"IT HAVE MORE SOFT WORDS":
A STUDY OF CREOLE ENGLISH AND READING
IN CARRIACOU, GRENADA
RONALD F. KEPHART
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Ronald F. Kephart
I first saw Carriacou in July of 1971 as a Peace Corps
Volunteer in training making a visit to my assigned job
site. Six weeks later, I began teaching Spanish in the
newly-opened Junior Secondary School. The school was
intended specifically for those children who reached the age
of 12 years and were unable to attend high school, for
either academic or financial reasons. Previously, these
children had to remain in their primary schools, often
sitting in classes with much younger children. The hope, at
the time, was that they would be better served in a separate
junior secondary with less emphasis on academics and more
"practical" work, such as cooking, woodworking, and the
Although children at the Junior Sec were considered for
the most part less scholastically capable than those who
entered secondary school, I was soon very impressed with
their quickness in learning Spanish. I had taught for a
year at a prestigious private boys' school in the States,
where the students were highly selected for and expected to
be university material. I saw no difference between their
capacity to learn and that of my new charges. Of course,
there were big differences in the materials at our disposal.
In Carriacou there was no language lab, and there were no
textbooks; everything had to be improvised. I eventually
produced little text booklets for the children on the
school's mimeograph machine. Despite these shortcomings,
the children were enthusiastic about Spanish and several
became Spanish teachers themselves.
I was at first puzzled by the complaints of teachers in
other subjects who talked of the children's "backwardness"
in such things as history, language arts, mathematics, and
science. In some cases, the very same children who were
doing fine in Spanish were failing these other subjects.
Perhaps because I was a language teacher, I soon began
to feel the glimmerings of a possible solution to this
paradox. Although Peace Corps had offered no serious
insight into the language of Carriacou I began to realize
that the language the children and others in Carriacou spoke
was very different from the English I had grown up with. I
began to see patterns in both phonology and grammar, and
although I did not realize it at the time I began
unconsciously internalizing a good deal of Carriacou Creole
At the same time, I knew that my method of teaching
Spanish, the "direct" method, made as little use as possible
of any language other than the target. Objects, drawings on
the chalkboard, and in-class situations served to get the
meaning across rather than lengthy explanations in English.
Of course, in the children's other classes, this was not the
case; there English was the medium through which the
children were expected to learn. I began to feel that this
was one of the major stumbling blocks to the children's
in-school learning, although at the time I did not have the
knowledge of linguistics and anthropology to understand the
In 1979, after five years away from Carriacou, I found
myself a student of anthropological linguistics at the
University of Florida. Quite by chance I discovered in the
library Bernadette Farquhar's grammar of Antiguan Creole
(Farquhar 1974). Reading the grammar, I recognized some
patterns as similar to those I remembered for Carracou
Speech, while others seemed very different. I received a
grant from the Inter-American Foundation to return to
Carriacou and produce a grammar of the Creole English spoken
there (Kephart 1980). In the course of this work, I found
that the differences between Carriacou Creole and English
were even greater than I had originally thought.
The research reported in this dissertation is a direct
outgrowth of both my experience in the Peace Corps and my
attempt at writing a grammar of Carriacou speech. The basic
idea was to give a group of Carriacou children access to
literacy through their native language, using an orthography
developed as part of my grammar. At the same time, the
children were periodically tested in English to test the
hypothesis that learning to read their own speech would help
them in their reading of English. The research eventually
took on the informal name "Carriacou Literacy Project" and
in this dissertation it will be referred to as such, or as
simply "the Project," for ease of reference.
Texts in Carriacou Creole have been written in the
orthography used in the literacy project. This facilitates
the printing of the dissertation and makes it more
accessible to non-linguists, who might be intimidated by
phonemic symbols. These texts have been enclosed in slant
lines (/.../) as a reminder that they represent a phonemic
From time to time it has been necessary to utilize
other than phonemic representation. In these cases, the
traditional linguistic symbols have been used:
[...] indicates phonetic notation.
/.../ indicates phonemic notation.
[...] indicates morphemic notation.
Articulatory phonetic symbols used are from the
American adaptation of the International Phonetic
Association's symbols (see e.g. Pike 1976).
In several places I have given long texts in
traditional English orthography to make the content more
accessible to those not familiar with Carriacou Creole.
One term notable for its absence from this dissertation
is "dialect." I have avoided this term because of the
negative connotations which it carries in the West Indies,
as well as the difficulty even dialectologists have in
defining it and using it consistently. Instead, following
Hudson (1980), I have preferred the term "variety." In
addition to lacking negative undertones, the notion of
"varieties of language" conveys more realistically what
actually happens in human language. Varieties of language
can be designated at all levels, thus while all humans speak
a variety of Human Language, English is a variety of
Germanic, General North American is a variety of English,
and Southern Appalachian is a variety of North American
Carriacou Creole English
I have used this term to refer to the language used by
most people in most unmarked social interactions in
Carriacou. This language is not as "deep" or "basilectal"
as, say, rural Jamaican Creole. Nevertheless, it is
certainly closely related to other varieties of Caribbean
and West African Creole English. Typologically, it is a
variety of Lesser Antillean Creole English with some
syntactic patterns similar to Sierra Leone Krio. In
this work Carriacou Creole English is abbreviated CC.
In general usage, the term literacy covers at least
three different concepts (Pattison 1984:119):
1. The mechanical skills of reading and writing at least one
2. The ability to use the skills of reading and writing to
get along in a cultural context ("functional literacy");
3. The use of reading and writing in maintaining a
relationship with the intellectual life of a particular
culture. It is this third concept of literacy, which might
be called "critical" literacy, that is most difficult to
attain in a non-native language. This difficulty is
increased when access to and information about the target
language is controlled by the elite and no alternative
outlet for self expression in the native language is
For the most part in this dissertation I have tried to
stick to the terms reading and writing when only the
mechanical skills are meant, reserving literacy to refer to
the last two domains.
