3rd February, 2002.

Taylor uses the term Island-Carib to describe a language variety or cluster of language varieties in use in the islands of the Lesser Antilles, from Antigua to Grenada at the time of the arrival of the Europeans. Taylor (1977:25) refers to Breton, a mid 17th century source on the variety spoken in Dominica, as reporting that the names these people used to describe themselves were Calliponam in the women’s speech and Callinaga in the speech of the men. These have reflexes in the speech of the Garinago or ‘Black Caribs’ of Central America, the only speech community currently using a language variety which can be traced back to these earlier speech varieties. For the modern Garinagu of Central America, the reflex, Garinago is used for the ethnic group, and Garifuna for the language.

The pre-Columbian inhabitants of these islands, before the coming of the Carib or Karina, were speakers of a language which Taylor reconstructs as an Arawakan language which is sometimes referred to in the literature as Igneri. At an as yet unspecified time prior to the arrival of the Europeans, but often estimated at around 1200 A.D, contact seems to have developed between the resident Arawakan speakers and Carib or Karina speaking peoples migrating into the Caribbean from the South American mainland. According to the most common historical version of this contact, male invaders conquered the Igneri, massacring the men and seizing the women as wives. The outcome is supposed to have been, in the first generation, a society consisting of women speaking an Arawakan language and men speaking a Cariban language, Karina.

By the time of the first European attestations of language within the community, in the mid-1600s, the bilingual situation involving the pre-existing Arawakan language, on one hand, and the imported Karina, on the other, had already broken down. It had been replaced by a single language, an Arawakan one, with gender based diglossia, involving partial bilexicalism. Taylor (1977, 28) suggests, based on evidence from modern Garifuna, has not entirely been outgrown three centuries later. Taylor (1977, 96) refers to Breton (1665; 1666; 1667). His works describe the language of speakers in Dominica in the mid-17th century and, from his description, considers himself to be describing just one language. This, he divides into (i) forms characteristic of men’s speech, (ii) forms common to both men and women, and (iii) forms characteristic of women (and of children) which he disparagingly refers to as ‘the children’s jargon and the women’s dialects’. What differentiated the language forms described by Breton as men’s speech was a higher level of lexical and sometimes structural influence from Karina than that which characterised the forms common to men and women or those which were restricted to ‘the children’s jargon and the women’s dialects’.

Did Breton indeed observe three different speech forms coexisting in mid-17th century Dominica and by extension similar other speech communities in the Eastern Caribbean? Let us examine the following observations made by Taylor (1977:96) of Breton’s work. For men’s speech, Breton gives as pronoun prefixes, i- ‘first person singular’, a- ~ e- ‘second person singular’, and k- ‘first person plural’. These are all lexemes of Karina origin. At the same time, Breton gives, as ‘common to the speech of men and women alike’, l- ‘third person singular, masculine’, th- ‘third person singular, feminine’, h- ‘second person plural’ and nh- ‘third person plural’. These latter are all lexemes of Arawakan origin. On the very next page, however, in presenting a word of Karina origin, uhémbou ‘belly’, he presents this word with the four pronouns of common speech, as well as the three pronouns, n- ‘first person singular’, b- ‘second person singular’ and ua- ‘first person singular’, which were, by virtue of not being listed with the men’s pronouns, part of women’s speech. The conclusion from the above, supported by other aspects of Breton’s description, is that there were, in fact, only two varieties. One was restricted to males. The other was common to all members of the community.

A sample of the 17th century relationship between male and common speech, can be seen as follows. In the two sentences, nebouiátina tibónam and chileàtina tone, the first is characteristic of male speech, the second of common speech. In the former sentence, the roots, /nemboui-/ and /-ibónam/ mean ‘come’ and ‘go’ in Karina. This contrasts with the female speech, which would take the Arawakan roots, /chile-/ and /-óne/, also meaning ‘go’ and ‘come’ respectively. What both sentences share is a single set of grammatical affixes, /a-ti-na/ (perfective aspect, 1st person singular) and /t-/ 3rd person singular female. (Taylor 1977:27). The Arawakan affiliation of even the men’s speech in Breton’s time can be seen, with grammatical morphemes of Arawakan ancestry being used with Karina lexemes.

