Creole Lexicon*(Le lexique du créole)

Alongside the many words that have come from French and evolved into creole through basic sound changes (See ‘The Evolution from French to Creole’, link: ‘Etymologie’ <http://creoles.free.fr/Cours/etym.htm>), there are a number of other sources of creole lexicon.

Creole has developed variously from:

words originating from dialects from the North and West of France, cf. tchenbé/kinbé in Martinique or Guadeloupe, for example, which comes from ‘tiens bé’ or ‘tiens ben’, (‘to hold/keep’ from French ‘tenir’)

terms borrowed from some languages of the slaves, namely West African languages, relating principally to flora and fauna but also to aspects of daily life, such as malanga or tarot meaning an edible tuber and didiko (‘light meal’ or ‘breakfast’ - kikongo?)1

a few terms that have also been borrowed from Caribbean languages. This is confirmed by the recent new edition of Father Raymond Breton’s 1665 Dictionnaire caraïbe-français ([n.p.]: IRD/Karthala, 1999). Some examples include canari (‘clay pot’), mabouya (‘large transparent lizard’ similar to the gecko), and ajoupa meaning a temporary shelter or a hut made of foliage.

In spite of these roots and external borrowings, creoles are not necessarily any more composite than other languages since the processes of evolution are systematic.

Contrary to past claims and to what is often still apt to be maintained today, creoles are no more heterogenous than any other language. Comments that spring to mind in particular are those that tend to make Haitian Creole, for example, a language composed of Ewe syntax and French vocabulary or even ‘French words cast in the mould of African syntax’.2 More recently, Sylviane Telchid writes in 1997 on the subject of Antillean regional French that ‘the majority of its words, expressions, and phrases are borrowed from creole and hence from the languages that produced creole (Amerindian, African, Old French, regional French, French dialect, Spanish, English, and Indian)’.3 The names given to define the languages 1 For further study, reference may be made to Pierre Anglade’s Inventaire étymologique des termes creoles des Caraïbes d’origine africaine ([Paris(?)]: Harmattan, 1998) but caution is advised when consulting this work since often the etymologies suggested are highly imaginative with no attempt made to verify that the terms have not actually originated from French. Suggestions as to source languages are made in view of simple phonic comparisons without there really being any possible semantic explanation. Compiled within the most stringent methodologies of comparative linguistics, on the other hand, is Annegret Bollée’s Dictionnaire étymologique des creoles français de l’Océan Indien, ([Hamburg(?)]: [Helmut Buske(?)], 1993) of which Part II, ‘Mots d’origine non-française ou inconnue’ (‘Words of non-French or unknown origin’) is a unique source of information. Unfortunately, its subject is the Indian Ocean creoles and therefore, although it is sometimes possible to find explanations for terms common to both the Caribbean and the Indian Ocean, it is not tailored to present needs.2 Suzanne Comhaire-Sylvain, Le créole haïtien: morphologie et syntaxe (Geneva: Slatkine Reprints, 1974), p. 195.3 Sylviane Telchid, Dictionnaire du français régional des Antilles ([n.p.]: Editions Bonneton, 1997), p. 223.

1

here, such as ‘African’ and ‘Indian’, are as fanciful as the perception of Spanish or English as languages ‘that produced creole’. There is no doubt that creole has sometimes borrowed from these languages, but no more than any other language which has or has had contact with these languages and certainly less than French, for example.

It is worth remembering that the French language, which has a large Latin stock, also has a significant Germanic stock, acquired during the Germanic invasions in the fourth century, and that it has also borrowed from the various languages with which it has been in contact throughout its history through wars, marriages, and so on, such as Arabic, Italian, Spanish, Portuguese, English, German, and Russian.

One of the traits found to different degrees in the majority of French creoles is that of multifunctionality4 of lexical forms or, at least, of lexical base forms prior to any derivation which happens to be quite limited in creole. This means that the majority of single morpheme words can serve as noun, adjective, and even as adverb at the same time. Other words too can occasionally be either verb or noun according to their environment, with the particles indicating whether the phrase is nominal or verbal. In noun phrases, a subject or object phrase is indicated by the position of the word in relation to the main verb. The lexical base form is often not marked however, for example, manjé is a noun as well as a verb:

i ka manjé / i ké manjé (‘he is eating’ / ‘he will eat’) - verb phrase manjé-la / manjé-a-li (‘the meal’ / ‘his meal’) - noun phrase

It is possible to distinguish a number of recurrent phenomena in Antillean lexical creation and these historically attested processes, which have characterised the transition from French to Creole, are often behind new lexical creations that have remained in existence:

Agglutination of the article: the French article is often no longer separable from the base word so that creole words, such as lari (‘road’), monpè (‘priest’), and divin (‘wine’) take a creole article when spoken to give, for example, on lari, monpè-la, and divin-la-sa. This process can be compared, in contemporary terms, to a prefixation and seems to be used, in Martinique at least, to create abstract nouns, for example, lachichté, ladévenn, lakontantman, lakonportasyon, latranblad, ladoutans, and lakrintitid. This phenomenon stems historically from incorrect segmentation of words, bearing in mind that neither the masters nor the slaves in the initial stages of colonisation knew how to read or write, and is associated with the retention of the liaison [z]5 in words beginning with a vowel producing such forms as zarb, zozyo, zanoli, zèb, zyé, and zafè which are retained as such both in the singular and plural, cf. on zozyo and sé zozyo-la.

4 On the subject of multifunctionality, reference can be made to volume 1, ‘Les parties du discours’ (‘Parts of speech’) in Travaux du Cercle Linguistique d’Aix-en-Provence (1983). In this collection of articles, G. Hazaël-Massieux on the one hand and Daniel Véronique on the other discuss this question, the former in relation to parts of speech in Guadeloupe Creole (pp. 73-85) and the latter in a reflection on the existence of an adjectival class in Mauritian.5 International Phonetic Alphabet (IPA).

2

Aphaerisis, or loss of the syllable at the beginning of a word, has given forms such as pliché from ‘éplucher’, gadé from ‘regarder’, krazé from ‘écraser’, and blijé from ‘obliger’.

A few prothetics are found in front of ‘s’ + a consonant, for example, eskandal, espésyal, èspésyalis, and èstati from ‘statue’, although these are more rare.

Various metathesised forms ranging from érisi and présepsyon which have evolved from ‘réussir’ and ‘perception’ to phonemic changes such as siléma which is used alongside sinéma while koulfizi, koulpyé, koulwòch and a few others are considered ‘normal’ forms.

Some epenthesis, notably through the introduction of an ‘r’ between two vowels, for example, klouré from ‘clouer’ and pwaryé from pwayé (native ‘pea tree’).

While aphaerisis and agglutination of the article are generalised phenomena whose production is not linked to a particular social environment, it should be emphasized that metathesis and prothesis are, on the other hand, characteristic of popular or basilectal creole and are currently incorporated into the lexical creation processes which are presented at various times by the Groupe d’Etudes et de Recherches en Espace Créolophone (GEREC) since the aim of these creations is to fill gaps in the basilect which is being set up as the common language.

While inflection is limited in creole, it should still be pointed out that an exploration of lexical items which have different functions depending on their form as a result of parallel introduction into creole of inflected French forms would be important since it is possible that an assessment of the premises of possible grammaticalisations may be made from this. The Lesser Antillean creoles have retained some morphological elements suited to indicating the active/passive voices which other French creoles, in the almost universal absence of inflectional morphology, have to convey by different processes. Oppositions such as fè/fèt and pran/pri in Guadeloupe and Martinique allow the active/passive voices to be expressed:

i fè on bitin (‘he has done something’) >< i fèt Lapointe (‘he was conceived in Pointe-à-Pitre)i pran on liv (‘he has taken a book’) >< i pri (‘he has been caught/taken’).

The verb form ending in –é stands in opposition to the agent noun form ending in –è and examples include chanté/chantè (‘to sing/singer’), bagoulé/bagoulè (‘to deceive/swindler’), gouvèné/gouvènè (‘to govern/governor’), péché/péchè (‘to fish/fisherman’), bétizé/bétizè (‘to say or do silly things/someone who does or says silly things, a joker’), and tayé/tayè (‘to make clothes/tailor’). Since the average dictionary lacks many agent forms, merely listing the verb form for the most part, this process might naturally be extended to create new words since there is no reason not to have péyè (‘payer’) alongside péyé (‘to pay’), kachè (‘someone who hides’) alongside kaché (‘to hide’), and so on. As is also the case with French –eur however, it should be remembered that -è is not only found as a suffix marking an agent since there are French-derived forms, such as chalè (‘heat’), where the final –è is clearly not a suffix. By a quirk of history, some creole verbs also end in –è, for example ouvè or wouvè (‘to open’) and dékouvè (‘to discover’), most probably having formed from past participles (‘ouvert’, ‘découvert’). This is undoubtedly the case for many creole

3

verbs since initial verbal structures used in speech in the sixteenth century did not correspond to infinitives because the oral forms of infinitives and participles were identical for verbs of the first conjugation group (ending –er) and probably for verbs of the second group (ending –ir) and –ir verbs of the third group. In these circumstances, not forgetting as well that there are also quite a number of verbs ending in –i from the second and third groups of French conjugation such as fini and kouri, it is not possible to systematically extend the é/è derivation process to all creole verbs. There are also verbs ending in –é which have too passive a meaning to be able to generate an agent noun, for example, the verb graté means that someone has been made to itch by something such as a button or an allergy, as in the phrase bra a-moin ka graté, and hardly presupposes the existence of an active ‘itcher’.