Export English 2
The term "Export English" or EE is used in place of the
more usual terms "standard English" and "Internationally
Accepted English." The term "export" is used here to refer
to the highly idealized, prestige variety of a language
which is usually presented to the world outside the
metropole, as well as to those within, as the "standard"
No one actually has Export English as their native
language. Its closest spoken realization occurs only in the
most circumscribed contexts, as for example the language
used by announcers or "readers" on the BBC. The ideal form
of Export English, however, is print, because print is not
dependent on individual speakers' variations. Also, it is
possible and even customary for syntactic structures which
rarely if ever occur in normal speech to occur frequently in
print. In the form of print, Export English includes the
prescriptive rules of spelling and "grammar" which apply
only to print, such as the difference between cats and
First of all, I'have to acknowledge the patience and
willingness to endure economic hardship of my wife Willy and
our son Tommy, both of whom accompanied me during most of my
sojourn in Carriacou. Willy's main contribution to the
project was to keep me fed and healthy in mind and body--a
full-time job if ever there was one. Nevertheless, she
found time to do voluntary teaching at Bishop's College and
to participate in the life of the community. Tommy spent a
full school year away from his best friends, but he made
many new ones and rapidly became a fluent speaker of CC.
Secondly, I have to acknowledge the contribution of the
Carriacou Junior Secondary School students who participated
in the Project: Godwin Adams, Laurine Ashby, Agnes Bibby,
Brian Brooks, Jane Cornwall, Godwin Coy, Kester Douglas,
Jefferson Hector, Joseph James, Dennis John, Brian Joseph,
Jeffrey Joseph, Dexter Lambert, Israel Lendore, Bernice
McFarlane, Roy Newton, Fitzroy Noel, Louise Noel, Sherma
Noel, Album Patrice, Susan Phillip, Roderick Prime, Stanley
Quashie, Francis Reece, Anastasia Simon, Beverley Teka,
Anthony Thomas, Shane Thomas, and Lincoln Vesprey. I
believe these young people have added significantly to
knowledge about literacy and education in the West Indies.
Mr. Patrick Compton, the school Principal, readily
accepted me into the school's routine, made many
suggestions, and willingly adapted to my sometimes
inconvenient requests. The teachers, some of whom were my
students in Peace Corps days, were always accepting and
willing to allow me to observe and tape their classes.
One teacher in particular made a spectacular
contribution: Beverley Bartholomew. Normally the home
economics teacher, she enjoyed reading the materials I had
prepared for the children and willingly gave up her off-duty
time to work with the class, freeing me to observe more
Of course, the Project never could have been undertaken
without the cooperation and support of the Grenada Ministry
of Education. The Education Minister, Jacquelyn Creft, gave
permission for the Project and ensured the cooperation of
the Curriculum Development Unit of the National In-Service
Teacher Education Programme. Two members of the Unit, Merle
Hodge and Janice Hamilton, spent time discussing the Project
with me. While they did not always agree with my methods,
their criticisms and suggestions helped. Another Ministry
official, Carlyle John, discussed the work with me on
several occasions and offered very useful insights and
Victor Ashby, an old friend from Peace Corps days,
provided room, board, and comradeship on the shortest
possible notice whenever I made the trip to Grenada. As
Headmaster of Grenada Boys' Secondary School, he is acutely
aware of the problems I was researching and we spent many
hours sitting on his veranda drinking Carib beer and
discussing the pros and cons of letting children read Creole
English. He was a most effective devil's advocate and only
the members of my committee have grilled me harder on the
theoretical underpinnings of the Project.
In a similar way, Judith Hemming and Michael Simons,
both educators from Great Britain, forced me through their
criticism to look carefully at some aspects of my work.
Again, it is important to note that disagreement can be far
more productive than agreement.
On a more personal level, I have to thank the people of
Carriacou who have always made me feel "homely" and who just
might be the friendliest people on Earth. In particular,
our neighbors the Edmunds shared much friendship and food
across the fence. Roy Benjamin arranged a very fine house
for us and later taught me how to fish on the rocks; as a
teacher, he was also interested in the Project and discussed
it with me often. His uncle, Mr. Peter Benjamin, another
old friend, provided us with land for a garden and taught me
a great deal about Creole French and life in Carriacou in
the early,1900s; we also spent many hours playing music
together. Sydney Cudjoe, a former colleague at the Junior
Sec, provided many hours of conversation and comradeship
strong enough to survive the clash of born-again
Christianity with secular humanism. Others who deserve
special mention for various reasons include the Bethel
family of Windward, Flossie Fleary, Euthan Samuel, the
Daniels, Alfred George, the Bullens, Aimee McDonald, Dolly
Fortune. All these people, and others too numerous to
mention, make Carriacou a place I always feel homesick for.
At the University of Florida, I have to thank
especially Dr. M. J. Hardman who introduced me to
anthropological linguistics and has guided me through the
rigors of both my master's and doctoral research. She and
the other members of my committee, Allan Burns and Jerrie
Scott, have been willing to read and criticize the
dissertation in pieces as it was written, which I think
makes for a stronger and more useful final product. Others
at Florida who have been especially helpful are professors
Terry McCoy, Helen Safa, Russell Bernard, Paul Doughty, Gary
Miller, Robert Lawless, Norman Markel, and Doreen Ross.
Outside the University of Florida, a number of
professors have been very helpful in various ways.
Johnnetta Cole of the University of Massachussetts helped me
get in touch with then Education Minister Jacquelyn Creft
and later gave advice and guidance during a visit to
Carriacou. Don Hill, State College of New York at Oneonta,
has corresponded with encouragement and enthusiasm.
Lawrence Carrington, Richard Allsopp, and Dennis Craig, all
of the University of the West Indies, offered interest and
encouragement. Ian Hancock, University of Texas, has
corresponded extensively and read my preliminary report on
the project at a conference on education in societies
without a written tradition in London. Robert Calfee,
Stanford, discussed the Project during a chance encounter on
a plane flight. John Honey, Syracuse University, questioned
me at length in Carriacou on the theoretical basis of the
A number of fellow graduate students at UF have also
helped either by example or by their willingness to discuss
various aspects of the research: Lawrence Carpenter, Shoko
Saito Hamano, Kofi Akwabe Ameyaw, Jim McKay, Dale Stratford,
Monica Lowder, Jane Collins, Mike Painter, Dan Cring, Tom
Jacoby, Jean Gearing.
