sign in sign up
<<
Home , Creole language, Middle French, Pidgin

r The study of and creole lenguages

PiererMuysken and Norval Smith

r.r Introduction This book is concernedwith pidgin and creolelanguages. This statementmight well give the impressionthat we know preciselywhat is meant by theseterms. In fact they arethe subjectofmuch debate.Creolists agree neither about the precisedefinition ofthe terms pidgin and creole,nor abour the statuso[a number oflanguagesthat havebeen claimed to be pidginsor creoles,Mixed ,introduced in chapter4, havegenerally not been mentionedat all. To turn first to pidgin languages,it is generallyagreed that in essencethese represent speech-formswhichdo nothavenative speakers, and are therefore primarily usedas a means of communication among pcoplewho do not sharea common .The degreeof developmentand sophisticationattained by sucha pidgin dependson the qpe and intensiry of communicativeinteraction among the i$ users.Miihlhiusler (r986) makesthree basic distinctionsamongst speech-forms that crcolistshave referred to aspidgins - (ratherun- stable)jargons, stablepidgins, and expandedpidgins (seefurther chapter3). To turn to creolelalguages (or just creoles),one vital differencefrom is that pidginsdo not havenativespeakers, while creolesdo. This is not alwaysan easydistinction to make,as one aspectof theworldwideincrease in linguisticconformiry and the concomi- tanrreducdon in linguisticdiversity, is that extendedpidgins are beginning to acquirenative speakers.This has happenedfor instancewith , English, and Sango(),to name but rhreecases. In panicular this hastended to occur in urban environments,where speakers from different ethnic groupshave daily contactwitheach other. The pidgin thenbecomes the town language.The childrenof mixcd marriagesfrequently grow up speakingthe home language- the pidgin - astheir native language.

r.z Historical and the definition ofa creole I A creolelanguage can be defned asa languagethat hascome into existenccat a point in time that can be establishedfairly precisely.Non-creole languages are assumed (often in thcabsenceofdetailed knowledge oftheir precisedevelopment) to haveemergedgradually. i

I [: Pitter Mulsken and Norua! Smtth + Thestu4 of? lgin anl creob langages

from other spokenlanguages. Malry of rhemt€nd to hav€certain features in common, bur So Archaic developedinto ClassicalLatin, the popular variery of which rn turn creolistsare divided asto the interpretationofthis fact,and a languagelike Chineseresem- developedinto Vulgar Latin, which among other thingsdevelopedinto 'which some blesmanycreole languages in its .This meansthat beforewe can claim alanguage developedinto Middle French,which in turn developedinto Modern French \Vhile others'we to be a creole,we needto know somethingabout its history,either linguiscicor social,and st^g., ofthi, d*.lop-ent involved more radicalchangcs in tfrelanguage thar preferablyboth, As we know comparativelylitde about chedetailed development ofmost caricleim r.rith somejustification to be ableto tracethe line ofdevelopmentfrom Modern litde languagesin the world, and virtually nothing of the history of most ethnic groups,this Frenchback toArchaic Latin - the earliestrecorded stage ofLatin, with on thewhole it is fairly inevitablymeansthat theremaybe many untecognizedcreol€ languages around theworld. difficulty.Before thatwe haveto rely on linguistic reconstruction,but onceagain to One problem in the identificadonofparticular languagesas creoies is causedbythenot obviousthat Latin is a typical Indo-Europeanlanguage, and can thus be safelyassumed of possibly unusuaicircumstance that creolestend to be spokenin the samegeographical regions as havedeveloped from Proco-lndo-European,through the intermediatestages have been &e languag€sthat provide the greaterportion of th€ir lexica (their donor languages,ot Proto-ltalo'Celtic and cerrainly Proto-ltalic Proto-lndo-Europeanitself may around B C' languages).In somecases we 6nd a continuum ofsPeech_formsvarying from the spokensom€wherein SouthernRussia (an ? nachro nistic term ' ofcourse) 5ooo languages' creol€at one end ofihe spectrum(the basilect),through intermediateforms (mesolectal This kind ofstatementwecan de6nitely not makewhentalking about creole Sowe cannot varietied, ro the lexifier languagefthe acrolect).Sometimes speech-forms exist which Theseexhibir an abruPt breakin the courseoftheir historicaldevelopment r8) derivesin any apparently representcases where ei*rer the originalmesolect has survived, while the basilectai saythat Sranan(the major EnglishJexifiercreole ofsurinami seechapt€r historicalprecur- creoie,and sometimes also ihe originallexifierlanguage have not. Suchcases may be referred eradualfashion from EarlyModern English irs mostobvious immediate ofSranan ro aspost-creoles. Oth€rcases seem ratherto involvepartialcreolization, or inlluencefrom io.. Er,.., .omparison ofEarly Modern Eng)ishwith the earliestforms " ",rrrory two form ofthe samelanguage. These languages maybe t€rmed semi-creol€s orcrco- (6rst recordedin r7r8) will make it abundantly obvious that we ere dealing with acreolized Modern English loids. seems likely to havebeen the resultofsome such While Iinguisa completelydifferent formsofspeech. There is no conceivableway that Early Process. Even would not in generalwish to recognizethis languageas being a fuli creolc,many aspects .oul.1 h"u. d.u.lop"d into the very different Srananin the available7o or so years wholesale ofAfrikaansare reminiscent ofthe things rhat happenduring .Other cas€sof the phonologicalJevelopments required would be extreme,not to speakofthe putativecreoloids ar€ American Black English, arrd at leastsome forms ofBrazilian P"rtu- chaneesthat would havehad to havetaken placein the syntax that they guese. S"ocreole langu.ges a.e different from ordinary languagesin that we can say A quite different situation involving an 'intermedia[e'staaus is the caseofthe mixed cameiDto existenceat somepoinc in time. Applying the techniquesofhistoricai linguistics que\tion of languages.This type which has until now been the object of comparativelylittle study, to creolesis therefor€not simple, and in addition presupposesanswering the provided the involvescases where two languagesclearly make a signiFcantcontribution to language- which languagesthe creoleshould be comPatedwith: the languagcwhich structure frequentlyone language pfovides the contentwords,and anotherthe grammar.Hete thete l.