Inflectional markers are also sometimes used for masculine/feminine opposition in order to indicate a person’s gender. The è/èz opposition works quite well and, as a creative process, can be applied as necessary to all agent nouns. If chantè/chantèz is possible, so is kolpotè/kolpotèz (‘a gossip’) and chanslè/chanslèz (‘a person who brings luck’), although ganmèz (‘an elegant person’) seems to be exclusively feminine! It seems, however, that current dictionaries have not always sought to systematically list both feminine and masculine forms of agent nouns and merely give the masculine form for the most part.

In terms of derivation, there a certain number of suffixes, mostly inherited from French but generally applied with different rules and values, which engender changes in grammatical categories, however it should not be forgotten that zero derivation is a highly fruitful process in creole and that it is sufficient very often to attach the necessary functional morphemes to a lexical base form to make it a noun or a verb.

Nevertheless, the possibility of creating verbs by the addition of a suffix –é or –té is perhaps on the increase in contemporary creole:

bétizé (‘to do something stupid’) from bétiz, gazonné (‘to mow the lawn’) from gazon, biginé (‘to dance the biguine’) from bigine, and balkonné (’to be standing on the balcony’) from balkon

kadoté (‘to give a present’) from kado and pyété (‘to stamp one’s feet, trample’) from pyé.

Nouns too can be created (mostly abstract as a result) by means of various suffixes but, once again, these processes are found mostly in popular creole. Examples are kontantasyon (‘contentment’) and tébètitud (‘being an idiot/imbecile/mentally retarded’); the etymology of tèbè is unknown. The use of any verb accompanied by the necessary articles is sufficient to make an abstract noun in creole, for example, biginé-a-li (‘his/her way of dancing the biguine’) and manti-a-ou (‘your lie’). It is worth pointing out the following possible suffixes (there are certainly others), however it is advisable not to exaggerate their true effectiveness for reasons previously discussed:

-aj: examples include dékolaj meaning dry rum taken in the morning on an empty stomach (to soak off the lizard lodged in the throat!), vonvonnaj (‘buzzing’), and dékatyaj (‘explanation’), a GEREC creation

-ans: examples include oubliyans (‘oversight’) and souvènans (‘memory’)

4

-ité: examples include lapépasité (‘inability’ or ‘impossibility’) and lapérézité (‘fear’)

-iz: examples include kouyonniz (‘stupidity’ or ‘foolishness’) and vayantiz (‘courage’)

-man: examples include kòzman (‘conversation’), protèjman meaning a small object worn as a lucky charm to protect against ills, and dékatman (‘analysis’ or ‘explanation’), a GEREC creation. This suffix is not to be confused with the adverbial suffix homophone –man, cf. touboulman, dousman, and so on

-ri: examples include kouyonnri (‘stupidity’ or ‘foolishness’), mantra (‘lie’), and toumantri (‘torment’ or ‘obsession’)

-syon: examples include modisyon (‘curse’) and profitasyon (‘cowardice’ or ‘abuse of power’)

-zon: examples include mantézon (‘lie’), toufézon (‘stifling atmoshere’), and soulézon (‘drinking bout’ or ‘drunkenness’).

In terms of composition, it should be pointed out that there are many, very productive processes as well as potential problems posed by classification inasmuch as it is precisely the multifunctionality of the majority of words that prevents classification as noun, verb, adjective, and so on. The order of priority that is widely accepted however is head-modifier, with the exception of a few nominal forms in which the first part could tentatively be said to be playing more the role of a prefix than of a true adjective. Without claiming to be either exhaustive or definitive, four categories are therefore suggested for the grouping of compounds:

(i) adjective (actually a prefix) + noun compounds: examples include timoun, tiponch, vyékè, and jènjan. The modifiers ti-, vyé-, jèn-, plus a few others (cf. gwo-, bèl-, etc.) belong to a closed category since the usual position of the modifier in creole is in postposition to the noun; even a noun that is in postposition to another noun is said to be its modifier, cf. timoun fwè-moin (‘the son of my brother’) and fwè timoun-moin (‘the brother of my son’), and so on.

(ii) noun + noun compounds (head-modifier): examples include on zinyam-pakala (‘a pakala yam’), on krab-sirik (‘a cirique crab’), on tanbou gwoka meaning a particular type of drum, on boutèj luil (‘a bottle of oil’), fanm-mayé (‘wife’), fanm-déwò (‘mistress’ or ‘common-law wife’), and kaka-zorèy (‘earwax’).

(iii) verb + noun compounds (tentative classifications): examples include on salibouch (‘an appetiser’), on bonbé-séré (a dance in which the dancers’ waists are tied together), and on bénirété (a formalisation of cohabitation, a marriage). The outcome can also of course be verbal: there are a considerable number of compounds with fè, for example, such as fè lous (‘to play the big tough guy’), fè kim (‘to froth/bubble/lather’), fè labé (‘to enter into holy orders’), and fè mas (‘to dress up for the carnival’) as well as many compounds with bay such as bay bal (‘to have fun’ or ‘to live it up’), bay blé (‘to lie’), bay do (‘to lose interest’), and bay lavwa (‘to spread the news’). The majority of these creole ‘verbs’ can give rise to compounds or, at least, to constructions

whose overall meaning is often very different from that of the parts

5

which can by and large be substituted for single words that constitute a substitution frame with the compound word

where the parts cannot be separated by the insertion of any element.

(Note: In languages with extensive derivation, an additional criterion in the identification of compounds is considered to be the frequent option to construct derived forms from the whole compound, for example, ‘right-mindedness’ is derived from ‘right-minded’, however this peripheral criterion does not apply to all compounds and, a fortiori, cannot be envisaged for creole, a language with marginal derivation.)

Of these classic criteria, which are often used to identify compound words, non-separability is accepted above all as the one that distinguishes true compounds from a momentary combination of two words in speech.

The ambiguity of the nominal and verbal character of forms appears clearly with some compounds, for example béni-rété (cited above) where, depending on context, rété can either mean ‘cohabitation’, as in on rété, or it can mean ‘to cohabit’, as in i ka rété. The question is whether on bénirété should be considered a compound of nouns or verbs.

(iv) verb + verb compounds also give noun as well as verb compounds, for example, i ka méné-alé + complement (‘he takes X’) and on ti méné-alé. Once again, the question is whether the combinations are of verbs or nouns.

The average creole dictionary contains very few compound forms and it is worth referring to texts in order to study the productiveness of such processes. In respect of the orchestration of creole, namely in the development of a coherent orthography, it would be advisable to undertake the systematic marking of compounds with the presence of a dash between the constituent elements.

* NoteThe sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

6

The Evolution from French to Creole*

(L’évolution du français au créole)

Once the regular sound changes involved in the evolution from French to creole have been understood, the etymology of any unfamiliar word can be found.

In Lesser Antillean Creole, for example, the French front rounded vowels [y], [ø], and [œ] automatically become [i], [e], and [ɛ].

The main exceptions to this rule are of a geographical or sociolinguistic nature:

in Les Saintes, dependency islands of Guadeloupe, the pronunciation of [y], [ø], and [œ] is regular.

it is not unusual to hear some Antilleans retaining the rounded variant through hypercorrection in order to appear either as a French-speaker or as distinguished or cultivated, for example, ‘rue’ [ry] becomes [lari], ‘queue’ [kø] becomes [ke], and ‘fleur’ [flœr] becomes [flɛ]

In an article in 1967, Guy Hazaël-Massieux reported this popular story from Guadeloupe which is a perfect illustration of hypercorrection:

A certain village craftsman, in order to hoist himself above his social standing, claimed to speak only French. One day, having acquired a dinghy, he asked the local primary school teacher to suggest a name for said dinghy which was painted blue and, in view of the colour, the teacher suggested ‘firmament’. On the Sunday, friends who had come to the unveiling were surprised to read ‘Furent ma mère’ on the side of the boat. Triumphantly, the craftsman explained that he wasn’t going to be duped so easily and he was perfectly aware that one said ‘furent’ not ‘fir’ in French and that ‘maman’ was just a vulgar form of ‘ma mère’. 6

Articulation of ‘r’

The French [r] regularly disappears at the end of a syllable in the Antilles with the latent ‘r’ also protecting the traditional contextual nasal vowel, hence ‘corps’ [kɔr] becomes [kɔ], ‘terre’ [tɛr] becomes [tɛ], ‘coeur’ [kœr] becomes [kɛ], ‘porte’ [pɔrt] becomes [pɔt], ‘morne’ [mɔrn] becomes [mɔn], and ‘charme’ [ʃarm] becomes [ʃam]. It also becomes modified in certain contexts, for example, the generally uvular ‘r’ (cf. French ‘r grasseyé’) moves forward to the velar point of articulation to become velar ‘r’ which is in fact [w]. This is why, for example, [wu] is heard for ‘roue’ and [wɔʃ] for ‘roche’. Rounding adds to this phenomenon, thus for example on ti bwen is heard for ‘un petit brin’ and pwan for ‘prendre’.