The research reported here was funded by an
Inter-American Foundation Doctoral Fellowship. I would like
to thank Elizabeth Veatch, Mel Asterken, and the members of
the fellowship selection committee, all of whom provided
immeasurable support and guidance during the period of the
fellowship. It is worth noting here that the basic research
on the language of Carriacou, without which this research
would have been impossible, was also supported by the
Inter-American Foundation. When the Project was interrupted
by the political events of October-November 1983, and
additional funding was needed for a return visit in the
summer of 1984, the Inter-American Foundation, the
University of Florida Graduate School, and the Florida
Foundation all contributed.
Despite all this excellent help, I have to claim full
credit for any errors contained herein.
Permission to Reproduce
Permission is hereby given by the author to educators
in Grenada, Carriacou, and Petit Martinique, and other
officially English speaking areas of the Caribbean Basin, to
reproduce any part of this dissertation for use in
furthering the understanding of language, culture, and
education in these areas.
1. Personal communication from Ian Hancock.
2. The term "Export English" was first suggested by Dr. M.
J. Hardman. The concept grew out of our discussions about
my Peace Corps experience and later linguistic fieldwork in
Carriacou. See Hardman and Kephart (forthcoming) for a more
TABLE OF CONTENTS
PREFACE . . . . . .
Background . . .
Terminology . . .
Acknowledgements . .
Permission to Reproduce
Notes . . . .
ABSTRACT . . . . . .
I INTRODUCTION . . . . .
The Problem . . . . .
Project Goals . . . .
Research Design . .
Why Anthropology? . . ...
Orthography . . . . .
Carriacou: The Land and People
Education in Carriacou .
The Junior Secondary School
Notes . . . . .
. . . . . 1
II ECOLOGY OF LANGUAGE IN CARRIACOU . .
Classification . . . . . . .
Mythology of Afro-American Language . .
Origins of Afro-American Language . .
Contributing Languages . . . . .
Context of Language Use . . . . .
Glottopolitics: "Nostalgia as Repression"
Notes . . . . . . .. .
III SYSTEM AND CONTRAST . . . . ..
Phonolgy . . . . . . . .
Morpho-Syntax: Noun Phrase . . .
Morpho-Syntax: Verb Phrase . . . .
Some More Syntax . . . . . .
. . 34
. . 62
. . 62
. . 71
. . 75
. . 86
. . . . . iii
EE vs. CC: Some Areas of Contrast . .
Notes . . . . . . . . .
IV A MODEL OF READING AND ITS IMPLICATIONS .
. . 91
* . 98
* . 99
A Model of Readin
The Importance of
Two Other Ways
g . . . .
n and Reading . .
Orthography . .
the Carriacou Projec
. . . . . .
Notes . . . . . . . .
V THE HISTORY OF AN APPLIED LITERACY PR(
September-December 1982: . . .
January-March 1983 . . . . .
May-July 1983 . . . . .
September-November 1983 . . .
The Final Session . . . . .
Four Children . . . . .
Notes . . . . . . .
VI CONCLUSION . . . . . .
Results . . . . . . .
Limitations and Further Research .
Implications . . . .
REFERENCES . . . . . . . .
A FIRST PRETEST . . . . . .
B SECOND PRETEST . . . . . .
C THE PRIMER . . . . . . .
D LETTER TO PARENTS . . . . .
E WAN WIK-END IN KARYAKU (SAMPLE PAGES)
F DI LAJABLES CHAYL (SAMPLE PAGES AND T
G CC WORD RECOGNITION TEST .. . .
H FIRST READER . . . . . .
I FIRST POSTTEST . . . . . .
J TEXT OF /A MAN IN DI VILIJ/ . . .
K CC READING TEST . . . .
. . . 99
S. . 104
. . . 107
: . . 116
. . . 120
. . . 124
)JECT . 125
S . . 138
. . * 158
. . .. 176
. . . 196
* . . 204
. . . 204
. . . 219
. . . 220
. . . 223
.* . 232
. . 236
. . . 239
* . 255
. . 258
EXT) . 260
. . 263
. . 264
. . 279
S. . 284
L SECOND POSTTEST . . . . . . .
M PHONICS DRILL SHEET . . . . . .
N SECOND READER . . . . . . . .
O BRIEF ACCOUNT OF THE COUP AND INTERVENTION
P TEST ON /DI GOT/ . . . . . . .
Q TEST ON /WAN GREN A FIG/ . . . . .
R CARRIACOU SONGBOOK (SAMPLE PAGES) . .
S THIRD READER . . . . . . . .
T FOURTH READER . . . . . . .
U THIRD POSTTEST . . . . . . .
V LIZARDS . . .
W PARALLEL PASSAGES IN EE AND CC . . .
X CASSETTE . . . . . . . . .
BIOGRAPHICAL SKETCH . . . . ... .
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
"IT HAVE MORE SOFT WORDS":
A STUDY OF CREOLE ENGLISH AND READING
IN CARRIACOU, GRENADA
Ronald F. Kephart
Chair: M.J. Hardman-de-Bautista, Ph.D
Major Department: Anthropology
Children in the officially English speaking West Indies
have two major problems in learning to read. They must
adjust to the multiple patternings and irregularities of
traditional English orthography, a hurdle they share with
all people learning to read English. They must also deal
with a grammar which differs significantly from their native
Creole English. This makes it difficult to develop an
internal model of the reading process. The result is that
few West Indian children attain true literacy in English.