*i.orr, o. th.langu"ge(s)whichwere responsiblefor mostaspects ofgrammatical isnot necessarilyanyquesdorrofsimpli6cation. Awell-known caseofthis ro bestodied - inasmuchas it is possibleto identi$ these' rype (and' Media Lengua (lit. 'middle language')(Muysken r98rb) is spoken in Ecuador,and It is cleerin fact that creolclanguages develoP as thc resuirof'linguisric violencc' - to reckon with a involvesSpanish lexical items, combinedwith basically morphology,and aswe shali see,frequently socialviolence too) ln other words' we have Quechuasynta-x, (1992)has referred to this kind ofsituation aslanguage intemvining develoPmentofthe languaBe,the natural transmissionof a language phonology.Balker breakin the natural 'We not English referthe readerto chapter4. This whole subjectha-s just startedto be studied in any from generationto The Parentsofthe 6rst speakersofsrananwere important' they detail.Sometimes a creoieinvolves substantial mixture at all levelsoflanguage structure. ,p."klr. rtl, btt.p.akers ofvarious Afrrcan languages,andwhat is more "t the A casein point is BerbiceDutch Creole,described in chapter19. did not gro* .,p in .n environment where English was rhe norm How creolization' various theoriesare Othercaseswherelanguages have become simplified to someextentare oflingua ftalcas d.u.lop-.rrr oi language,takes place,or at leastwhat the "r.ol. " at this juncrurc-this is acontrover- (not rheLingua Francaofthe Mediterranean)and koines.These come into existenceunder concerninghowit tekesPlace, wecannot reallygo into below' similar circumstances one speech-formbecomes widely usedby non-nativ€ sPeakers, sial matteithat will be dealtrvith in chapters8 thtough rr' and briefly a degreeof simpliGcation.Here, the processseems to be gradual- in other What is clearisthat creolelanguages arc not in the slightestqualitatively distinguishable undergoing 6 The sndl ofpidgin and crcoh hngtzgzs Pieto Muyshenanl Norval Snith words, no linguisticor socialviolcncc is involved.Vc speakofa linguafranca when speakers pcan colonial languagcs,afthough even herewe may be certain thar some languagesremarn of various different languagesare involved, and of a koine when the didec$ of a single undiscovered.In the ese ofcreoles and pidgins not involving a European baselinguists have language are involved. bccn faccdwith the above-mentionedproblem that the history ofvcry many langtagesis In chapterz6 there is an annotatedlist oflanguageswherc thescdistinctions and some verypoorly known. And aswewill discovertime and dmeagain in thc courseofrhis book, further ones are used to classifrover too languagcsand .To comPlicatemaller a knowledgeofthe historyofa languageis ofrcn esseniia.lfordctermining its crcoleor pidgin speechforms maychangeinstatus ovcr time. Variousscenarios or life-c1'ctes(cf. HaJlI966' starus,or the lack ofthis. This meansthat creolesthet cameinto cristenc€hundrcds ofyea$ who usedthe termsomewhatdifferently) have been proposed for thc developmentofcreoles. agomay only bc recognizedas such in modern limes. Miihlheusler(Iq86l presents thr€e such sc€narios: The smallsize ofmany creole-speakingcommunities also militatcs againsr rheir recogni- tion. A small linguistic community will more casilybe assumedto rcpresenta (deviant) Typ" t "ryp,, Typ"t dialectofa largerlanguage than a largeone will. Small commr.rnidesalso ger overlooked jargon jargon Jargon more eesily.So the Vutun ''of Qnghai provincc, China has been recognizedas ll involving a problem in classiicationby Chinescscholars for quite somctim€. This mix€d stabilizcdpidgin stabilizedpidgin Amdo Tibetan-KansuMongol- has certainly bcen in existencefor several I *p""al oiari' hundrcd years.It had beenvariously claimed to bc Chincse,Monguot and Tibetan. lts I esscntialrymixed staruswas 6rst recognizedby Chen (1982).The facr,however, that the .Ju languagchas only zSoospeakers in 6ve villageshas not helpcd it ro appcarin any lisr or Straits New Guinea Hawaiian Creole Torrcs classificationofthe world's languages.For instance,it doesnot appcarin the rrth edition English Tok Pisin English Crcole of (Grimes r988). We havecited the questionofprejudice above. This is especiallyrelevant in the caseof As will bearguedinchapterJ, however,not all jargonsor pidginsare Partofsucha lif€-cycle, pidgins.Pidgins, by their very nature,tend towardsinstabiliry both in termsoflinguistic and neither can we show that all cr€oleshad a jargon or pidgin stage.It is in this rcspect system,and in termsoftheir function. Iftheydo not belongto thc smallgroup ofpidgins that mixed languagesdisplayan important differencefrom creoles.On theone hand,mixed that bccomestandardized, or nativized,or borh, rheymay well disappearcompletely when larrguagesdid come into existcnccat a particular moment in time, on the other harrdthcy the socialneed that