Contextual Nasalisation

6 Guy Hazaël-Massieux, ‘Remarques sur les créoles français des Antilles: problèmes de convergence linguistique’, in Proceedings of the 10th International Congress of Linguists (Romania: Editions de l’Académie de la République socialiste de Roumanie, 1969), pp. 727-731.

7

Contextual nasalisation is also a characteristic of Lesser Antillean dialects. Any oral vowel followed by a nasal consonant becomes nasalised, for example ‘maman’ is pronounced in creole as manman and ‘aimer’ becomes enmé and this rule is valid for both Guadeloupe and Martinique. It should be mentioned that there is also a strong tendency in Martinique towards progressive nasalisation, meaning the nasalisation of any oral vowel preceded by a nasal consonant, hence, combined with the regressive nasalisation, ‘aimer’ readily becomes enmen and ‘mais’ (mé in Guadeloupe Creole) becomes men in Martinican Creole.

* Note

8

The sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

9

Written Creole (L’écriture du créole)

As is the case with the majority of creoles, Guadeloupe Creole has no fixed orthography due to the fact that it is essentially an oral language. Writers have very often adopted their own system of notation and one system in particular has been seen to emerge since the 1970s as a result of a certain coherence that has been cultivated by some of the writers, together with an adherence to certain rules that they have collectively laid down.

In the three publications of their journal, Mofwaz,7 between 1977 and 1980, GEREC, together with Jean Bernabé (who revisited the subject in his thesis),8 proposed an orthographic system that was very close to a simple phonetic transcription. It is now widely used in the Lesser Antilles even though such a system is bound to present communicative problems as will be shown below.

Marie-Christine Hazaël-Massieux put forward an assessment, first in 1987 but principally in Ecrire en Créole9 in 1993, of the main problems encountered by an essentially oral language during its transition to writing and suggested a system of notation that would integrate very useful morphological and lexical data to make written communication easier throughout the Lesser Antilles. As a consequence, it is possible to summarise some of the problems encountered during the development of an orthographic system for a language.

How should creole be written?

Introduction to the problem

The idea of a written creole is still contested by many speakers in creole societies who undoubtedly assume that a language can only be either written or oral. While there have always been oral and written languages at any given time in history, the statuses of languages can change and a language that was purely oral (cf. Francien and all the Roman dialects in the Middle Ages when written communication was in Latin) can progressively arrive at written status just as French, Spanish, and Italian have, while others such as French Provençal continue to develop but remain essentially oral. Although there may have been various attempts at writing these oral languages, they have been insufficient in terms of turning them into real written languages for use in all the functions that are associated with many national languages. In short, the arrival of a language at written status, its use in multiple functions, its status as a national or official language, and so on are always a result of linguistic policy and, of course, of the willingness of the speakers themselves.

The creole situation: the existence of transcriptions

7 Jean Bernabé, ‘Ecrire le créole’, Mofwaz, 1, 2, and 3 (1977-80).8 Jean Bernabé, Fondal-natal. Grammaire basilectal approchée des créoles guadeloupéen et martiniquais ([Paris?]: Harmattan, 1983).9 Marie-Christine Hazaël-Massieux, Ecrire en créole (Paris: Harmattan, 1993). Further references to this text are given in parentheses after the reference.

10

It is certainly true that in any creole area at the present time, many samples of French creole writing are available and these are all essentially more or less phonological and phonetic transcriptions rather than orthographies in the strict sense of the word. This is due to the fact that no standard has been decided on either in terms of the notation principles or the variety that should be selected from the assortment spoken in all the different countries and French creole-speaking ‘départements’. Haiti is the only country where an orthography has been fixed by decree10 (in 1979) but not all writers observe this orthography and it is still questioned on a number of counts.

It is clear that the transcription of creole does not present any technical problems because, as with all languages, creole can be transcribed using, for example, the IPA and observing the principles laid down by the founders of such a notation system whereby one sound equals one symbol, in other words, each creole sound would correspond to just one graphic symbol. In contrast to the French sound [o], for example, which can be written as ‘o’, ‘au’, ‘eau’, ‘aux’, and so on, all [o] sounds in creole will be written as ‘o’, if the rules of phonetic transcription are observed, with a distinct symbol to avoid confusion such as ‘ò’ or ‘ô’ for the open ‘o’ sound which also exists.

Differences between transcriptions and orthography

Having established this, it is however necessary to underline the ways in which a transcription system is not an orthography. An effective orthography has to integrate more information than just the representation of pronunciation if it is to facilitate reading and the correct receipt of the message. There is a certain amount of information that can be passed on through orthography, for instance the different value of phonically identical words (homophones) can be indicated in the written form. Orthography also serves as a reminder of grammatical relations, for example, the interdependence between noun and determiner can be shown through agreement (as it is in French), by a dash, or by some other functional marker. These are just some of the elements that clearly mark the difference between a simple transcription, which is always easy to produce once the transcription alphabet is known, and a true orthography, which has communication designs beyond that of just the linguistic community and which is laid down to be used as a benchmark or norm (of both ortho- and -graphy).

The speakers faced with creole notations

In order that an orthography is accepted and therefore used by speakers, however, it also has to meet with their expectations to an extent. An orthography that has been developed without taking the speakers’ practices into account, which has symbols that are disconcerting, unfamiliar, or considered ‘foreign’, for example, can lead to its rejection.

The notation of creole presents many problems in this respect. Two systems, or more precisely a spontaneous notation which does not respond to any rules and which does not really constitute a system on the one hand and a system which basically advocates

10 The text of this decree of 18 September 1979 is published in Etudes Créoles, 3, no 1 (November 1980), 101-5.

11

phonetic transcription with no other real developments on the other are clearly in confrontation. Each writes as it pronounces and if a word is pronounced differently in different regions, it is written down differently with the result that comprehension completely comes to a halt the moment someone does not live in the same town or even the same district of a town or does not belong to the same social class. From this perspective, a Martinican and a Guadeloupean are unable to understand each other when communicating in writing in spite of the fact that they can understand each other perfectly in oral communication even though they may notice they each have different accents. People from the south of France do not pronounce French in the same way as those from Lyon who, in turn, do not pronounce French in the same way as people from Nantes or Alsace but everyone writes in the same way and this is precisely because the representation that was proposed was not aimed solely at the exact reproduction of pronunciation.

It should be pointed out that, in the Antilles, this phonetic transcription was quite widely rejected by the public and contributed to giving creole a bad public image. It must be said too that, even if the speakers cannot always say clearly why they rejected it, it is possible to bring out some important points from their comments:

Force of habit: this is certainly one of the points that must be taken into account. In the majority of French creole areas, those who write creole already write in French (or English in Mauritius, the Seychelles, Dominica, and Saint Lucia) and therefore show, a priori, considerable reluctance when it is suddenly suggested they write ‘manger’ as manjé or ‘courir’ as kouri, and so on. They also instinctively write creole words that resemble French words in accordance with French orthography, often getting the origin or the relationship of the word in question completely wrong into the bargain, for example, it is impossible to recognise the verb (vouè) and even less so the third person object pronoun in moin voueille which is so often written for moin vouè-li (‘I saw him/her/it’), often pronounced ‘moin vouè-y’. Not to mention the number of times moin caille is written for moin ka alé!

Comment: Force of habit cannot be permanently used as an excuse for not setting up a coherent system for creole which, moreover, must not under any circumstances be a repeat of the French system because, even if part of the creole vocabulary comes from French, it is a different language with different demands and a different history. When someone learns a foreign language, such as an Englishman learning French or a Frenchman learning German or English, they accept that the notation rules of the new language are not the same as those used for their mother tongue. From this perspective, it will be clear that, as a complete language, creole can have its own rules of representation. At the same time however, it is important, if only to facilitate learning, not to seek to systematically go against the writing conventions of the creole speakers. This is illustrated by the fact that, while ‘in’ and ‘en’ are both possible representatives of the front nasal vowel in creole, it is certainly preferable to retain ‘in’, which better suits the needs and pratices of those who have gone through the French school system (cf. Hazaël-Massieux, ch. IVe), rather than ‘en’ which would in fact lead to serious confusion. Two words spelt, for example, as ‘lapen’ and ‘chaben’ would be pronounced [lapɑ̃] and [ʃabɑ ̃] before the reader realised that the words in question were, in fact, meant

12

to be lapin and chabin. Moreover, this latter term, which is a pejorative label for a man, forms its feminine ending in –ine to give chabine. In the same way, it seems completely absurd to suggest writing the word peigne (actually pronounced more or less as [pɛ̃ɲ] in creole) as ‘penny’ in an English-speaking context such as the Seychelles. While the spelling ‘penny’ would be a possibility in Haiti where it would be directed at monolingual speakers who know neither French nor English, it would hardly make sense to suggest it for English speakers.

Absence of redundancy: a more serious issue is the fact that a phonetic transcription which is in line with the strict principles of the IPA cannot function as a writing system because there is too little redundancy (the IPA is used for scientific purposes, particularly in communication between researchers studying the oral form of a language). Any communication has to have a certain amount of redundancy to be effective, that is, it must give a piece of information not just once but several times and a system which, by definition, prohibits any redundancy is deprived of functionality. Comments such as ‘I cannot read creole,’ ‘creole is very difficult,’ or ‘you need to read things twice to understand them’ are evidence of the fact that the burden of redundancy has passed from the writer to the reader who has to read each word or sentence twice in order to extract the meaning. Redundancy implies that the information is given with largesse rather than the reader having to send out a search party for it!