Since the skill of reading is not language-specific, a
possible solution is to give children access to literacy
first through Creole English. In 1982-84 this was attempted
with a small group of twelve year olds who had failed to
learn to read competently. A phonemic orthography was used
to represent the children's speech. Reading materials were
based on stories, anecdotes, etc. contributed by the
children. The children were tested at regular intervals in
English and these results were compared with a control
The research showed that reading Creole English neither
confused nor impaired the children's reading of English, as
predicted by some educators. While it was not possible to
prove conclusively that reading Creole English helped the
target population's reading of English, the enjoyment and
enthusiasm displayed by all children in reading the Creole
materials strongly imply that West Indian children should be
allowed to read and write Creole as part of their language
arts programs. Other children and adults who were already
relatively literate were able to read the materials with no
difficulty. Finally, the reading materials were prepared at
very low cost with technology available to most schools,
refuting the claim that provision of materials in minority
languages such as Creole English is too expensive.
Children in the officially English speaking West Indies
have a double burden in learning to read. First, as do all
English speakers, they face a spelling system which was not
designed for English but rather adopted piecemeal from a
variety of sources, with the result that native Anglo Saxon
words are spelled with one system, Latinate borrowings with
a second, etc. (Pitman 1966). English speaking children
have to learn, in effect, which words of the language are
spelled in which system, and how each system operates, with
the added complication that none of the subsystems
efficiently represents the phonology or morpho-phonology of'
modern English. This is no simple matter and the time it
takes children to learn to read and the overriding concern
in the English speaking world with spelling tests as an
important part of the educational process testify to this
The second hurdle many West Indian children have.to
overcome is the fact that the language they speak does not
have the same underlying grammatical structure as the
language they are trying to learn to read. It is true that
much is shared: lexical cognates, SVO word order, and
predominance of pre-positional marking in both NP and VP.
These surface similarities mask and make it easy to minimize
underlying differences. As an example, note the following
They go in the store.
In Export English (see Preface) this sentence is an answer
to a question like "Where do they go / What do they do (on
Saturdays)" The sentence is non-past in tense and habitual
in aspect. In Carriacou Creole, the meaning is quite
different. The non-stative verb go is unmarked; therefore
aspect is punctual and tense is simple: the process has
begun but is not yet completed, and the sentence glosses as
"They went into the store and they haven't come out yet."
Quite a difference, despite the apparent equality of the
surface morphemes and syntax.
The teaching of English in the West Indies has
generally followed a deficit model rather than a contrastive
one. That is, rather than teach children the differences
between their speech and English, the usual policy is to
consider them deficient in language and in need of being
filled up with "proper" English. Even specialists in
language education are, in my experience, unaware of the
depth at which West Indians' speech differs from EE. Little
or no provision is made for making use of the children's
native language skills. Some children are adept at picking
up on the differences between their speech and EE but most
are just normal children. They don't know why they have
such a hard time learning to read, but they are very soon
(sometimes at age 5 or 6!) labelled "dunce" and left to
their own devices while teacher concentrates on the "bright"
children. As one boy in the Treatment group put it,
"/d6 giv yu a buk an d6 tel yu tray an rid it an
d6 liv yu an g6/" ('they give you a book and they
tell you to try to read it and then they go off and leave
you'). It is no coincidence that the "bright" children are,
as a group, better dressed, come to school with all the
proper tools, and arrive on time, all signs that they come
from a household which is better off economically and
probably get more exposure to EE (cf. Scott & Smitherman
It is possible to glimpse quantitatively the extent of
the reading problem in the West Indies by looking at how
students perform on English Language exams. The "Ordinary
Level" ("O-Level") exams are set at British universities
such as Cambridge or the University of London. Until
recently these exams were virtually the only criteria for
student achievement in high school. In addition, a "pass"
in English Language is a minimal requirement, like a high
school diploma in the United States, for all but the most
menial work. Craig (1969:1) states that the failure rate
for the English speaking Caribbean in general "fluctuates
between 70 per cent and 80 per cent." The actual figures
for Trinidad and Tobago appear to confirm this statement:
TABLE 1.1- "O-LEVEL" EXAM RESULTS FOR
TRINIDAD & TOBAGO 1974-78
YEAR SAT EXAM PASSED %PASSED
1974 6905 1402 20%
1975 5685 1480 26%
1976 6517 1476 23%
1977 7185 1333 19%
1978 8889 2077 23%
TOTAL 35181 7768 22%
(Republic of Trinidad & Tobago 1978)
As Craig points out, the problem is actually worse than even
these figures show, for these represent the elite, the 10
per cent or so who made it to high school and spent at least
four years there before attempting the exam, all the while
attending classes whose sole purpose is to prepare them for
the exam. So perhaps 2 or 3 per cent of the total population
passes the exams.
The explanation for this is that West Indian children
enter school at age five or so having, as their native
language, some variety of Creole English. As we shall see,
West Indian Creole English differs from the official
language, Export English, at all levels: phonology,
morphology, syntax, lexicon, sociolinguistics. These
language differences result not from a lack of knowledge of
Export English (deficit model) but rather from the existence
of a separate, fully developed, natural language system
inside the heads of the children (difference model). This
rich natural resource, which all children carry with them to
school, is at best ignored in Carriacou and other parts of
the West Indies. At worst, it is actively repressed.
Since formal schooling takes place in Export English,
children are asked to learn the language of instruction and
the content of instruction at the same time, with little or
no effective teaching of Export English as a foreign or even
quasi-foreign language. The problem is particularly acute
in the area of reading. Fluent reading is a holistic
psycholinguistic process which depends heavily on the match
between what is on the page and what is inside the heads of
readers (see Chapter IV). Because of this, the reading
process, which has been shown to be essentially the same
across languages, is more easily and efficiently acquired in
the readers' native language. The mis-match between Creole
English, such as that spoken in Carriacou, and Export
English is great enough to impair the acquisition of this
The introduction of the Marryshow Readers into the
Grenadian primary schools in 1983 was an attempt to deal
with this problem. The Readers will be discussed in greater
detail in Chapters IV and V. For now, it can be said that
the main problem with the Readers was that although they
were a radical departure from previous language arts methods
and materials, their goal was still the introduction of
Export English as soon as possible. The children's own
language was to be left behind as quickly as possible. The
feeling of educators working on this program seemed to be
that time not spent in moving toward EE was time lost.