causedthem to come into exisrenccpasscs. An eventso rrivial as dre wereformed from ordinary languageswirh nativc sPeakers- therewas no jargon or pidgin disruprion ofa market may make x particularpidgin redundant.Population mov€ments phase. may haveth€ same€ff€ct. So thc raisond'€trc of the Pidgin Russianspoken in Harbin, Manchuria,between Russiqns and Chinese,disappearedwhen most ofthe Russiansleft in the 6fties. r.3 Distribution ofpidgins and creoles The questionofthe distribution ofpidgin and creolelanguages is one ofthe growth areas r.4 History o[ pidgin and creolestudies in linguistics.Because of their mixed characterthese specch varieties have frequently not been accordedthe statusoflenguagc. The fr€quentprejudice againsr their recognitionas Vhy should there be a field of pidgin and creolelanguage studiesi Since the group of properlinguisticsystemshas m€ant tha! listsoftheworld's languages,produced up tillfairly languagcsas awhole arenot geneticallyrelated, norspoken in the samearca, the languages recently,tended to ignote thesespcech varieties \Vhile many linguists,and sometrmes mustbe consideredto havesomething else in common, in order to be meaningfullystudied educationalists,recognize the fact oftheir existence,this isby no meansuniversellythe cas€' asagroup.ln the field rhereis an irnplicit assumptionrhar lhe creolelanguages share some The eFect of this is that new crcolcsand Pidgins arecontinually being addcd to thc lists propert,'that cellsfor en exPlanatorytheory of such languages. Thc carliestwrittcnsourcesfor manycreoles datc from the rSrhcentury whenmissronar- Recognitionhas come quick€stfor thosccrcolcs and pidgins (partially)bascd on Euro- icsstarted dictionaries,and tmnslaring religious texrs inro rhelanguages ofthc slaves. Th tudy ofpidg;n anl crcoh hnguages

(e) The 6rst time the term 'crcole'was applied to a languagewas 1739,in the Virgin Islands, Creolelanguagcs are olien assumedto be morc simple than other languages.There when rhe very youthful Dutchlexifier creoleNegerhollands was refeftedto es clniolsche isa wide-spreadbeliefthat creolelanguagcs are notjust morphologically,but alsosyntacti by a Moravian missionary(Stein 1987).The 6rst grammar ofa creolewas writren in the callyand phonologicallysimpler than other languages. Virgin Islandsby J.M. Magens,a scion of a local planter family (rzzo). In addition to $) Creole languagesare often assumedto have morc mixed grammarsrhan other missionarics,travellets or other laymen occasionallywrote briefdi:Joguesetc. in the local languages.Many peoplehave drawn parallels between language and biology,whcn rhinkrng ,mixed' creoies,a! lhat time generallyreferred to as Negro-English,Negro-Dutch, ctc. There are ofcreoles.Itis assumcdtharjusr asmanyspeakers ofcreole languages haw African, reasonablehistorical recordsfor a number ofcreole languages,including Negerhollands, European,Asian and in some casesAmerindian anccstry the languagesthcy speakare StananandSaramaccan (Surinam), ,andJamaican. These allow usto study likewisesimply a combination ofa bit ofEuropeanvocabularywith someAfrican orAsian the historicaldcvclopment ofthe creolelanguages (see chapter ro). syntaxand . (a) Creole studies originated as a s).stematic6cld of rcsearchovcr a cenrury ago, with Pidgin and creolelanguages are often assumedto cxhibir much more internal Schuchardt's(r842-r927) importantseries ofarticles. These starred as an attemPtto account variabilitythan otherlanguages.Theyare xsumed to be highly dynamic languagesysrems for a more complex setofdevelopments in lhe history ofthe Romancelanguages than was and often coexistwith their lexifier languagesin rhe samespeech communiry possiblein the Neogrammarianpreoccupation with the regularityofsound change-Hesse- Theseassumptions play a role in the varioustheoties ofcreoie oriqin that havebeen ling's (r88o-I94r)work origina.llysrarted out from an explanationofthe developmentsin proposed.The rheoriesof otigin havc becn developedin p"rt ,o.*plain rhe assumed Greek,from the earlydialects through koine Gr€ekunderthe Roman Empire,to Byzantine sirnilariry,simpliciry, mixing, and variabilityof the creolelanguages. \fe havechosen to and modcrn Greek. Both scholarsfound it ncccssaryto allow for more complex typesof group thesetheorics into four categorics,in chapters8-rr. Hcre we will btiefly summarize linguisricchange: mixture, simplification, reanalysis,and the complexiryoftheir analyscs the principalhypotheses pur forward.References will be providedin the relevantchaprers. characterizesmodern creolcstudies as well. Until r965the field remained,howevet, rarher marginal. Creole languages wcre studied r.5.r The Europea:rinput by a few cnthousiastichistorical linguists - usudly Anglicisa or Romanists,Geldworkers Somemodels attempt to rracerhe propertiesof the pidgins and creolesback to speciGc with an adventurousbent, or folklorists ahead of their time. Now the study of crcole antecedentsin (seefirrther chapter8)_ The PortuguesemonoG)genesis model has Ianguageshas moved to rheccnter oflinguistic tcsearch,a researchprogramwith univcrsalist undergoneseveral modifications. Crucial to all ofrhesc is the existenceofa tradelanguagc th€oreticelpretensions, half-way between theorcticrl lirguistic and sociolinguisticsReasons with a predominantlyPortuguese , used in rherith to tgth centuricsby tfaders,slave for this developmentare manifold, but include the political and cultural emancipationof raiders,and merchants from throughout the rhen incipient coloniaJ societies. The monoge, certainparts oFtheCaribbean (most notablyJamaica),an intetestin Afro-Americancuhure, netictheoryholds thar