It is in this sense that the discussion on whether it is preferable to use long or short forms merits inclusion. There is actually always the option in all French creoles of expressing a sequence of words more slowly or more quickly for any grammatical elements that differ. The shortened sequence notation often hinders reading considerably by impeding identification and, in addition, different shortenings can arise from dialectal varieties within the same creole language with the result that the speakers or readers of a different variety to the one that has been transcribed will neither be able to understand the word nor identify it and reproduce the pronunciation that is normal for them. The long form notation, on the other hand, has the advantage of acknowledging any of the different pronunciations and even of allowing silent reading, which is the most frequent method of reading for all adults, where a word is identified by sight rather than by pronunciation thereby shortening reading time considerably.

This is not to advocate a creole orthography with as much redundancy as that of French and a balance should certainly be sought between too much and too little redundancy (cf. Hazaël-Massieux). Any proposals in this respect will also have the advantage of bringing about a certain unification, which is the only way in which the distribution of literary works can become possible, as opposed to maintaining the dialectal dispersion which is at odds with a widespread comprehension of the language. No editor is going to agree to print and distribute a work which, because of the way it has been notated, will automatically remain exclusive in the sense that only those from Basse-Pointe in Martinique or from Fort-de-France will be able to read it. This is certainly the problem encountered by many writers who, more often than not, have to

13

resort to vanity publishing if they want their works printed in creole, although if it is success they are seeking, they must write in French.

In short, it is difficult to find solutions to the problem of orthography since they must reconcile:

scientific coherence political factors (such as division and unification of dialects that are close to

one another) economic factors psychological factors (such as representations and attitudes of the speakers)

Nevertheless, in the absence of any norm or established written form and although various issues remain unresolved, writers still write. There is a wealth of works in creole including stories inspired from oral tradition, poetry, drama, short stories, even novels and attempts at technical texts but people rarely read these works, often confessing not only to a great reluctance to do so but also to great difficulty. The fact is that people have to learn to read creole in the same way that they learn to read any other language and it is a little naïve, when confronted with a creole text for the first time, to confess to difficulties in reading when there has been no basic learning first. After all, it would not be possible for an Englishman to automatically be able to read French or Spanish. School certainly has a role to play in this learning, to educate any readers who are interested and likely to ask for more texts, thereby encouraging writers to produce. However, the often extremist political stances of the champions of creole do not make progress in the matter any easier because they cause confusion among the most well-meaning speakers. The legitimate demand, for example, for a place for creole in schools in the French overseas ‘départements’, where French would be perfectly compatible with creole language teaching, is mistaken for a demand for political independence since the independent parties often present creole as the ‘language of independence’. While changes in attitudes in linguistic matters are always long and difficult to achieve for reasons of linguistic insecurity, language status, and so on, they become impossible in a context where representations are distorted from the outset by deliberately sustained confusion.

The future of creoles

So what does the future hold for creoles? Will these languages that are still essentially oral one day become written languages in the same way that French, Spanish, and Italian progressively replaced Latin over the centuries? The situation is, in fact, very complex because each creole is only spoken by a limited number of speakers (except in Haiti) and is also always used in conjunction with a language that has international status (French or English in French creole areas) and which fulfils all formal and written functions. These two factors, together with the current importance that is attached to schooling (often compulsory to a relatively advanced age), where only the diglossic ‘high’ language is used, do not leave much hope for the various creoles in the long or short term future. This is in spite of the fact that the whole population speaks creole and it is the first language children learn and use until they start school which is a very different situation to that of the French dialects in France which are spoken almost exclusively by the elderly sectors of the population.

14

It is therefore difficult to give an opinion as to the exact future of creoles11 because there is no shortage of examples in history where a language that was reduced in one capacity or another has been able, given the right circumstances, to enjoy a real revival and to develop and prevail over the language that formerly enjoyed the more favourable status. There is no doubt, however, that writing now plays an essential role in the development and recognition of a language and if written creole is not taught and if creole is not used in all areas of expression, it is a pretty safe bet that all French creoles will end up one or two centuries from now in the same apparently barely recoupable situation that is faced by Louisiana Creole today, in other words in a situation of decline or even death.

Note: On the subject of Réunion Creole, some comments and/or examples that might be of interest are suggested by Pierre Louis Mangeard in his Master’s dissertation, ‘La détermination nominale en créole réunionnais. Essai de grammaire syntagmatique’ (University of Rouen, September 1996), which sums up certain aspects of the problem. His thesis, ‘L’écriture du créole réunionnais. Histoire, description et essai d’analyse’ (University of Réunion, 1998, p. 242), is also recommended reading.

Creole Pronunciation(La prononciation du créole)

As young, basically oral languages that have been used essentially for daily exchanges and do not have long written traditions as a result, the present question facing French creoles is clearly which is the best system of written representation. Depending on when they were written, any texts that are available have either been

transcribed according to the model that was available to the writer at the time which was French until very recently, or

transcribed (particularly since the 1970s) in accordance with the principles of a phonetic transcription (generally the IPA) as the linguists responsible for notation had become competent in various phonetic alphabets.

True readability of a text involves not just a simple phonetic transcription but a writing system that has been developed (See ‘Written Creole’, link: ‘L’écriture du créole’ <http://creoles.free.fr/Cours/ecriture.htm>) to take other factors than just immediate phonic reproduction into account. It is important that the linguistic unit is recognisable, on the one hand because adults read by sight when reading quickly and, on the other, so that the unit can be expressed correctly when read aloud as opposed to merely repeating the pronunciation since this can change according to immediate context or to the speaker’s region of origin. If one were to settle for a simple phonetic transcription, the pronunciation would undoubtedly be correct but the ability to see each syllable would not make for easy reading, just laborious deciphering. Indispensable grammatical information can be conveyed to readers through a number of graphic elements for fast and accurate comprehension of the text in front of them.

Once transcription of the creole language has been established, it is desirable that rules of pronunciation should be formulated for any reader who encounters it. This is especially true if the reader in question is ordinarily accustomed to reading other languages (which may, for example, contain many silent letters since many are

11 For further information, cf. M. C. Hazaël-Massieux, L’indispensable survie (Paris: Editions Entente, 1999).

15

pronounced in creole as will be seen) or is new to the language. The reader will know that a letter or group of letters is not always pronounced in the same way from one language to another (for example, ‘ch’ is pronounced [ʃ] in French, [k] in Italian, and [ç] or [x] depending on the context in German) and will, quite rightly, query the correct pronunciation of each letter.

The written forms used in older texts from the seventeenth to the mid-twentieth centuries are quite heavily etymological and comprise many silent letters, however, for contemporary written Lesser Antillean Creole, it is possible to state that:

all letters are pronounced, although a few digraphs should be noted, for example ‘ch’ and ‘in’ which represent one phoneme and not two successive sounds. It is suggested that a single silent letter should be introduced to facilitate the correct pronunciation of consonants, particularly nasal consonants, which are sometimes taken as digraphic elements, for example, ‘in’ is pronounced [ɛ̃] and ‘ine’ is pronounced [in] as in chabine, the feminine form of chabin;

letters are always pronounced in the same way.

The table below summarises the written forms in use, both single letters and digraphs, giving their phonetic value and creole examples of each sound.

LETTER PHONETIC VALUE (IPA)

CREOLE EXAMPLE MEANING

i [i] liv book

é [e] léléwhisk made of one branch dividing into two or three smaller ones for whipping sauces

è [ɛ] lèlè a gossip (male)a [a] kako cocoa

ô or ò [ɔ] bò side, (river) bank, rim, edgeo [o] bo a kiss, to kiss, to embraceou [u] foufou hummingbird

u [y] chuval; butin (cf. note 1 below) horse; thing

eu [ø] keu; dufeu (cf. note 2 below)

tail, bottom, handle, stalk, rear, queue; fire, light

eu [œ] keur; seul; jeune (cf. note 3 below) heart; only, alone, lonely; young

an [ɑ ̃] adan inon [ɔ̃] on, yon a, one

in (written ‘en’ by GEREC) [ɛ̃] chabin; fin pejorative label for a man; hunger

p [p] papa; lapo dad; skinb [b] ba; bitin to give; thingt [t] tini to haved [d] di to say, to tellk [k] kè; kinbè heart; to hold, to keep g [g] gèl; gadé mouth or ‘trap’; to look at, to watchf [f] fanm; fèy woman; leaf, sheet of paperv [v] voyé to sends [s] sòs; sè sauce; sisterz [z] zozyo; zyé bird; eye(s)ch [ʃ] chimin path, lane, track, way

16

j [ʒ] janmé; jounou never; kneem [m] manman; manjé mother; to eat, mealn [n] nonm man

ny (or ‘–gne’ in word-final position)

[ɲ] nyanm; montangne yam; mountain

l [l] lari; lékòl road; schoolr [r] diri ricey [j] pyé; yanm foot; yam

u [ɥ] lanuit (cf. note 4 below) night

w (sometimes written ‘ou’ in certain contexts)

[w] wou; bouè wheel; to drink

Table 5.1 Written forms in use in Lesser Antillean Creole

17

With reference to the notes in parentheses in the table above, the following points should be noted:

1. (‘u’) Although the realisation of a rounded palatal in some creole words, often in place of ‘i’, is without doubt a regional (common in Saintes) or social variable (some speakers take every opportunity to introduce the ‘u’ even when it is not required as it is considered distinguished), it is necessary and even constructive to make provision for its written representation due to the fact that people will not always know to replace the spoken ‘u’ with ‘i’ or ‘ou’ in writing. Even though the forms chival and chouval do exist and bitin is the most common form, they are not produced by the same speakers so it is not possible to totally forego the symbol ‘u’. This is all the more true because, as can be seen below, ‘u’ also quite readily represents the semi-consonant [ɥ] which is totally indispensable insofar as it is the only form available to render words such as uit (‘eight’), lanuit (‘night’), and juiyé (‘July’),.