The relatively superficial impact of Karina can be seen with reference to some Karina loanwords which survive in modern Garifuna. The evidence suggests that where Karina grammatical morphemes did occur, there is clear evidence that the users had no sense of the significance or original use of these morphemes. These simply were swallowed whole along with the lexical morpheme to which they were attached. Thus, in a Garifuna sentence ka siuámai bubáu hádagee ‘Which one of them do you like?’, the interrogative ka ‘who, what, which’ is Arawakan. The lexical item, siuámai ‘like’ is Karina in origin. The auxiliary verb, bubau ‘you do her’ is inherited, and the functional ha-da-gee ‘them-among-from’, has a Karina base, -da (Karina ta ‘in’), with Arawakan affixes. However, the lexeme, siuámai ‘like’, though treated as a verb which requires the auxiliary bubau, is itself an inflected verb in Karina, complete with subject and object, i.e. si-wama-e ‘I please him/her/it’. (Taylor 1977:98).

There were three trends already in existence in 17th century Dominican. There was firstly that involving the use of Karina lexical linguistic features to differentiate exclusively male speech from common speech. There was secondly that involving the erosion of Karina influence within the speech community and thirdly the trend towards the reduction in the level of difference between male speech and common speech. This latter involved both the loss of Karina features in male speech and the absorption of some Karina features into common speech.

We can see these trends in historical perspective with reference to modern Garifuna. There has been maintained, in modern Garifuna, certain relics of Karina influence that serve to distinguish between male speech and common speech. Thus, in modern Garifuna, there are emphatic personal pronoun forms, au ‘I, me’ and am?r? ‘you’, items of Karina origin in exclusively male speech now as in the mid-17th century, in opposition to nuguya ‘I, me’ and buguya ‘you’ which are Arawakan in origin and which occur in common speech. However, the trends towards the erosion of Karina influence and towards the reduction in the difference between male speech and common speech have also been maintained. In modern Garifuna, the three Karina personal pronoun prefix forms of men’s speech in the 17th century which we have already referred to, i.e. i-, a- ~ e- and k-, have died out, leaving in place only their common speech equivalents, i.e. n-, bu-, and ua-, all of Arawakan origin.

Lexical statistics support these observations. In Breton’s time, in the mid-17th century, there was 56 synonymous pairs of ‘non-cultural’ words, i.e. items from the 100 word Swadesh word list, whose individual members were used respectively in men’s speech and common speech. In the 20th century Garifuna, the differentiation between exclusively male speech and common speech has survived. However, the number of pairs of items differentiating between exclusively male speech and common speech, had shrunk from 56 to 6. The male speech item in each of these 6 pairs continues to be of Karina origin, with the common speech equivalent being of Arawakan origin. In the meantime, the number of Arawakan words on the Swadesh word list without a male speech competitor of Karina origin, has increased from 33 to 77. This provides supporting evidence for the other two trends, i.e. the erosion of Karina influence and the narrowing of the difference between exclusively male speech and female speech. The presence of 1 African and 16 Carib words which have no specialised male speech competitor, points to the trend towards a narrowing of the gap between exclusively male and common speech, but via a route which does not simultaneously involve the erosion of Karina linguistic influence (Taylor, 1977:76, 81).

What all of this represents is, on one hand, the persistence over several centuries of the need for the marking of difference between incoming males from the Karina group and women speaking an Arawakan language. Even when the ethnic element of the need for differentiation has long disappeared, the gender element in this need for differentiation has persisted. At the same time, however, the trend towards the erosion of differences between the incoming male conquering group and the female conquered group has been inexorable. The historical reality of being different is pulling in one direction, and the present reality of belonging to the same society is pulling in the other. The latter is winning over the former but only over a very protracted period of time.