theslavcslearned the Ponuguesepidgin in rhe slavecamps, trading particularly in the U.S., and a paftial reorientationoflinguistic research. fons,and slavcships oftheir earlycaptiviry and rhen took this languagc,realiy no mor. thanajargon, wi!h rhem ro the .ThediFerent crcole languages as we know them arcbased on this jargon, but havereplaced the Portuguesewords with words Fromother r,5 Theories of origin in creolestudies and dreoreticallinguistics Europeanlanguages. The supposedsimilarity ofthe cteolelanguages is dueofcourse to the The main rcsearcheffort in pidgin and crcolestudieshas been to 6nd a principledcxplana- underlyingPonuguese jargon, and their simplicity to the simplicity ofthis jargon. tion for thc gcnesisofthe languagesin"olved. There isan implicit assumPtionthat thc creole fie restricted monogenesishypothesis is lessambitious. It is mosrly limited to tne l'X/hat languagcsshare some properry that calls for an cxplanatoryrheory propcrry this is Englishand Frcnch-lcxi6ercrcole languages oftheAtlantic and ,and proceeds dependson the theory concerned.Any offour Propertiesare assumed to play a role: fiom the ideathar rherewas a jargon or pidgin spokenalong rhe coastof Vest that (r) Crcole languagesare oftcn ass.rmedto bc more ulike than other languages.As we laterformed the ptimarysourcelor a wide rangeoIcreoles. The common leaturesof these will see,creoles share manystructural Gatures, and many researchersbelieve that theseresem- ctcolesare then assumedto be due to theseearly pidgins. blanccscannot be simply due to the similarity berweenthe languagesofwestern Europe, The Europeandialect origin hyporhesisholds that creolescssentially dcveiooed from or accidcntal. non-standarddialecrs ofthe colonial languagesin an ordinary way, and arc rhe resultof Then'dy of pilgin and ocole languagcs Picm Mrytkcn and Norual Snith migration by dialectspcakers to the newly foundedcolonies, compoundedwith thcexistence r.t., DeYelopmentalapproaches ofa strongly dialectalnauticrl language.In this thcory similaritiesbetween creoles hold Many researchersstudypidgins and creolesfrom a devclopmentalperspective, as gradually only for thosederived from a singlecolonial languagc;creoles may be simple becausethe evolvingand continuouslychanging sysrems rather than as stablesyst€ms lhat emerged non-standardvaricties w€re simpler than th€ written national standerd. rapidly.'JTithin this approach,expansion ofpidgins throughtheir continueduse and growth ln otherappoaches,proccses involvingthc rransformationofthe Europcanlanguages in functionaldomain isstrcsscd above strictly grammatical or cognitiveaspccts, [n cnaprer play a central role, thtough imPerfectsccondlanguage learningor the reductionofspeech II we retum to various developrnental approaches. dircctcd at foreigners.The or foreigner talk theory is similat to rhe imperfect The common social context theory adoptsa such stricdy functional perspective:the secondlanguage learning thcory in postulatingthat creolesare frozen (i.e. fossilized) stages slavcplantations imposed similar communicative rcquircmenrs on rheslavcs, ncwly arrived, in thc secondlanguage learning sequence.The diFcrcncelies in the fact that in dre baby and lackinga common languagein many cases.Thc commonality ofthe communicative talk theory the responsibilityfot thesimplificationis shifted from drelearners to thespeakers requirementsled to the formation of a seriesof fairly similar makeshiftcommunicative of European languages,who provide a simpli6ed model. The similarity bcwccn crcoles systems,which then stabilizcdarrd became creoles, would be due, in this view, to universal ProPertiesof the simPli6ed input. Thc type of evidenceadherents ofthc babytalk hypothesisarc looking for thus includessimpli6cations r,;.4 Universalistapproachcs made by native speakers,not by leaners, in pidgins, such asthe useofinfinitivcs' Univcrsalist models strcsslhe intcrvenrion ofa spccific generalproccss duriog thc transmissi- ln the imperfect secondlanguage learning thcory creolesare the crystallizationofsome on oflanguagefrom generationto generationand from speakerto speaket(see chapter u). stagein the developmentalsequence ofsecond languageacquisirion. Thc spcakcrsof the Thc processinvokedvarics: a gcneraltendency towards semantic transparency, 6 mt langr.g. proto-creolesimply did not have sufficient acccssto the model, and had ro make up an learningdriven by universalprocesses, or genera.lproc€sses ofdiscourse otganization, approximativesystem. In thisview the fact thatctcole-sare simple is dueto thc simPli6catior The semantictransparency theory is not afull-blown genesistheory, bursimplyclaims inhcrent in the second languagelearning Process.For some adherentsof this view the thatthe structureofcreole Ianguages direcdyreflects universal semantic structures. The fact possiblesirnilarities among the creole languagesale duc to universalProperties ofthe learning that they arealiLe, in this vicw, is due to thc fact tha! thc semanricsr.ucrures are universal. They aresimplc becauscthe scmanticstructures involved are fairly directly mappcd Process. onto surfacestructures, eschcwing any very complcx transformationaldcrivation. An example r,;.2 The Non-Europeal input ofthis maybe the factthat creolelanguages have separate tcnse/mood/arpecr particlcs, which TheMro-genesis modcl reallydeals mostlywith thecreolelanguagesspoken in theAtlantic reflectseparate logical opcrato$, ratherthan incorporatingtense, etc. into the region: \g€stAfrica and the Catibbean,and postulatesthat theselanguagcs have cmerged of the . through the by the slavesof the Vest African languages,rhe so-called substrate The