2. (‘eu’) Exactly the same problems are found with this other more open, rounded palatal which is found in place of ‘é’ in some acrolectal and regional creole varieties. The most frequent pronunciation for these words, however, is ké and difé.

3. (‘eu’) There are so few examples of this sound, which is really only found in a context where the French consonant is latent, that one would not expect a different notation to the one given above. It should be written as ‘eu’ whenever the need arises and the notation of the consonant that follows it (cf. jeune) will ensure any confusion is avoided and will guarantee an open pronunciation. The preferred usage almost everywhere, however, is a non-rounded pronunciation as in kè, sèl, and jèn with, once again, ‘eu’ only being found in high acrolectal varieties, some examples of which can be found in Lapin ki vlé mandé Bondyé lèspri.12

4. (‘u’) The reuse of the letter ‘u’, already used for the vowel [y] above, does not present any problems given the rarity of the semi-consonant which, in any case, appears in a different environment to that of the vowel ‘u’.

Creole and Translation*(Créole et traduction)

While all languages present translation problems (‘non-equivalence’ of vocabulary, incompatibility of different grammatical systems, and so on), the problems are intensified in the case of creoles by the existence of so-called diglossic situations (See ‘Diglossia’, link: ‘La Diglossie’ <http://creoles.free.fr/Cours/digloss.htm>). The ‘birth’ of creoles during the era of European colonisation in the seventeenth century is relatively recent and there is little doubt that the links between them and their source languages, which is mainly French in French creole areas, have still not quite been severed.

In order to have a better grasp of the translation problems connected with diglossia, it would be useful, first of all, to cite a definition of translation and J. Dubois’ definition seems quite acceptable in terms of giving an initial rough idea:

12 Marie-Christine Hazaël-Massieux, ‘Lapin ki vlé mandé Bondyé lèspri’, Groupe Européen de Recherches en Langues Créoles, (last updated 15 July 2006) <http://creoles.free.fr> [accessed 1 September 2006] (link to ‘Un cours de linguistique et littérature créoles’).

18

Translation is the expression in another language (or target language) of what has been expressed in a source language with semantic and stylistic equivalence preserved.13

Everything revolves around these two terms “semantic” and stylistic” and the question in this case is whether there are actually two languages available in diglossia. It is worth briefly mentioning the English/French translation specialists, Vinay and Darbelnet,14 who noted that the translations of bilingual speakers are not the same as those of monolingual speakers living in their monolingual domain. Taking a Canadian and a French translator as respective examples, they wrote that the French translator is able to identify suitable equivalents immediately (familiar reality) and would know, for example, that the motorway sign ‘slow’ should be rendered by the verb ‘ralentir’ (‘to slow down’) rather than by ‘lentement’ (‘slowly’) and that ‘slippery when wet’, which would be translated literally by the Canadian as ‘glissant si humide’, should be rendered as ‘ralentir par temps de pluie’ (‘slow down when raining’) or ‘chaussée glissante’ (‘slippery road’).

The aim of this introductory comment quoted above from the two authors is to make one think about what translation really is. It is not a matter of simply repeating the original text in another language but of searching for stylistically acceptable equivalence by taking the cultural context into consideration which can, of course, lead to important ‘adaptations’. From this point of view, it may not be so much the bilingual, native speaker of the original or source language who is considered the best translator but rather the native speaker of the target language, brought up within the context of the target audience’s language, who is undoubtedly quasi-bilingual but who is, above all, imbued with the culture of the target language and capable of manipulating linguistic signs and connotations in the target language.

Vinay and Darbelnet emphasise the fact that translation also, above all, presents technical problems and that ‘one cannot suddenly become a translator just because one is bilingual’ (p. 24). It is in this sense that we present the problem of the diglossic context where creole-speaking writers are by turns writer and translator and we will attempt to show that not only is it not possible to ‘suddenly become a translator just because one is bilingual’ (p. 24) but it is often even a handicap to be in a diglossic context.

A diglossic situation is characterised by two languages that make complementary use of registers; what is said in one cannot be said in the other and vice versa. Moreover, speakers in diglossic contexts have often not had the opportunity to develop a register in one of the languages since they systematically go over to the other language when confronted with a situation that implies the register in question, hence the register is missing altogether. In respect of the ‘high’ language such as French, this is only a problem for the speaker who does not have access to the more informal registers, although people do make skilful use of ‘français familier’ (informal French) and ‘français populaire’ (working-class French) when translating the more informal creole registers. In respect of the ‘low’ language, in this case creole, drawing on a more formal register is much more problematic because speakers generally switch to French when required to speak in an ‘elevated’ fashion with the result that the high creole 13 J. Dubois, Dictionnaire de linguistique (Paris: Larousse, [2001(?)]).14 J. P. Vinay and J. Darbelnet, Stylistique comparée du français et de l’anglais. Méthode de traduction. (Paris: Didier, 1958). Further references to this text will be given in parentheses after the quotation.

19

register becomes, to a fairly large extent, confused with French. This is what makes it particularly difficult in the present state of creole development to render words or texts of a formal register in creole and, because of this almost inevitable gap (which writers are often happy to play on), tragedies become laughable and the Bible becomes a rather comical work.

The complex question of translation in diglossia can be represented in the following way:

In the case of strict diglossia (Fergusonian), the relationship between the two languages is represented below, where the black indicates an absence of language register:

acrolectal Frenchbasilectal creole

Translation is not possible when there is no white area opposite another.

In the case of bilingualism, on the other hand, which implies even-handed usage of two languages, everything that is said in one language can be said in the other:

acrolectal Frenchbasilectal creole

At present, diglossic situations are rarely strict. There is no doubt that the Antilles are moving towards relative bilingualism and, while not everything can be easily translated, there are a number of registers in both languages. Translation is feasible in conditions such as these where there are areas of possible equivalence which means that one is not forced to produce the incorrect translations that are all too often an inevitable consequence of diglossic situations lacking in any intervention where, for example, aristocratic French is rendered as working-class creole or informal creole is rendered as French that is too formal:

20

acrolectal Frenchbasilectal creole

This is of course a conjunctural problem linked to the state of development of the majority of creoles and not a ‘defect’ permanently hanging over them. Registers can, and must, be gradually developed to allow everything from the most serious issues to the most complex feelings to be expressed in creole and rhetoric must also be developed in order to be able to play on language registers. It has to be said, however, that development projects on Antillean creoles, for fear of relying on any French, have been concerned above all with developing the basilect (See ‘Continuum’, link: ‘Continuum’ <http://creoles.free.fr/Cours/continu.htm>). To this end, Jean Bernabé and GEREC are officially proposing to fill in the gaps in the basilect by gathering together all forms that are furthest removed from French, whether they are used in real communication or not, aiming to ‘construct’ a creole from voluntarist reasoning.

In Fondal-Natal,15 Bernabé undertakes to describe this ‘basilectal creole’ and justifies its selection on sociolinguistic and political grounds. In the same way that ‘the real basilect inevitably has features borrowed from French due to decreolisation’ (Bernabé, Fondal-Natal, p. 15), this theoretical ‘basilect’ is also partially (re)constructed because, when ‘gaps’ appear in it, Bernabé makes use of the concept of maximum deviation and retains the form furthest removed from French; in his own words, ‘the basilect aims to establish itself on the basis of maximum deviation in relation to French’ (Fondal-Natal, p. 16). Bernabé had already presented the concept of deviation as the founder of his reasoning in an earlier article16 where, within the framework of lexical research, he showed why a deviatory term should be created each time one similar to the French term appeared in speech. Hence his suggestion that the creole term jansiv should be replaced and how, ‘within a group of about ten secondary school pupils […], we created an original word to refer to ‘gum’ ‘ (Bernabé, ‘Recherches sur le créole spécifique’, p. 19), the result of which was ‘dyenndan’. (Extract from M. C. Hazaël-Massieux, Ecrire en créole (Paris: Harmattan, 1993), pp. 29-30)

Fortunately however, due to gaps between theory and practice, there are some excellent translations into creole. Among those particularly worthy of appreciation are the works of the Martinican, Georges Mauvois, whose excellent translation of Molière’s Don Juan was especially successful. Neither Molière, who handles many abstract concepts, nor the creole language have been betrayed in this work and the achievement merits the attention of all those interested in translation.