In contrast, the goal of the present research project
was to test whether learning to read Creole English, using
materials designed not for transfer to EE but rather for
making the children feel comfortable with their own
language, would assist them in reading EE.
This goal is in direct conflict with the usual goal of
"literacy and development" projects, which is the
acquisition of mechanical and functional literacy in the
superordinate language. This reflects the fact that
literacy programs are usually designed and carried out by
the dominant class in a society and have as their unstated
but nevertheless very real goal a society in which most
people can function i.e. work in a way that supports the
dominant class without rocking the boat. I feel that this
is true whether the society in question be socialist Cuba or
the capitalist United States. There is little or no
interest here in "type three literacy," the critical
literacy, which is generally learned only by the elite in
special institutions (see Pattison 1984). This explains the
sometimes almost rabid desire to avoid people's native
language in literacy and development projects: if we allow
people to experience education through their native language
for too long, too many of them will take control of their
own self-expression and the stability of the stratified
state will be threatened.
There were three other questions which I hoped the
project might be able to address. One was whether Creole
speaking children would or would not find it difficult to
read their own language. It had often been expressed to me
that the children would have trouble learning to read a new
spelling system, and some educators expressed concern that
attempting to do so would "confuse" them. My own feeling
was that there is no good reason why people should have
trouble reading their native language, provided that the
writing system used was designed for the language, and I
wanted to collect data on this question.1
Another problem involves provision of materials. One
of the most widespread arguments against using non-official
languages in education is that the cost of providing
materials in languages spoken by relatively small numbers of
people is too great (see e.g. Bull 1964). My intention was
to show that materials could be prepared locally in the
schools, by teachers themselves, using machines such as the
Gestetner mimeograph which are generally available in
schools throughout most of the West Indies.
A third problem is the question of whether or not
children can learn the content of school subjects through
Creole. The linguistic design feature of openness (Hockett
1960) makes this possible, but at least one West Indian
educator told me that Creole was suitable only for folk
tales and songs. For reasons that will become apparent
later, I was able to investigate this question only to a
To investigate the various questions that I was
interested in, a year-long project was undertaken in the
Carriacou Junior Secondary School. This school was chosen
because I had taught there in the early 1970s and I knew the
Principal and many of the teachers. A group of 29
twelve-year-olds who, in the judgement of teachers, were in
the bottom third of their age-group was chosen as the
treatment group. For these children, part of the regular
language arts program was replaced two to three times a week
by sessions in which they were given access to reading
through their native language. An eclectic "language
experience" approach was used with some phonics instruction
but with an emphasis on reading texts. Most of the texts
were contributed by the children themselves, in the form of
personal anecdotes, folk tales, etc. These were either
taped and later transcribed, or dictated by the children in
class. A few of these texts were constructed by me with the
aid of other teachers; this was done to test their reading
of previously unknown material.
To test their reading of EE, a simple reading inventory
was devised. I did not feel it was desirable to use one of
the standardized reading tests since these tests are
designed for middle class North Americans, and comparison of
Carriacou children with such a population would be
pointless. The inventory contained three different types of
tasks: word recognition, sentence completion, and passage
comprehension. No writing was required; in each case, the
children responded by simply circling the answer of their
choice. The treatment group was pretested twice to check
reliability of the inventory, and then posttested at the end
of each 13-week school term. For each testing, question
items were chosen randomly from a bank of similar items.
Due to the small size of classes in the school, and the
nature of the school timetable, it was not possible to
obtain a control group identical in all particulars to the
treatment group. I decided to use children in the bottom
third of the 13 and 14 year age-groups as the "controls."
At the beginning of the Project there were 22 children in
the 13 year old group and 24 in the 14 year old group. This
introduced age as a possible variable, but the alternatives
were having a treatment group of only about 12 children or
trying to compare children across ability groups, both of
which seemed less likely to produce meaningful results. The
controls followed the regular language arts program, and
they were tested each time the treatment group was tested.
Performances of the treatment and control groups were
compared, using standard statistical formulas (Hatch &
Following is a summary of the Treatment and Control
groups at the start of the Project:
TABLE 1.2- PROJECT POPULATION*
AV AGE GIRLS BOYS TOTAL
(IC) 12.5 9 20 29
(2C) 13.6 12 10 22
(3C) 14.6 4 20 24
*As of December 1982
The original intention of the research project was to
involve children just entering primary school. The design
was changed at the request of the Minister of Education, who
was concerned that taking children out of the new Marryshow
Readers program would be detrimental to them. In fact, I
was specifically asked to work instead with poor readers in
the Junior Secondary school, to see whether my approach
would help them. This change in the treatment population
introduced another significant variable, namely the fact
that the children I was working with had been failures in
the school system for about six to seven years. It also
showed that, even in a revolutionary setting, children
classified as "failures" may be considered more expendable
than those who have "succeeded" in the system.
There is nothing that we as anthropologists do
that is not better done, more thoroughly, in more
specialized detail in some other department, except
integrate. What gives our studies greater value
than, say, an equivalent study by a sociologist, is
precisely that we are under the obligation to see the
area under study not just from the social point of
view, but also to be able, consciously or simply as
part of our own cognitive framework, to bring to bear
the linguistic, historical, and physical aspects that
may be relevant--and to know whether such are relevant
or not. Not, as is the wont of normal narrow
specialists, dismissing any such relevance out of
ignorance. (Hardman 1982:3)
There are, indeed, aspects of this study that might
have been done more thoroughly, or at least differently, by
specialists in education. For example, such specialists
might have preferred to use one of the standardized reading
tests and they might have insisted on a strictly matched
control group for comparison. I believe, however, that the
perspective which anthropology brings to the study more than
compensates for these possible shortcomings. Anthropology
insists on bringing holistic, cross-cultural, and
evolutionary perspectives to bear on any given problem.
These perspectives have enriched the present research in a
number of ways.