I would suggest that the historical process outlined above is characteristic not just of the pre-Columbian era but the period since then. The process is one which can be characterised as involving different stages of what may be called conquest diglossia, i.e. diglossia imposed by the military conquest of one language group by another. The H language in a situation of conquest type diglossia, is imposed on a population rather than by choice adopted by that population, as is the case of German-speaking Switzerland.

‘… H can succeed in establishing itself as a standard only if it is already serving as a standard language in some other community and the diglossia community, for reasons linguistic and non-linguistic, tends to merge with the other community. Otherwise, H fades away and becomes a learned or liturgical language studied only by scholars or specialists and not used actively in the community. Some form of L or a mixed variety becomes standard.’ (Ferguson 1959:437)

The earlier Arawakan/Karina relationship represents the ‘conquest type’ diglossia of the Fishman type involving distinct and unrelated language varieties. Later, as links with the mainland Karina became more attenuated, there is a shift towards a diglossia of the Ferguson type, one in which the language varieties are related and can be viewed as varieties of the same language. The language situation in the post-Columbian ‘Anglophone’ Caribbean may be characterised as involving one or the other type of conquest diglossia, and/or a transition from one type of conquest diglossia to another. It is this model which will form the basis for the rest of this sociolinguistic description.

Garifuna is arguably the most successful of the indigenous languages of the region.

A major question arises about the relationship between the various varieties of the Arawakan language, referred to by Taylor as Island Carib, previously spoken in the Lesser Antilles. The two varieties of Island Carib for which we have some basis of comparison are Dominican and Vincentian, the two varieties which survived into the 20th century. These two varieties are themselves likely to have been dialectal amalgams of dialects spoken in other islands. This is the case since Dominica and St. Vincent were the two last islands controlled by the indigenous populations and became a refuge for Island Caribs driven out from other islands as a result of European conquest. Modern Garifuna, spoken in Belize, is an offshoot of the variety spoken in St Vincent.

A comparison of Modern Garifuna with the 17th century Dominican variety points to a pattern of shared idiosyncracies in the adaptation of loanwords. In some words beginning with /p/ in the language of origin, the reflexes in 17th century Karifuna and Modern Garifuna are /p/ and /f/ respectively. In other loanwords with /p/ in the language of origin, the reflex is /b/ in both 17th century Karifuna and Modern Garifuna.

Modern Garifuna 17th Century Karifuna Presumed Source

buírihu bouírocou puerco

búroburo boúrbr?ê pólvora

fádiri pátri padre

fáluma pálma palma

sabádu sabátto zapato

isíbuse (i)chiboúchi espejo

gaburána caboúranê caparona (?)

fanídira pántir bandera

In the above examples, a knowledge of the form of the loanword in one variety predicts the form which will appear in the other. However, the form of the item in the lending language cannot predict the form of the reflex in either of the two borrowing languages. The relationship between /p/ in the lending language and /b/ and /f/ or /b/ and /p/ in the borrowing languages is idiosyncratic. This suggests that the borrowing of these loanwords took place once, whether into the Dominican or Vincentian variety, and then spread to the other, rather than occurring twice. The reason for this conclusion is that exactly the same idiosyncracies associated with the exactly the same items are unlikely to have occurred twice.

In the meagre record of the 20th Dominican variety which Taylor (1977:79) was able to unearth, he finds eighteen Old World loanwords unknown to Breton. In this list, loanwords with a voiceless unaspirated alveolar stop in the Dominican variety systematically corresponds to the Garifuna voiced alveolar stop (Taylor 1977:78-9).