While this is the most important issue in French/creole translation, the differences between the languages and cultures, such as geographical or anthropological gaps, also of course present problems, as is the case with all languages, and can lead to

15 Jean Bernabé, Fondal-natal. Grammaire basilectal approchée des créoles guadeloupéen et martiniquais ([Paris?]: Harmattan, 1983). Further references to this text will be given in parentheses after the quotation.16 Jean Bernabé, ‘Recherches sur le créole spécifique. Part I: la désignation des parties du corps humain’, Espace Créole, 2 (1977), 13-38. Further references to this text will be given in parentheses after the quotation.

21

improper uses of words or faux-amis and so on. If an expression that makes reference to a daily reality in one language, such as two Russian men kissing on the lips, is translated literally into English for example, where this gesture is only conceivable between two lovers, it takes on an entirely different meaning to the one it actually has in a Russian novel. A cultural transfer is essential in English in order that the exact meaning from the Russian text is not distorted.

The difficulties in crossing from a language that says less to a language that says more should also be pointed out. While it is always possible to cross from a language that says more to a language that says less, for example it is not difficult to render the English ‘mutton’ or ‘sheep’ by the French ‘mouton’ (with the option of stipulating whether reference is to the living animal or the meat on a plate), it is far more complex to cross from a language that says less to a language that says more. Citing a Swedish example, Malmberg underlines what is automatically added by the translation ‘min farbror’ of the French ‘mon oncle’(‘my uncle’) because the translator has no choice but to convey in addition that reference is actually to the subject’s father’s brother as opposed to their mother’s brother:

The essential difference between languages consists therefore of the fact that two semantic elements follow one another in some languages but appear combined and without order in others. In the case of the word ‘oncle’, there is a dimension of order and hierarchy in French where it is ‘paternal’ that determines ‘oncle’ and not the reverse, but in Swedish the question of order does not arise. This encapsulates the (semantic) structural difference between languages and although it often complicates translation, it does not preclude it.17

Malmberg goes on to say:

We can thus summarise by saying that transformations of this type entail linear changes or changes in the succession of a combination of elements in the target language which do not know any dimension of order in the source language. Any transformation in this direction is therefore a structural simplification and any transformation in the other direction is a complication. Translation will entail excessive specification, unnecessarily so in the eyes of the target audience, where the source language represents the least developed structuring and will have an irritating lack of concretion in the reverse case. (pp.252-3)

When translating from a language that says more to a language that says less, there is always the possibility of syntagmatically adding some meaning in the form of determiners, parentheses, interpolated clauses, and so on or, at a pinch, by recourse to the ‘translator’s note’. It is often possible to compensate for paradigmatic absence with syntagmatic exposition provided that the translator is careful to make additions advisedly, although some expositions and periphrases, faithful no doubt to the meaning of the source word, cannot possibly cross over to the target language. Initiative plays a major role when translating from a language that says less to a language that says more; the translator needs to make good choices and adding meaning becomes an obligation rather than a risk. It is enjoyable to speculate on how to avoid making the distinctions that are, a priori, necessary in the target language and how to reduce or neutralise meanings. How can ‘uncle’ be said concisely when the 17 B. Malmberg, Signes et Symboles: Les bases du langage humain (Paris: Editions Picard, 1977), p. 252.

22

language does not allow it? Additional characteristics are unavoidable in this situation and, where it is not made clear in the source text, the translator will also have to decide whether the maternal or paternal ‘uncle’ should be retained from the various possible terms which all offer superfluous distinctions. There is hardly ever a solution here (with the exception once again of the great ‘translator’s note,’) and the translator is always forced to add extra meaning to the text and not always satisfactorily so opting, in the absence of a hypernym that corresponds to the general term in the source language, for a hyponym which is probably far too precise. It is interesting to note that while ‘seat’, for example, may be a hypernym for all terms such as ‘armchair’, ‘chair’, ‘settee’, ‘stool’, and ‘pouffe’ which were the subject of Pottier’s well-known semic analyses, there is no hypernym to replace such terms as ‘sideboard’, ‘dresser’, ‘wardrobe’, and other tall storage units. Every language has its richness and its gaps which are not always easy to manipulate. As Jakobson famously wrote,

Languages differ essentially in what they must convey and not in what they may convey.18

* NoteThe sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

18 Roman Jakobson, ‘On Linguistic Aspects of Translation’ in On Translation, ed. by Reuben A. Bower (Cambridge, MA: Harvard University Press, 1959), pp. 232-9 (p. 236).

23

24

Creole-Speaking Countries and their Populations*

(Les pays créolophones et leurs populations)

(French-lexifier creoles)

COUNTRYSURFACE

AREA (km2)

POPULATION

CENTRE OF ADMINISTRATION

LANGUAGES SPOKEN

Dominica 751 100,000 Roseau English, Antillean Creole

Grenada 344 100,000 Saint-George English, Antillean Creole (residual)

Guadeloupe 1,709 422,496 Basse-Terre French, Antillean Creole

French Guiana 91,000 157,277 Cayenne

French, French Guiana Creole, various Amerindian languages, Hmong, Chinese, Haitian Creole, various Businenge languages (English-based creoles)

Haiti 27,750 7,000,000 Port-au-Prince Haitian Creole, French

Louisiana 235,675 4,000,000 Baton Rouge English, Louisiana Creole (residual), Cajun

Martinique 1,100 381,441 Fort-deFrance French, Antillean Creole

Mauritius 2,040 1,100,000 Part-Louis

English, French, Mauritian Creole, Bhojpuri, Hindi, Urdu, and some other Indian languages, Chinese

Réunion 2,511 707,758 Saint-Denis French, Réunion Creole

St. Lucia 616 150,000 Castries English, Antillean Creole

St. Thomas 83 56,000 Charlotte Amalie English, Antillean Creole (residual)

Seychelles 410 70,000 Victoria English, French, Seychelles Creole

Trinidad 5,128 1,300,000 Port-of-Spain English, Antillean Creole (residual)

Table 7.1 An overview of creole-speaking countries

N.B. The figures indicated here for the four French overseas ‘départements’, Guadeloupe, French Guiana, Martinique, and Réunion, have been taken from the 1999 census.

NOTE: Do not be misled by this table which merely aims to give a few geographical details. The reason for showing the languages spoken in the final column is to complete the information with respect to the possible multilingualism of the ‘département’ or State in question but it is in no way possible to deduce from this the number of speakers of one or other of the languages spoken. While Haiti, with over seven million inhabitants, can be said to have seven million creole speakers, this is certainly not the case in Louisiana which has only a derisory number of people still speaking creole

25

among its four million inhabitants and where the language has almost died out as a result.

The maps below will prove useful in establishing the locations of both the Caribbean Islands, many of which are creole-speaking (Map 1), and the Indian Ocean Islands (Map 2) while the somewhat stylised Map 3 shows the Seychelles in more detail.

Map 1 Caribbean Islands

Map 2 Indian Ocean Islands

26

Map 3 Seychelles

More comprehensive information on the position of different creoles in the world in terms of their status and their actual usage will be found by consulting M. C. Hazaël-Massieux’s L’indispensable survie, p. 310.

* NoteThe sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

27

28

Diglossia(La diglossie)

There is a need to clarify the precise meaning of what is traditionally called diglossia since it is a term that is frequently returned to in this text due to the fact that many creoles must be characterised as functioning alternately with another (generally European) language according to certain rules (which are open to change).

One of the first to have systematically developed and defined diglossia was Ferguson in a famous article, ‘Diglossia’,19 which appeared in the Word journal in 1959, in which he endeavoured to define this type of language contact through four distinct situations:

- Greece: the alternation of Katharevusa and Demotic - Switzerland: the alternation of Swiss German and German- Arabic-speaking countries: the coexistence of literary and dialectal Arabic- Haiti: the alternation of Creole and French

Since the first definitions from Ferguson were reviewed and amended by Fishman (1971) and subsequently by many other sociolinguists over the years, we will avoid the extraneous question of whether diglossia implies that the high (H) and low (L) languages are inevitably related (Ferguson spoke of two varieties of one language). It will be quite clear that Ferguson’s analyses, which were perhaps valid in 1959, are often no longer a true description of the relationships between the languages he described. Demotic has now completely prevailed over Katharevusa and Haitian Creole is no longer restricted to the situations assigned it by Ferguson, having progressively taken over many areas of expression. Furthermore, it is questionable whether the concept of diglossia is still actually relevant in Haiti given that approximately 90% of the population are monolingual Creole speakers while only a fringe minority of the population is actually diglossic (or possibly even bilingual).

Clearly, taking Ferguson’s 1959 descriptions of Haitian Creole usage relative to French as the starting point measured against the description that might be given of them today could give the impression that there has been a major evolution. It is even possible to express some doubts as to the soundness of Ferguson’s description or, more precisely, to the population sample used for this description. Examining the correlation between French and Creole, Ferguson describes the status of these two languages in Haiti in the following grid:

19 Ferguson’s famous article has been republished in particular in two works which are quite readily accessible: Dell Hymes, Language in Culture and Society ([n.p.]: Harper, 1964), pp. 429-39 and Pier Paulo Giglioli, Language and Social Context ([n.p.]: Penguin Modern Sociology Readings, 1983), pp. 232-51.