Perhaps the most glaring problem with much of the
literature on reading is its emphasis on the reading of
English. Traditional English orthography presents special
problems for people acquiring literacy--these problems will
be discussed much more fully in later sections. The point
to make here is that an anthropological linguistic
perspective allows me to freely consider Afro-American
varieties of language as we would any other previously
unwritten language. At the same time, I can bring in
findings on reading in other languages which shed light on
the problem I am trying to address.
Another problem is the influence of the
transformational-generative model of grammar on some reading
theorists. It happens that one aspect of this model, the
elimination of the phonemic level of representation, is
critical for the design of alphabets. For this reason, I
have preferred an eclectic structural approach to language.
This-approach allows me to use aspects of generative grammar
when they are appropriate, i.e. when they most economically
account for what is going on in the language without doing
violence to the language itself. But it does not commit me
to forcing the data into a TG model, or any other model. In
this approach structure is extracted from the data, not
imposed on it.
Another important perspective gained from
anthropological linguistics is the notion that all natural
human languages are equally adequate for the purposes of
education, and by extension worthy of study and respect.
This runs counter to the language-deficit model, which holds
that some languages are not "developed" or "elaborated"
enough to serve as vehicles for the concepts required in
education (see e.g. Bernstein 1964). While this model has
lost ground in recent years with the entry of
anthropological linguists into the educational sphere, it
has by no means disappeared completely (for a good example,
see Morse 1980).
A full explanation of and justification for the
spelling system I used in the Project will be presented in
later sections (see Chapters III and IV). The spelling
system is given here for convenience, so that examples of
Carriacou Creole can be read more easily. Readers are
reminded that the system is phonemic, not phonetic; i.e.
it provides sufficient information for native speakers and
others who have learned the language to reconstitute it into
The system is given in "alphabetical" order:
TABLE 1.3- PROPOSED ORTHOGRAPHY FOR CC
a f 1 p v
b g m r w
ch h n s y
d i ng sh z
e j o t zh
6 k 6 u
NASAL VOWELS: a e
As a very rough guide, the symbols /a, 6, i, 6, u/
represent phonemes corresponding approximately to those
represented in Spanish by /a, e, i, o, u/. The symbol /e/
represents a vowel like the vowel in English set. The
symbol /o/ represents a vowel something like that in the
British pronunciation of caught but slightly more open and
with lips less rounded. The symbol /j/ represents an
affricate similar to English jump. The symbol /zh/
represents a fricative like the middle consonant of English
measure. The symbol /r/ represents a retroflex, not a
flap. The other symbols have approximately their
Carriacou: The Land and People
The island of Carriacou is the largest in the
Grenadines, a series of rocks and cays that stretches
between Grenada to the south and St. Vincent to the north.
Climate is maritime tropical. Slopes are often steep, but
there is not enough land mass nor altitude to affect oceanic
weather patterns. The result is that, with virtually no
springs and no freely running water, plant, animal, and
human inhabitants of these tiny islands are dependent on
chance rainfall for the fresh water they need. Otherwise,
the islands are relatively benign: the only dangerous (land)
animals are centipedes and scorpions.
Carriacou itself has an area of about 13 square miles,
all but the steepest parts of which were cleared and planted
in export crops during the 17th, 18th, and 19th centuries.
Today much of this is second-growth forest, mostly
/pika-tri/ (Acacia farnesiana?) and other hardy species.
Some particularly vulnerable areas on the windward side are
heavily eroded and support only patches of /rachet/
(prickly-pear cactus) and other plants that can withstand
the dessicating effects of the almost ever-present northeast
The people of Carriacou are descendants, for the most
part, of slaves brought from West Africa and other parts of
the West Indies in the 17th and 18th centuries. About 1,000
people live in or around the only town, Hillsborough. The
rest live in five less densely populated areas called
"villages." Most families have some land which they use
for subsistence crops and for pasturing sheep, goats, and
cattle, which are raised for food and for sale in Grenada
and Trinidad. Until about 1975, many also raised cotton and
limes as cash crops, but the cotton and lime factories are
no longer operating. When rain falls well the volcanic soil
is very fertile, and a great deal of food in the form of
maize, peas, squash, root crops such as sweet potatoes,
cassava, yams, /dashin/, etc. is produced. This is
supplemented by fish, many species of which abound in the
Grenadines, as well as wild protein in the form of land
crabs, iguana, and /maniku/ (opossum).
Despite this abundance in good times, it is never
possible to be totally self-sufficient in Carriacou, given
the metropolitan notions of life-style and consumption that
prevail. Certain items, which are defined as necessities,
can not be produced in Carriacou and must be bought with
cash. These include salt, kerosene, rice, sugar, matches,
cooking oil, wheat flour, powdered milk, tinned meats, and
canned milk. Carriacou people, therefore, must supplement
their subsistence activities with cash-earning activities
such as driving a truck or taxi, day labouring, working on a
boat or in a shop, teaching, working for the civil service,
or perhaps raising minor cash crops such as lettuce,
watermelon, tomatoes, and pumpkin, which can be sold
locally. In addition, most depend heavily on remittances
from family members working overseas.
Since 1838, Carriacou people have relied heavily on
their ability to migrate to places both within and outside
of the West Indies to work for cash. This migration has led
them to such disparate places as New York, London, Cuba, the
Venezuelan oilfields, the Panama Canal, and Ascension
Island. Until relatively recently, most migration has been
by men, who typically worked overseas long enough to earn
the money to buy land, build a house or boat, or perhaps
open a shop. Through the years, migration has been a major
safety valve, considering the limited possibilities for wage
labour in Carriacou itself (For more details on the
ethnography of Carriacou see Smith 1962 and Hill 1977).
Education in Carriacou
Education has a long tradition in Carriacou. During
the early French plantation period this took the form of
religious instruction. The French Catholic priests were,
apparently, more concerned about the spiritual well-being of
the slaves than were their English Anglican counterparts.
