad enough to encompass, at times, everything but the high acrolect). 1.4 The data and orthography Much data below is cited from written records. Cassidy’s phonemic orthography (1961) has served as a model for many other Creole writing systems, but is little-followed by Jamaican writers. Uncredited data (and most translations) are by the author or recorded informants, and generally follow Cassidy. While creolists generally prefer a diachronic perspective, and seek out “pure” basilectal forms as evidence of earlier stages of language development, the description below is synchronic and does not privilege the basilect. This may affect some analyses, e.g. whether to treat se ‘say’ under complementation or verb serialization. 5 Jamaican Creole grammar 2 Tense, mood and aspect marking 2.1 A Creole TMA system? All descriptions of basilectal JamC agree that it combines invariant pre- verbal particles with unmarked verb stems to express these grammatical categories, where native Englishes typically use verbal auxiliaries, inflectional suffixes and agreement-marking. It is also generally argued that contrasting linguistic categories and semantic values underlie and constrain these formal differences. The most influential account is given by Bickerton (1975, 1981) for creoles in general. Three main categories –- anterior tense, irrealis mood, and non-punctual aspect –- each have a principal preverbal marker, which must combine in the order T-M-A. In creoles, Bickerton argued, states, habitual situations and progressive events can all be described as having nonpunctual aspect. Further, verb stativity is said to crucially affect the occurrence and interpretation of markers of past-reference: bare nonstative verbs receive a default past-reference reading, while statives are nonpast unless preceded by a tense-marker. These claimed syntactic and semantic properties together describe a grammar that “clearly bears no relation to the system of English” (Bickerton 1975:47). This gives the following paradigm: Stativity Pre-V Marker Meaning Examples (1) +stative none present, habitual Mi Ø lov im (2) -stative none past Mi Ø run (3) +stative (b)en/did past Mi ben lov im (4) -stative (b)en/did past-before-past Mi ben ron with the translations: (1) ‘I love her’ (now) / ‘I love her’ (habitually) (2) ‘I ran’ (3) ‘I loved her’ (4) ‘I had run’ (before some other past event or action) Bickerton argued that creole basilects, including JamC, do not have an absolute past tense, but rather a relative anterior tense. Instead of taking the moment of speaking as an absolute reference point (with past tense required for events before it, and future for events after), this point is relative. For stative verbs it is the moment of speaking, but for verbs of action it is some relevant 6 Peter L Patrick, July 2003 earlier moment. Thus when they are preceded by a past marker (ben in 5), they refer to a past-before-past action, sometimes called remote past. (5) Father Manley fight and mek black pickney go a St Hilda’s school, where no black pickney couldn’t ben go first time. (Sistren 1987: 105) ‘Manley fought so that black children could go to St Hilda’s school, where no black children had been able to go in the old days.’ While Bickerton’s description often matches JamC utterances at surface level, the analysis is flawed. It is widely conceded that this scheme fails to account for the full range of facts over many creoles (Singler 1990), and articulates poorly with general TMA and typological studies (Winford 2000). However it is rarely noted that, as a categorical analysis assuming privative oppositions, it misconceives the nature of creole grammars, including JamC. That is, it predicts a strict form-meaning isomorphy which does not hold: e.g., in order to convey a past-before-past meaning, a nonstative verb must be marked with an anterior marker (basilectal ben and variants wen, en, min; mesolectal did); and when so marked, it must receive such a reading. In reality, exceptions occur in both directions. The prediction is worth refuting because many other linguists give such idealized accounts of creole grammars. 2.2 Habitual, progressive and completive aspect Progressive aspect is uniformly signalled by preverbal a (6-7), while habitual aspect is often unmarked (1), though at an earlier stage both were marked alike in a single imperfective category with (d)a (da and de persist in western Jamaica, Bailey 1966: 138). It is still possible to mark habitual with a+Verb, just like the progressive. Aspectual a is tense-neutral in JamC, and may be preceded by tense-markers (ben+a, did+a, ben+de, was+a etc.). (6) -stative a, de progressive Mi a ron (7) -stative ben/did + a/de past progressive Mi ben a ron (6) ‘I’m running’ / ‘I was running’ / ‘I (used to) habitually run’ (7) ‘I was running’ / ‘I used to habitually run’ Completive aspect is signalled by don, which unlike other TMA markers may occur not only preverbally but after the verb phrase (8-9), or even both. 7