29

H LSermon in church or mosque xInstructions to servants, waiters, workmen, clerks xPersonal letter xSpeech in parliament, political speech xUniversity lecture xConversation with family, friends, colleagues xNews broadcast xRadio “soap opera” xNewspaper editorial, news story, caption on picture xCaption on political cartoon xPoetry xFolk literature x

Table 8.1 Ferguson’s 1959 illustration of the status of French and Creole in Haiti20

This distribution is well known and has been discussed many times. Even if it is accepted that the position attributed to French thirty years ago undoubtedly corresponded to the more visible Haitian reality, it is also possible to find elements that show that this situation was already on the move and this is confirmed by accounts and documents that can be gathered today which reveal a significant evolution.

Although the general tendency is to turn to French in formal situations and to creole (which appears in different varieties according to the social class of the speaker) in more informal situations, there are many usages that cannot be ignored which conjure up a more complex21 representation showing clearly that diglossia is not typical for everyone, both because there are speakers who are actually monolingual, although their numbers are decreasing in Martinique and Guadeloupe, and because there are people who are likely to use either Creole or French in the same situation and who are therefore moving towards true bilingualism.

There is no doubt that there is a ‘français populaire’ in the Antilles but there is also an acrolectal, or distinguished, creole and it would be wrong to see creole only as the basilect and French as the high language. This classical outline of diglossia has had its day and is certainly to be substantially reviewed for the Lesser Antilles.

Superposing creole and French simultaneously on the Antilles, it is clear that both are subject to geographical as well as sociological variation. The lects of rural speakers are not the same as those of urban speakers and the lects of speakers of high social status, such as doctors or lawyers, are not the same as those who work, for example, as heavy vehicle drivers on building sites, even though they may be from the same town. It should be emphasised that, even among the working classes, French is not restricted to administrative and formal usages but is also a widely used second language. While this informal French, which again varies according to social class, may not be the same as that spoken in Paris or Lyon (but then neither is the informal

20 C. A. Ferguson, ‘Diglossia’, Word, 15 (1959), 324-40 (p. 329).21 M. C. Hazaël-Massieux speaks of a double continuum in ‘Continuum’ (link: ‘Continuum’ <http://creoles.free.fr/Cours/continu.htm>).

30

French spoken in Paris the same as it is in Lyon), there is no denying that there is an Antillean ‘français populaire’ whose features are yet to be systematically described. The literary French forged by such authors as Chamoiseau and Confiant is also quite obviously not the French that is used spontaneously by speakers even if it has been inspired by this ‘français populaire’ on occasions.

Summary

Diglossia is a term used to classify communication situations in societies that make complementary use in daily exchanges of two distinct codes which are either two language varieties or two languages. Certain circumstances imply the use of one of the codes, language A, to the exclusion of the other, language B, which can only be used in situations from which the first language is excluded. This definition comprises many variations, however. Although there are diglossic situations in a way in the majority of societies, for example in Britain, there is a contrast between the English used in exchanges between friends or while shopping and that used in university lectures or at public conferences, it must be emphasized that the preferred use of this term is in reference to societies in which the contrast is particularly marked and often backed up by the recourse to two distinct reference terms for the varieties in use (for example, standard language/patois, Katharevusa/Demotic in Greece, and French/creole in the majority of French creole-speaking areas). Generally, these diglossic situations are situations of language conflict whereby one of the languages is termed the ‘high’ variety in contrast to the other which is considered ‘low’ with the former being used in communication situations considered ‘noble’ (writing, formal usage, and so on) and the latter being used in more informal circumstances (conversations with close family etc.). Speakers have sometimes been known to question whether this second variety is, in fact, a real language.

31

Variation(La variation)

Variation is the term used to refer to the appearance of lexical units in different forms and is a phenomenon that exists in all languages but is perhaps most marked in creoles. As oral languages, creoles have been only minimally used as a tool for writing with next to no standardisation or provision of the teaching manuals which, by exerting a normative pressure on those who consult them, inevitably generate a certain linguistic unity.

It soon becomes apparent when studying a creole that pronounciation and grammatical forms can differ from one village to the next and that more or less the same object or concept can be referred to by different names in different areas.

In linguistics, distinction is generally made between:

diachronic variation, or variation through time, also called historic variation, for example, the second person singular, which is now represented by ou, would commonly have been expressed as to in nineteenth-century Lesser Antillean Creole.

diatopic variation, which is variation according to place or geographical variation, for example, ‘my hair’ is expressed as chivé-an-moin in Point-à-Pitre but as chuveu-an-moin in Saintes.

diastratic variation, or variation according to social class or to the social group to which a speaker feels they belong. In the diglossic situation that characterises the Antilles, diastratic variation often appears in the transition from French, representing the formal or higher level, to creole, representing the more informal level. Always with the deep-rooted feeling of being more ‘correct’ in the first case and more informal in the second, speakers will alternate between ‘il m’embête’ and i ka anmèkdé-moin or between ‘il est fatigué’ and i las.

finally, diaphasic or ‘stylistic’ variation, or even individual variation. This is more difficult to characterise clearly especially for those creoles that lack sharp description. However, among all the choices offered by the creole lexicon, it is possible that where X will prefer to say marakudja and Y will prefer pom-kalbas is perhaps of the same order as the variation that makes the distinction between gwayav, gouyav, and goyav. Once again though, it would be necessary to carry out surveys in these cases to confirm that geographical or sociological factors are not contributing to one or other of these choices; this is known as variation analysis.

Obviously, this internal variation in a language or dialect, sometimes called intralinguistic variation, should not be confused with interlinguistic variation which also exists in creoles, for example, speakers will say ich for ‘child’ in Martinique but ti-moun in Guadeloupe. There is, in fact, very little interlinguistic variation between these two creoles which clearly come under the same linguistic set (the same language in fact) albeit with some topolectal, or geographical, variants. There is considerable interlinguistic variation, on the other hand, between various other creoles, particularly between those of the Indian Ocean and the Caribbean. A Mauritian, for example, will use bann as a plural marker (bann liv-la) where a speaker from Guadeloupe uses sé

32

(sé liv-la); while both these forms are preposed, the Haitian will use the postposed –yo as a noun plural marker, as in liv-yo (‘the books’).

33

The Status of Creoles*(Le statut des créoles)

Although there are a number of French creoles, it should be borne in mind that the focus here is on Lesser Antillean Creole. In spite of (often deceptive) similarities associated with their common origin (essentially seventeenth-century ‘français populaire’), all creole languages are different and it is important to note that the command of one creole does not automatically lead to the ability to understand another.

It should be mentioned at this point that there are also creoles stemming from other colonising languages:

- Portuguese creoles: spoken, for example, in Cape Verde, Guinea-Bissau and in other countries, notably America

- Dutch creoles: Negerhollands, which has almost died out however some elements of Dutch are perceptible in Sranan and Saramacan even though these creoles have been heavily relexified in English

- English creoles: examples include Jamaican, Gullah, Sranan, and Krio (Sierra Leone)

- there are also, without doubt, some Spanish creoles: Palenquero (Columbia) and Philippine Creole can be identified as such and there is still large-scale discussion on whether the flourishing language, Papiamento, is a Spanish- or Portuguese-based creole

Although creoles are certainly languages in the sense that the word can be used to refer to ‘any linguistic system used to communicate’, it is primarily their status that is the issue for those interested in learning about creoles.

It is worth clarifying at this point why it is not uncommon for languages to be contrasted with dialects and patois and why the term ‘language’ is opted for when political and cultural recognition is explicit and why one or other of the latter terms is used when a ‘language’ is considered to occupy an ‘inferior’ position.

Languages do not, in fact, all fulfil the same functions nor have the same status. Some languages are national languages, some have official status and others are only used by a section of the population in daily exchanges. There is no reason why idioms, which are not written and which do not have official language status, should be denied the name of ‘language’ since the term refers to any doubly articulated linguistic system used to communicate, however, it is not uncommon to hear all non-written or local languages being called dialects or patois. J. Dubois gives the following definition:

A dialect is a form (we would say a variety) of a language with its own lexical, syntactic, and phonetic system which is used in a more restricted environment than that of the language itself. A dialect, which currently refers to a regional dialect as opposed to a language, is a system of combinatory signs and rules which has the same origin as another system,

34

called the language, but which has not acquired the cultural and social status of this language from which it has developed independently.22

The primary meaning of the word dialect is given, first and foremost, as a ‘specific form conferred on a language through diachronic evolution which is differentiated according to region’.23

These meanings concur but are, a priori, not pejorative and the term ‘dialect’ in no way marks out a language as inferior as long as one does not venture beyond these definitions. The confusion between dialect and patois in France is caused by the fact that, at present, all dialects are in a patois-type situation and have had seriously diminished status since the nineteenth century. It would be preferable, however, to keep the term ‘patois’, or preferably ‘patois-type situation’, to refer to languages or dialects of diminished status that are often dying out since it is regrettable to lose one of the terms by merging them. The term ‘dialect’ will therefore principally refer to a historically-derived geographical variety while the term ‘patois’ will refer to a language or dialect, often of diminished status, which is dying out and which can be principally characterised by the following traits:

- a non-written language as opposed to a standard variety responsible for all written functions

- a language in which not everything can be expressed because the speakers are accustomed to making use of another language for certain functions and where the so-called ‘noble’ functions are entrusted to the ‘high’ language (See ‘Diglossia’, link: ‘La diglossie’ <http://creoles.free.fr/Cours/digloss>)

- a language which is spoken by the oldest sectors of the population and, for this reason, is dying out. The vitality of a patois can be assessed by the age of its speakers, for example, the over 70s or the over 30s, and so on.