They instructed the slaves in Catholicism, encouraged the
metropolitan forms of marriage and the family, and
discouraged promiscuity (Devas 1964:160). This tradition
was evidently strong enough to continue after the island
became British. A visitor in 1833 remarked on the high
incidences of marriage and stable families and reported that
a school, located in Hillsborough, was serving about 100
children (cited in Smith 1962:23-24). After Emancipation
(1838) the Catholic and Anglican churches continued to
operate schools on the island, and by 1897 the Grenada
government had a primary school there (Smith 1962:37).
Mr. Peter Benjamin, who went through primary school in
Harvey Vale during the years between 1900 and World War I,
remembers some aspects of schooling at that time:
Well, the school building is a very large house. A
very large house a large house with one room. It
haven't got no apartment. One room which inhabited all
the pupils all the children. . There wasn't no
wash-room. They only have a basin stand and a towel
where only the head teacher and the teachers generally
goes to wash their hands and so on according to their
work. . They have . two latrines. One for
male and one for female. . There was no kitchen
at that school, no cooking there.
. there was no decoration but, one picture they
used to have. The picture of the king and queen. .
The old queen [Victoria]. And Edward the Fifth, King
George the Fifth, and King George the Sixth.
You see now at that time they had a big Union Jack, the
English flag, when it hoist we know it ten o'clock. We
know when we are late, because they don't hoist until
ten o'clock. And we go, we off for lunch at one
o'clock, and call in two, two o'clock. And we dismiss
at four in the afternoon. See? So the school used to
We used to read the first book [ The Royal Readers ],
all the lessons been singing, but what I can remember a
lesson in the first book, first standard, called "One
fine day in summer, a very little boy was sent to
school by his mother." See? . Now when they have
to sing this lesson they say "o-n-e one o-n-e one." The
teacher say "next word." "F-i-n-e fine f-i-n-e fine."
You see that is the way they sing the lesson. "Day:
d-a-y day d-a-y day. In: i-n in i-n in. Summer: s-u-m
sum m-e-r mer s-u-m sum m-e-r mer summer." That is the
way. And then the teacher ask them to read "one fine
day in summer a very little boy was -
sent to school by his mother." And so on.
You see? So therefore, by singing the lesson they take
up the lesson faster.
The punishment that they got at school is for bad
behavior, and late coming to school. . They punish
by tamarind rod and strap. yes. When we go.to school
late, they always put you on a bench to stand and wait
for punishment. What we call licks. You see the strap
was given to the school . by the government you
know? But the tamarind rod was taken by the teachers
sending a pupil to get the tamarind rod. To lick
themselves! (Benjamin 1983)
The Present Situation
By the 1970s and 1980s some things had changed in
Carriacou, and some things remained the same. Primary
schools still were generally housed in one building with no
walls separating the different groups of children, resulting
in teachers having to shout their lessons as a matter of
course. Lessons in reading still tended to focus on the
mechanical aspects of literacy, with much emphasis on
spelling and correct punctuation. Reading materials were in
short supply, unless you count the castoffs from Canada and
England with cute stories about Euro-American people mowing
their lawns in summer and building snowpeople in winter.
The strap and tamarind rod were still liberally applied for
misbehavior, lateness, and not learning the lesson. One
teacher regretted that this was necessary but felt that it
was the only way to make the children perform as desired
given their "slave mentality."
One change was that the school day had shifted earlier.
Rather than starting at 10:00am as in the early 1900s,
primary schools now ran from 9:00am to 3:30pm. This gave
children less time in the morning to tend to the chores
typically done by Carriacou children before school: tending
animals, gathering firewood, hauling water, working in the
In 1982, when this study began, there were five primary
schools in Carriacou, serving the five principal population
centers. These schools took in children at age five,
sometimes at four if they had an infants' department.
Starting at age five in standard one, children progressed
through to the eighth standard, which they should have
reached by the age of twelve. At this point some students,
i.e. those who performed well on a nationally administered
exam called the "Common Entrance Exam," could enter
secondary school. There was one secondary school in
Carriacou, Bishop's College, with places for about 80
entering children each year. A limited number could also
attend secondary school in Grenada. But this still left a
sizeable number unable to get a place at a secondary school.
Children lucky enough to enter secondary school faced
five years of study in various subjects, including English
language, mathematics, history, biology, physics, general
science, religion (i.e. Christian dogma), Spanish, French,
etc. There were two chief limitations on the subjects
offered. First, could a teacher be found to teach it?
Second, was there an external exam (University of London,
Cambridge, CXC, etc.) that could be used to evaluate the
students' progress? If not, why bother, since internal
evaluation was not considered valid. In their fourth and
fifth years the students would sit the external exams, which
would be their sole evaluation for the previous four or five
years of study.
The Junior Secondary School
In an attempt to provide for those children who were
unable to get places in secondary schools, the Grenada
Government in 1971 opened up the first Junior Secondary
School in the nation's history in Carriacou. Along with
three other Peace Corps Volunteers, one Canadian, and about
twenty local teachers, I was a member of the original staff
of the school. The school was built with help from Great
Britain and Canada.
Previously, children who failed to get into secondary
school for one reason or another had the option of staying
in their primary school or quitting the system and looking
for work. The Junior Sec was envisioned as a means of
giving these children the opportunity to be with their age
mates and at the same providing them with "practical"
training in industrial arts, agriculture, home economics,
needlework, commercial arts, Spanish, and arts and crafts,
all of which were supposed to make it easier for them to
make a living.
The Junior Secondary School occupied one end of a
large, flat area of grassy land called "the Pasture"
(/pastya/) located in the middle of Hillsborough. It had
problems from the beginning, both geographical and
sociocultural. The school was placed at the extreme
southeast end of the Pasture so that the noise would not
disturb people whose houses ringed the other sides.
Unfortunately, this was the lowest part of the Pasture and
during prolonged rains it turned into a pond. Despite this,
no provision had been made in the school plans for water
catchment: some of the building walls were beginning to
settle and develop cracks by the night the school opened in
The school was conceived and built specifically to
provide "practical" education for children who could not get
into the academically oriented secondary schools. But
education in Carriacou, as in most of the Western world, was
seen as a ticket to elite status, one of the chief
characteristics of which was not having to do manual labour.