bioprogrem theory claimsthat creolesar€ inv€ntions ofrhe children growing up languages,under influenceofthe Europeancolonid languages(see chaprer 9). An alternatrve on the newly foundcd plantations.Around thcm thcyonly heardpidgins spoken, without explanation is in terms of the transfer ofAfrican langr.ragestructures in the processof cnoughstructure to funciion asnatural languages, and theyusedtheirown innatclinguistic learning the colonial lexi6er languages.The similarity ofthe languagesinvolved is due, in capacitiesto trarsform thc pidgin input from thcir parentsinto a full-eedgedlanguagc. this model, to the fact that they sharcthe sameAfiican linguistic features,mixed together Creolelanguages arc similarbccause the innatelinguistic capacity utilized is universal,and with the lexiconofthe Europeanlanguages. Thc main problemswitht}re Afro-gcnesis model tiey aresimple becausc thcy reflectthe mostbasic langu€e stnrctur€s.Onc fcatureshared in its sttict versiod are lhe large number of structural dift'ereocesbetween Vest African by all creolesthat would derivefrom the innatc capacityis the syst€mofpre-verbal tense/ languagesand creoleson thc onehand, and the linguisricdifferences among !h€ variousWest mood/aspectparticlcs. Not onlydo theyseem limitcd in thc creolelanguag€s to a parricular African languagesthemselves on the other'\flhat hasbeen claimed to savethe hyPothesis sctofmeanings, but thcy alsoscem alwa)'s ro occur in a particularorder. The systemof is that in the proccssofrclcxification cenain syntacticand semanticproperties ofEumpear t€nse/mood-/aspectparticles, its interpretation and its ordering would directly reflect univer- Iexical items were incorporatedas well. salaspects ofthe human languagccapacity. p The$udt ofPidgin ald rftok hnguage' r.5.5 Theoretical implications termsofthe hierarchicalorganization offeature syst€ms,or a synragmadcdimension, rn In all rhescmodels or theoriesnotions suchas alike, simple,mixed, andr.eriable play a role. termsperhaps ofthe notion ofgovernment(Chomsky r98r) as a centtalprinciple ofsynracric They ate in fact taken for granted,assumed to be what requiresto be explained,and there- organiza!ion, make Al importantgroupofcreole researchcrs has focused on the fore not callcdinto question.The contribution that the studyofcreole languages can ' dynarnicand variable a:pecrs oflanguage(Sarkoffr982; in our view,to granmatical theory is thar it can help to elucidatethese four concePts'al ik€', Bickenonr97t; Rickfordr987). Vhile linguistsworkingin terms 'simplc', 'mixed', and'variable" All four turn out to be r€levantto the central concernsof ofthe paradigmofgenerative grammar tend to abstractaway from variation and change, modeangrammatical theory ln order to help usunderstand this, let usexamine the concepts focusingon thc universaland invariableespects ofiinguistic competence,many creolists have involved more closely. tended to put variation and changeat the centerofattention; only by studying the 'Whenwc saythat languagesr andT aremore alike than/ andz, we areclaiming in fact changesthat languagesundergo and the waysin which rhesechanges are manifested in the pidgin that in the cotal (abstract)variation spaceallowed for by the human languagecapacity r speechcommunitycan we 6nd out about the phenomenonoflanguage. aad crcolc languagcs andy are closcr thany and z. Consequendy,the claim that the creolelanguages arc more form a natural 6cld ofstudy for theseresearchers, precisely because they present dike thao other languagesimplies a clustering in the variation space lfrve think ofthe so much internal variation and becausethey tend to changeso rapidly. The extent of variation (and varia!ion spaceas defined by pammeter theory (asin recentworkby Chomskyand othcrs), pres€nt this is parricularlyrelevant for pidgins) again raisesthe questrons rrying to developa notion of'alike' redly boils down ro developinga th€oryofparametcrs, mentionedabove with respecrto rhc internal cohesionofa grammaticalsystem and how parameters parametersalong which similarities and differencesbetwe.n natural languagescan be determinethe way languagesv..y. Keeping defined. this in mind, rhcn, thc contribution ofpidgin and creolestudies to linguistic theory Consider now the conceptofsimpliciry. The ideathat creolelanguages are simple has is clcar Ve havecome to gripswith oneor morc ofthe corenotions ofgrammatical thcory: beentaken to mean two things.On one levelit hasmeantthat creolclanguages do not have a rich morpholory, on aootherthat the overallgrammarofcreole languages is lesscomplex alikc parameterrheory than that ofother languages.Both interpretationsare rclevant to gtammaticaltheory' The simplc: morphology/ryntaxinteracrrons idca that absenceofmorphology is relatedro grammaticalsimplicityneeds to beevalrrated markcdnesstheory in the context of contemporary researchinto morphology/syntaxinteractions' and thc (Chomslcy1982; Rizzi It82, and others)and ofcase mixedr modularity tra-rnmaricalstatusofindection or rrrr variable parametertheorymodularity marking (Stowell, I98r). Even more importantly, the ideathat rhe creolelanguagcs are not grammatically complex in generalonly makessense if one has a theory of grammatical Studying complexiry to fall back on, and this brings in markednesstheory. creolelanguages implies a consrantconfrontarion with thesenotions, and helps onero dcvclop vocabulary Consider next the notion of mixing. Mixing implies that elementsfrom one language a capableofdealing with them. arecombin€d with elementsFrom another, and this in turn callsinto questionthe cohesion Further o[ the grammatical systemsinvolved. The tighter a particular