The etymology which has been suggested, in spite of criticism, for the word ‘patois’ might explain the pejorative value that has been attached to the notion. According to the Le Robert Dictionnaire historique de la langue française (edited by A. Rey), it is a deverbal noun from the Old French ‘patoier’ meaning ‘to wave hands around, to gesticulate (in order to make oneself understood in the manner of a deaf-and-dumb person)’, ‘to behave, to plot’, or derived from ‘patte’ (paw or animal’s leg) with the suffix –oyer.

Any description of Lesser Antillean Creole as a patois really has to be mitigated because:

- admittedly, it is almost entirely non-written, and

22 J. Dubois, Dictionnaire de linguistique (Paris: Larousse, [2001(?)]), p. 149.23 M. Arrivé, F. Gadet, and M. Galmiche, La Grammaire d’aujourd’hui. Guide alphabétique de linguistique française (Paris: Flammarion, 1986).

35

- usage is essentially limited to informal situations, for example, it is not really used in university lectures or administration and its philosophical vocabulary is, at the least, indigent

- but all children speak creole and any ‘foreign’ children sent to school in the Antilles learn it through exposure within a few months and use it in the playground.

It is in this last respect that the creole and Breton languages differ: - Breton is written very little or not at all, and- usage is limited to informal everyday situations,- but no children (whose schooling is in French) speak it and, at

best, it is only understood by those in the 40-60 age group who confess to not speaking it themselves.

A dialect or language that is no longer spoken except by the over 70s has hardly any chance of survival or revival. If spoken by the whole population, on the other hand, and with a well-managed linguistic policy, a language or dialect can gain status and, gradually, it will be possible to write and teach it, to endow it with literature, and so on. Once again though, its speakers must want to take part in its development and this is not always the case since there is often reluctance when it comes to a language of diminished status.

Creole situations are often described as diglossic (See ‘Diglossia’, link: ‘La diglossie’ <http://creoles.free.fr/Cours/digloss.htm>) which means that two languages coexist unequally within the same linguistic community (cf. R. Chaudenson’s Créole et enseignement du français (1989), p. 162). These two languages are often a creole and French or sometimes a French-based creole and English as in St. Lucia or Dominica in the Caribbean. In some cases, it is not a matter of just two languages coexisting but of three or even four languages being used in alternation according to rules which can seem complex but which a native speaker, integrated into his community, masters naturally. This is the case in Mauritius, for example, where alongside the French-based creole, Mauritian, there are other languages such as English, French, and various Indian languages including what is often defined as an Indian-based creole, Bhojpuri (the common language used essentially by the population of Indian origin). These languages, unequal in terms of status, are clearly not interchangeable and one rather than the other will be used depending on the circumstances and on the speaker or listener involved. Languages in diglossic situations are sometimes said to be in functional complementarity.

* Note

36

The sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

37

Continuum*(Quelques définitions du continuum)

An assessment of the term ‘continuum’ is offered by Bickerton in an article24 in 1973 in which he analysed some earlier articles, notably those of Alleyne,25 B. Bailey,26 De Camp,27 and Le Page and De Camp28 who had all used the term before him, and concluded by presenting an analysis of the situation in British Guyana through newly redefined concepts. Without clarifying their theoretical position in relation to Bickerton, Chaudenson and Carayol29 reused these various concepts but did not explicitly offer any new definition, merely applying the concepts in question to the situation in Réunion.

A continuum is made up of the set of variables attested in a linguistic community and this definition holds true without any additional preliminary need to question whether these variables have been taken from one or several languages, related or otherwise. These variables are organised in subsets, for example pronouns, verbs, and so on, and a lect (Bickerton first spoke of isolect) formed from retained variants can be determined at each point on the continuum. If the first person subject pronoun variable (v) has two variants an (v1) and moin (v2), lect L1 will be accurately characterised as having retained one of the two variants in use while lect L2 would be characterised by its selection of the other variant. Bickerton is insistent that no variable should be envisaged in simple binary terms, which at the extreme for example would be ‘presence’ and ‘absence’, but that several variants are possible. If the number of variables is large and the number of variants for each variable is also not inconsiderable, any continuum that takes into account all the variation can be highly complex as a result. From this perspective, lects must be organised on a scale, with one end labelled the ‘acrolect’ and the other labelled the ‘basilect’ while the whole intermediate zone is called the ‘mesolect’. As Bickerton points out in a note in his 1973 article, to be absolutely correct, one should speak of ‘mesolects’ in the plural because there is actually a set of lects between the two poles, however ‘mesolect’ is often used in the singular as a simplification although it is a misuse of language in a way. When this proliferation of lects becomes very unwieldy, the notion of implication comes into its own as a way of avoiding having to talk about a new lect every time a new variant appears. A lect is actually characterised, in a way, as a ‘bundle’ of variants which have implicational relations between them whereby the establishment of the presence of one particular variant also supposes the presence of another which is the product of another variable or even of a variable from another subsystem. Although these analyses can seem complicated, the underlying idea, also

24 D. Bickerton, ‘On the Nature of a Creole Continuum’, Language, 49 (1973), 640-69. Further references are given after quotations in the text.25 M. Alleyne, ‘Communication and Politics in Jamaica’, Caribbean Studies, 3, no. 2 (1963). 22-61.26 B. Bailey, Jamaican Creole Syntax (Cambridge: Cambridge University Press, 1966).27 David De Camp, ‘Towards a generative analysis of a post-creole speech continuum’, in Pidginization and Creolization of Languages, ed. by Dell Hymes (Cambridge: Cambridge University Press, 1971), pp. 349-70.28 R. B. Le Page and David De Camp, Jamaican Creole (London: Macmillan, 1960).29 Robert Chaudenson and Michel Carayol, ‘Essai d’analyse implicationnelle d’un continuum linguistique’, in Plurilinguisme, normes, situations, strategies, ed. by P. Wald and G. Manessy (Paris: Harmattan, 1979), pp. 129-72 and Michel Carayol and Robert Chaudenson, ‘Diglossie et continuum linguistic à la Réunion’, in Les français devant la norme, ed. by Nicole Gueunier, Emile Genouvrier, and Khomsi (Abdelhamid) (Paris: Champion, 1978), pp. 175-90. Further references are given after quotations in the text.

38

quoted by Bickerton, is actually that while variation is central to the theory, it is also necessary to establish that this variation is rule-governed and the linguist’s role is to extricate these rules which must make clear for a speaker the relative probability of rule x or y being realised in any given context or situation (i.e. the probability that one or other variant of a particular variable will appear).

While the theoretical outline is attractive because it introduces variation into the heart of linguistic description, the inevitable proliferation of lects is difficult to manage and the implication rules, specifiying that if a particular variable appears then another is also present, are the only solution to reducing the disadvantages linked to the description of variation. Each lect thus takes on a less random dimension.

The continuum described by Bickerton and the majority of authors after him is a continuum with two diglossic languages or dialectal varieties as its poles. In Bickerton’s study, ‘deep Guyanese Creole’ (p. 647) was the basilect at one pole and ‘Standard Guyanese English’ (p. 647) was the acrolect at the other. It is this outline that was taken up by Carayol and Chaudenson for Réunion who were conscious of the existence not just of a simple diglossia but of a ‘continuum running from the basilect to the acrolect through a series of intermediate degrees which constitutes the mesolect’ (p. 182). Their justification for the recourse to the concept of the continuum is their wish to integrate variation into theory and to no longer merely speak of ‘free variation’ as ‘simply an accident that is impossible to explain’ (p. 182).

A continuum is characterised then by the presence of a bipolar diasystem ranging from an acrolect, characterised by socially prestigious forms, to a basilect, corresponding to the language state that is socially low-status. Of course, the acrolect and basilect have a considerable number of linguistic traits in common and are only differentiated by a limited number of elements which allows a relative mutual intelligibility between the two poles of the continuum (Carayol and Chaudenson, p.182).

* NoteThe sections in Bradley Hand ITC font have been introduced to give a deeper understanding of certain concepts.

39

Code Switching(Qu’est-ce que le ‘code switching’ ou l’alternance codique?)

In line with J. J. Gumperz,30 code switching can be defined as the juxtaposition of passages within the same verbal exchange where the speech belongs to two different grammatical systems or subsystems. This phenomenon is very common among creole populations, within communities characterised by diglossic-type (See ‘Diglossia’, link: ‘La diglossie’ <http://creoles.free.fr/Cours/digloss.htm>) situations. Creole and French alternate in communication in this way according to complex rules which vary with the speakers, the themes of the exchanges, and so on. These various parameters and the way in which diglossic communication functions have been studied many times by authors including L. F. Prudent who has cited from and commented on a corpus31 of Martinican speakers which is characterised by code switching.

In order to reach a general theoretical level, it is important to attempt to interpret code switching data by extracting motivations for this alternation and conceptualising usages. The notion of continuum (See ‘Continuum’, link: ‘Continuum’ <http://creoles.free.fr/Cours/continu.htm>) is, of course, directly linked to that of code switching.

30 John J. Gumperz, Discourse Strategies (Cambridge: Cambridge University Press, 1982).31 This corpus can be found at <http://creoles.free.fr/Cours/intervi.htm>