Modern Garifuna Recent Karifuna Presumed Source

muládu mulátu mulato (Sp.)

dábula tábula table (Fr.)

sáuderu sáuteru chaudière (Fr.)

fúdu pútu pote (Span.), pot (Eng.)

e-tegi-ra e-thek-ra thank ‘ee (Eng.)

tásu thásu tasse (Fr.)

mútu múthu múntu (Bantu)

Voiceless aspirated stops in the Dominican variety systematically correspond to the voiceless stop in Garifuna. This is in spite of the fact that the input forms from the Old World varieties involves [?] in one case, [t] in three cases, and [d] in one case. The shape of the loanwords in the languages of origin do not predict the [t]/[d] versus [th] in these two languages. However, knowing the form in one variety allows one to accurately predict the form in the other. The conclusion here is that whatever idiosyncracies operated at the time these words were borrowed, applied equally to the two language varieties. This suggests that a plausible explanation is that the words were borrowed once, into one or other of these varieties, and spread to the other, idioscyncracies and all. {fn?}

Taylor (1977, p. 36) argues that Karifuna, as spoken by the so-called Yellow Caribs of Dominica, and Garifuna, as spoken by the so-called Black Caribs of St. Vincent and Central America, remained mutually intelligible until the former became extent early in the 20th century. If we combine this with the fact of the shared idiosyncracies in loanword forms and we have, at a minimum, evidence for sufficiently close contact between the two language communities for loanwords absorbed into one to become spread to the other community. It is particularly significant, for example, that the one item of undoubted African provenance in the basic vocabulary of these two languages, i.e. /mútu/ or /múthu/ ‘person’, is shared by both dialects.

One aspect of African/Creole influence which has affected at least Central American Garifuna but not Dominican Karifuna, is the word order of demonstrative adjective and noun. In Dominican, the form of such a construction would involve the demonstrative followed by the noun, as in thukúra hiaru ‘that woman’, whereas the Garifuna equivalent would involve the opposite word order, hiaru tugura. Taylor (1977:5) suggests that this might be the result of French Creole influence on Garifuna.

In the 1990s, as part of the commemoration of the arrival of Columbus in the Americas, there was the Gli-Gli mission, consisting of Caribs from Dominica which visited Carib communities in Guyana. They attempted to establish language revivals using the mainland Carib community as their reference point. Of course, to the extent that they succeed, they would be re-establishing a linguistic link not with their most recent linguistic history, which involves the use of an Arawakan language, similar to that spoken by the Garifuna in Belize. Rather, they would be re-establishing a link with Karina which was introduced into Dominica around 1200 AD and which had long ceased to be a language in general on that island centuries before Columbus arrived.

References
Taylor, D. 1977.  Languages of the West Indies. Baltimore: Johns Hopkins.

-------------------------------------------------------------------------------------------------------

Lánarime lamiselu (Traditional Garifuna Song)
Lánarime lamiselu oubouwogu hianruyun (g.s.) bürihayan iyanha (Rep.)

What a terrible calamity on this earth, women now do the whistling

Luba oundaragua lira küroud, O!, ‘Whee! Whee!’

Before the gathering crowd, ‘Whee! Whee!’

Liyan hiyanhan beluya busiganu nuagu, Mama, idaya nuba me?

Their whistling brings me shame, Mama, what will become of me?

Au neremuhaba liyawana narihin

I (m.s.) sing what I see

Nuguya rügübadina me magu lan anhana laremagua aluguranou nisagadi

I (g.s.) will be the only that won’t be hurt by it, I’ll sell my grass

Tebegi nisabadun aü furisunba nuba aü

Money for my shoes or prison will be my home

Au neremuhaba liyawana narihin

I (m.s.) sing what I see

Numada, waü nariyanhaba bun luagu tusan turaboun

My friend, I will tell you about her ways

Niduhe waü nariyanhaba bun luagu tusan turaboun

My cousin, I will tell you about her ways

Luba barüba lubei fugiabu wamadaraü lun bubara rigubabei me lan fulasurugu

It-before take he-has-him ‘poor one’ ? you-will ? you-have-it it-in ‘all over the place’

Idaya waba me? (Rep.)

What will become of us?