subsystem(e g the vowel reading The primarysourcefor documentation systcm;or thc systemofrcferential expressions)is organized,the lessamenablc it will be on the differentpidginsandcreoles isstill Reinecke's monumental (r975). to restructuringundct borrowing.Tightness oforganization in modernSrammatical thcory : bibliography Thcre areanumber ofintroductions to pidgin andcteole studies (r966), (r974 is conceptualizedin terms of modularity theory: the grammar is organizedinto a set of on thc marker,including Hall Todd r99o), Mtihlhiiusler (r986),wrur internally strucruredbut externallyindcpendent modules,the interacdonofwhich leads t muchinformation about the Paci6c,Holm (r988),strong on rhe history ofthe 6eld, and Romaine(r988), to the 6nal grammaticaloutput. For this reason'the notion ofmixing is important: it forces srrongonlinkswirh psycholinguisticresearch- In Frenchwe haveValdman (I978). us to think about which partsofthe grammar aretightly organized,and henceabout the In addition thereis a largenumber ofcollections ofarticles, ofwhich Hymes (r97r), Valdman(1977), (r98r) notion ofmodulariry and Valdman and Highfield are rhe mosr generalin scope, Tighrnessoforganizarion or cohesionmay haveeither a paradigmaricdimension, in Useful monographsby singleaurhors are: Bickerton G98t, which containsa hlghly The sx'j! ofPidgin and creok hnguagcs v z The socio-historicalbackground of creoles which documentsthe rcarlableexposition oF the broprogram hypothesis;AJleync G98r)' which presentstheview A-rends Afro-ge.r..i. hypoth.ris wirh a wealth ofdctail;and SankoF(r98o)' Jacques requiremcntswith ,i.,,i" *rrr.,rr.. of.r.ole languagesis finelyattuned to their functional journals' umal of Pidgin and ,'.,-b., of in"ightf.rl Th... t*o specialized /o " ",.i.[o. ".. is a newslcttet Thc Canicr Pidgin' Creob Languagu-end Eadzs Crcolet ln addition there

z.r Three typesofcreole

It hasbeen argued by somecreolists that creolescannot be defined asa distinct group of languagcson rypological,intralinguistic, grounds(e.g. Muysken 1988),If rhis is true, the questionarises whether there ar€ anyexternal, exttalinguistic, criteriaaccording to which theycan be groupedtogetherin one category The critcrionthat comesto mind most readily isthat ofrhe socialhistory oftheselanguages. Are creolescharacterized bya patticularsocial history a socialhistory that is common to all ofrheseianguages and that is not sharedby any other group oflanguagesiThis questioncannot be satishctorilyanswcred at the mo- ment, simplybecausethe externalhistoryof many creolesstill hasto be written, but there arcstrong indications that indccdin many casesthere are a number ofstriking similarities amongthe historicalprocesses rhrough which theselanguages came into being.One ofthese concernsrhe fact that manycreolesarose in rhecontcxtofthe Europeancolonial expansion from the sixteenthcentury onwatds.In many casesthis expansionwas accompalied by a spccifictypc of economy,which had asits most characteristicfeature rhe exploitationof relativelylarge agricultural units, plantations,for the production oflatgcly new products suchas sugar, coffee, and tobacco,for the Europeanmarkets. Plantations, however, were not the only situationsthat gaverise to creolization.Thcrefore, beforc wc go on to expiore the commonalitiesin the extcrnalhistories ofcreoles in general,we will firstbriefly discuss threedifferenr rypcsofcreoles that can be distinguishedaccording to differencesin their cxternalhistories. Accordingro rhcir external h istorythe following threetypes ofcreole have been distin- guished: creoles,fort creoles,and maroon creoles(Bicketon 1988).In addirion, a forth type may be distinguished:creolized versions ofpidgins haveemerged, e.g. in New Guinea and northern . In the Atlantic area,plantations wcrc worked by largc numbers of African slaves,who were purchased along rhe western coast of Africa from Senegalto Angola. In the initial stageofcolonization Amerindian slavcswcrc alsoused, as well asindentured laborers - poor Europeanswho wcrc conrractedfor a speci6ednumb€r ofIeats.In thccaseofthePacific and the Indian Ocean, was usually nor the primary meansofacquiring a labor forcc.lndenturedworkers from lndia, China,Japan,the Philip- pinesand thc South-WesiPacific, were recruit€d to work on the plantarionsin Mauririus, Queensland(Australia), and Hawaii. ThencioJ*torial bn&lround of oeoht

It is not only in the plantations,however, thar creolelanguages arose. Apart from rhe plantation creoles,which cmerged in the (e.g. in Jamaica,Haiti, Guyana, creoles(but cCAlleyne r98o for adifferentview). Other maroon crcoles,ourside Sunnam, Surinam), in West Africa (e.g. on the islandsof Annobon and Sao TomC off the Vesr include () and Angolar (SaoTom€), which are scill spokentoday. African coast),and perhapsin the southern parts ofNorth America aswell, a number of Finally,remnanrs ofanorhcr marooncreole have been found in rhe 'Maroon Spirit posses- creolesdevelopcd at rheso-called fons, the lortified postsalong the Vest Africancoast, from sion l-anguage'of Jamaica(Bilby r983),This is not empioyedin ordinary situations,but which the Europeansdeployed their commercial activities.In thc forts somc medium of it is uscdby peoplewhenthey are possesscd during religiousceremonics,to mlk to thespirirs communication must have been used, both among Africans from diffetent linguistic ofthose oftheir ancestorswho wereborn in Jamaica. backgroundsand berwecnAfricans and Europeans.More importantly,howevcr, interethnic The three-waydivision made hereat leastto some exrenrcuts acrossthe distinction communicarionexrended to the forts' surroundings where European men (so-calledlanga- betweenendogenous and exogenouscreoles. This distincrion was made by Chaudenson dos) were living in mixed householdswith African women, with whom they spokesome (1977)inorderto distinguishbetwecn crcoles that arosein areaswhere the nativelanguages kind ofcontacr languagc.In the courseof time thesecontact languages were expandedinto ofthe creolizingpopulation were spoken(e.g. somc African creoles,such as Kruba) and creolcs,in particular by the children that were botn into thesehouseholds. One ofthese thosethat did not, sincerhcy involvedthe massiverelocation ofrhe creolizingpopulation is the alleged'Guinea Coast Creole English, which, accordingto Hancock (t986), arose (c.g.the crcolesthat arosein rheNewVorld). The disrinctionis especiallyimportanr with out of rhe interaction berweenEnglish and African speakersin the settlemcntsin Upper respectto the potenria.lrole ofthe subsrratein creolegenesis: a creolethat arosein an area Guinea (SierraLeone and surrounding areas)and which may havc formed rhe basisofthe whcre its substratespeakers had ample opportunity ro cofitirru€ speakingtheir natrve CaribbeanEnglishJexi6et creoles. language(s)next to the emergingcreole is bound to showmore substtate influcnce than one A rhird type of socio-historicel context that hal give! rise ro the genesisofcteoles is thar did not (c[ Singler1988). marronage,which refersto the fact tha! slavesescaped from thc plantationsand subsequently formed their own communities in the interior in relativeisolation from the rest of the z.z Colonial expansionand the slavetrade colony. Maroon communities devclopcd in severalparts of the New Vorld (Jamaica, Colombia, Surinam) and in Africa aswell (S5oTomd). While most ofthese communities Thc history of Europeal cxpansiona-nd the concomitanrslave rrade cannot bc adequarery havebeen absorbed by the mainstreamculture ofthe societieswithin which they existed, describedhere, bur it cannot be excludedentirely eirher since i! constitutesthe socro- the Surinam , who are distributed over severaltribes, havepreserved their own historicalmatrix inwhich creolizationrook place.Therefore, inwhat followsa briefouttrnc traditions and rheir languagesup ro thc presentday. But since theselanguages probably will be given of this history as far as it concernsrhe Atlantic arca.The main Eur.g.., developedout of plantation creoles,we should not expectto 6nd impottant structural nadonsinvolved in the colonial expansionvr'€re Spain, , , Britain and the .What \i(4rile differencesbetween the two. may havecaused some divergence, however, is the fact Netherlands. thc Spaniardsand thc Porruguesewer€ rhc 6rsr to actually found that the maroon creolesdeveloped in telativeisolation from rhc metropolitan,European, setdementsin theNewVorld during the sixteenthcentury theywere followed by the others languagc.This issuehas asyet not been exPloredin any detail. e century later. In Surinam, nvo maroon creolelanguages can be distinguished.One, consistingofthe During the enrire slaverrade period some ren million Africans were captured aod dialectsspoken by rhe Saramaccanand Marawai tribes,is a'mixcd' creole,with tw" Euro- deportedto thc (Cr-rrtin 1969). Many ofthese did not surviu., .o-. di.d our,.,g peanlexiGcr languages, English and Portuguese(sce chapter 14). ln this respectit is clearly captivityin oneofthe forrsalong rheAfrican coasr, before theyhad even embarkcd on rheir different from the coastalcreole, Sranan, whose basic lexicon is English-lexifier,just likc middlepassage, thejourney to the New \0orld. Othersperished during $ansportas a resutt that ofthe other maroon creol€lanBuage, spoken by the Ndjuka, Aluku, Paramaccan,and ofdiseaseor other causesrelated to thc poor conditionson the slaveships. Of rhosewho Kwinti tribes.1*4rile some ofthe dialccts(e.g. Matawai and Paramaccan)have hardly been did arrivein the New World, manydied aftera relativelyshott period in the colonyrin rgth- studied at:Lll,Saramaccan has attracted the specialattention ofmany creolists,who regerd centurySurinamthe life expectanryupon arrivalwas somewhere berween 6ve and ten ycars. this languageas the most pure or radical creolclanguage cxtant today.According to these As far as the geographicalorigins and demographicbehavior ofthese Africansis con- scholarsSaramaccan, due to its supposedlyrapid formationand its subsequentisolation from ccrned,much remains to be discovercdby historical research.In the caseof Sunnam Jacque:Arcnzl: L9 extremelyd€tailed information has been made availableby the historian Postma(r99o)' 2.3 The plantation whosefindings arebased on archivaldocumenrs concerning the DutchAtlantic slavetrade syrstem port ofembarkation,dates ofdeparturefrom Africa Having discussed Thesefindings relatetosuch variablesas somerelevant socio-historical factors at themacroJevel,ir may be useful distribution, and the numbersofsleves that to go in some and arrival in the New \(/orld, ageand detail into a socialaspect of creolizationat rhe microlevel, i.e. the social Vhile we cannotgo into this in any detail,let ussumma- structureof the plantacion. rvereembarked and disembarked. After all, the plantation must have b€en the main locus of historicalresearch can contribute to the creolization. rizePostmfs main findings, asan exampleofwhat While most ofwhat followsrefers to Surinam(largely based on Van Strprraan and r8lt the Dutch shiPPedsome roo,ooo Africans 1993),it has study ofcreole genesis.Between I65o some rejevancefor orher creolesocieties roo. The srereorypicalimage of a than 2ooyearsafter imPortationbegan- th€ plantation to Surinam.At Emancipation,in r863- more colony asaseverel)'dichotomized sociery, wirh asmallnumber ofwhites holdine no more than some This showsthat during the powerover black population still numbered 36,ooo. largeflumbers ofAfrican siaves,needs some adjustment on the basisofwh"r very substantialpopulation reduction'due to an e\ceP- is known entire period ofslavery therewas a about how plantation life wassocially structured. Although there certainlywas low birth rate,while atthe sametime the rate a wide riooally high death rateand an excePtionaliy social,cultural and economicgap becween rhe small whiaesection ofa plantatron.s wasvery high As a resultofthis, the normal situationwhereby a language populationand the numerically ofimmigration dominant slaveforce (with ratiosreaching 1o:r ancl more), from one generationto the next, with second the actuel is acquired natively through transmission situation may havebeen a bir more complexrhan it might appearat 6rst sight. compieteiydisturbed. Although the precrse The figure languagelearning being only marginal, was belowchartsthe socialstratiEcation accordingto the divisionoil"bo. o., a rvpi.d linguistic consequencesofthis for creolegenesis are not entirelyclear, it seemsevident that Surinamplanr;rion. rhe roleof demographvshouldbe rakeninto :ccount As far as the geographicaibackground o[ the Surinam slavesis concerned,Postma providesa wealth of interestinginformation. From his Eguresit can be inferred (Arends to appearb) that duringthe slavetrade period therehave been substandal variations in the areasfrom which the Dutch purchasedtheir slavesThe generalpictule that emerg€sfrom thesefigures is that overthe entireperiod (161o-I8ti)the Mndward Coast(the areastretch- ing from SierraLeone to Ivory Coast)served as the main supplierofSurinam slavesHow_ ever,this areastarted to play this role only from r74o onwards During the first 7o years ofslaveimportation (I65o-17zo,the formativeperiod ofsranan and Saramaccan)theSlave Coast (Togo, Benin) and the Loango area(Gabon, Congo, Zaire,Angola) supplied more In rhe interveningperiod (r7zo-r74o),the than 9o7o ofall slavesimported into Surinam. domesticslaves Gold Coast (Ghana) servedas the main supplier ofslaves.Obvioudy' such a finding has important consequencesfor the investigationofAfrican survivalsin the Surinam "'"les' skilled slaves Thus, it seemssafe to assumethat in the formation of thesecreoles, l?indward Coast ndd slav€s ianguages,such asMande and \folof, despitetheir ultimat€ overallnumerical domrnance unproductiv€slavcs can only haveplayed a minor role, whereaslanguages spoken along the SlaveCoasr, such as Gbe, and in the Loango arca,such asKikongo, aremuch more relevantin this respect' researchmay drasticallyreduce rhe setof Vithin the This showsthat derailedhistorical-demographic black population therewas a division oflabor berween6eld slaves(who on creole(see further chapter Other coloniesfor average relevantsubstrate languages for anygiven 9) formed no more than around toyo ofa plantations blackwork force),house slaves, include Mautitius (Baker198z), Cayenne (Jennings to slavectaft:men, which similar work has been done and slavesperforming variousorher task, such as hunting and 6snrng. Guadeloupe(Singler I99zb). ThesediFerences appear),and and in funcrioncorrelared not oniywith diferencesin statusanJpower wirh in the black communiry, but alsowith the amounr of linguistic interacrionwith whites.A Thesocio-hitorical baekgroundof creoles

specialfuncrion was rhat ofthc so-calledcreole mame, a black woman - usuallyelderly - little is ftnown about difFcrencesin u vcrsusL2 use of€mergent who tookcareofthc youngerchildrcn. Sheis assumed ro havcprovided an important modcl creoles'it seemssafc to assume'on the basisofwhat is knowr for the acquisitionoflanguage by thesechildrcn, besidetheir patentsand oth€r r€latives, diff...n...*.."p....n, ;, ;;;;:r::ilJ Aparr from these,cvery slave community hadone or moreblack overseers, who occupied * o"li.fi.,T:::?[j::filHill ql*d p.ooss,thcy we.. ,o ;;;;.J an intermediaryposition in rhc power srructure,betw€en the white masterand the black tasksf 1, ",f ". ":*ooing "*.ign.i ;*, _** indudedinnoducing new slaves tothe l"d r..g"""g., "r, workforce.To him was delegatedthe execurionofpunishmcnt and allocationof task, as i... ,i;:; i".ii," .'*^ ,n* the model for Janguageacquisition was largely providedby blacra,not whites.In coronres also the decisionon when sick slaveswere fit to work. In addition to this, lher€ is some where,due to the demographicfacrors rcfer-red.; *"i"i,.r,,r.] evidcnccrhat hc was alsoa religiousleader in thc black communiry. ln many cases,quite r** increasedquite slowiy,rhe usk ofseaso "r.ii. "i.*.* surprisingll the black overseereven seems to havebeen in a more powerful position than u*n,",r,.,,r,",'ro.Jiy_il;;;il:;:::::l*'.::.j;Hft the whitcoverseer Probablythese diFerences in powerandstatus benveen different groups il,l:i#;n- *: .econd ianguageby the African-born ofslaveswere reflectedin their languageuse, just like in any other society,but, unfortu- I'i:":fjT.:l::,: slaveswoLrld be a second,not a nrsr tanguageversion of that creole.But again,it is exrrem.ly aitr.ul, nately,this cannot be empirically verified, duc to the absenceof documenlaryevidence. influence ,o ..,i_",. tt. this may havehad on the emergrngcreole. Apart from this, there must also havebeen considerable differences in rhe quality and Finally,two points have to bc mentioned,First, thc quantiry ofcontacrs belwe€ndifferent groupsofblacks on the one hand andwhites on the siavepopulation was not always aspowerless as is ofren assumcd,Because ofthei..h.., numb..., th. bf **. other.Thus, among the blaci