There was no well-developed concept of education for
practical life experience, the whole point of education
being to separate people from the practical side of life.
In addition, the Ministry of Education had not developed any
alternative evaluation scheme for children going through
Junior Sec. The only evaluation with any prestige attached
was still the school leaving exam, which was academically
oriented. Junior Sec children were required to take this
exam because the school was classified as a primary school
for administrative purposes. By the 1980s this led to some
of the Junior Sec children preparing for the same external
exams as the secondary school students, a clear
contradiction of the original purpose of the school.
Because recruitment to the Junior Sec was based on
negative criteria, i.e. inability to attend secondary
school, the school, pupils, and to some extent teachers as
well were held in relatively low esteem by members of the
community. The primary schools, which were all older and
played a more clearly defined role in the life of the
community, enjoyed a more favorable status and some children
preferred to remain in them, as they would have done in the
past, rather than attend the Junior Sec.
Another large problem was finding teaching materials.
Most of the materials originally given to the school were
out-dated rejects from school systems and libraries in
Canada and Great Britain. The money to buy new materials
was always short, with Government contributing only EC$0.50
(about US$0.25 in 1972) per child per year. The school was
able to add to that figure by holding a "School Fair" each
year. Several foreigners living on the island helped out
from time to time, and the Peace Corps Volunteers developed
the habit of shamelessly hitting up any tourists they
happened to run into.
In 1972 Great Britain provided the school with a
Gestetner mimeograph machine and a week long seminar in how
to use and maintain it. During the seminar, we also learned
how to design and print inexpensive booklets. My first use
for this technology was to produce my own Spanish
mini-texts, which were distributed to the children for a
small fee. Other teachers produced booklets for their math
and language arts classes, containing material directly
relevant to the children. This "guerrilla publishing"
sidesteps the issue of having to pay for expensive, usually
irrelevant, and sometimes racist educational materials
produced in the "developed" world, and I am convinced that
much more use should be made of it in places like the West
Indies. The concepts learned in this workshop led directly,
over time, to my ideas about how to provide reading
materials to children in the Literacy Project.
Children at the Junior Sec were grouped homogeneously
in two ways: by age, and by teachers' estimation of their
suitability for academic instruction. In 1982 there were
five age groups, called "forms", starting with the twelve
year old "first former" and ending with the sixteen year
old "fifth formers" Each age group was divided into
several ability groups, starting with the "bright" A's and
continuing down the alphabet. In 1971, when the school
population was around 500, there were as many as six ability
groups within some age groups; in 1982, the population was
much smaller and there no more than three.
In 1982, as in previous years, children entering the
Junior Sec from the various primary schools were given a
placement test designed by the Junior Sec teachers. This
test served to select children for the different ability
groups within the incoming age group. The top one-third
went into 1A, the middle third went into 1B, and the bottom
third became 1C.
In September 1982 the enrollment was as follows:
TABLE 1.4- JUNIOR SECONDARY
TOTAL ON ROLLS 285
Of course, the children knew perfectly well what the A,
B, and C designations meant, but there was a strong feeling
among most of the teachers that it was only possible to
teach efficiently in homogeneous groups. This mind set
pervades the entire education system of Carriacou and begins
when the children first enter primary school and begin to be
sorted for more or less attention from the teachers.
There were a total of 16 teachers at the school in
September 1982. Of these, only five had more than a year or
two of experience: the Principal, the two industrial arts
teachers, the arts and crafts teacher, and one other. One
of these was the only teacher left from the original staff
in 1971, a graphic illustration of the effect of migration
on Carriacou society. The other teachers were mostly fresh
from secondary school themselves, with little or no teacher
training, although some were participating in the
Revolutionary Government's in-service training programme
(see following section).
Timetable and Curriculum
Since fairly early in its history, the Junior Sec had
run on an abbreviated schedule starting at 8:45am and
stopping at 2:00pm. This made it possible to avoid
transportation conflicts with Bishop's College, which like
Junior Sec took in children from all over the island. The
school's timetable looked like this in 1982:
TABLE 1.5- SCHOOL TIMETABLE
8:45 9:00 Assembly
9:00 9:45 Period 1
9:45 10:30 Period 2
10:30 11:15 Period 3
11:15 11:45 Lunch Break
11:45 12:30 Period 4
12:30 1:15 Period 5
1:15 2:00 Period 6
The activity under "assembly" varied from day to day.
On Monday it was a general assembly, usually beginning with
a program led by one or two teachers followed by a talk and
announcements by the Principal. On Thursdays it was "House
Assembly" during which all children gathered in one of three
groups called houses which cut across age and status. These
houses were named for three men who played important parts
in the school's history: Fleary, the first Principal, who
died of flu on Christmas Eve 1977; Brathwaite, the Chief
Education Officer when the school was built and a
Carriacouan; and Mason, a Canadian who owned a home in
Carriacou and had helped the school in a number of ways.
Assembly on other days was used by teachers and pupils for
settling in, taking roll, etc.
Period 3 on Thursdays was set aside for "Religious
Knowledge." At this time, representatives of most of the
major religions practiced on the island came to hold what we
in North America would call "Sunday School."
One major alteration in the timetable was the
designation of Fridays as "Community School Day." On
Friday, teachers with no formal teacher training were given
classes by instructors employed by the National In-Service
Teacher Education Programme (NISTEP). This was a program
developed by the Revolutionary Government as a way of
getting training to teachers without their having to attend
Grenada Teachers College. It meant, however, that many
schools were very short-staffed on Fridays. To
counterbalance this, the Community School Day Programme
(CSDP) was born. Volunteer teachers from the community were
supposed to go to the schools on Fridays and teach the
children in whatever field of expertise they might have.
Unfortunately, enthusiasm for the program waned due to
transportation problems, lack of adequate reimbursement for
travel expenses, and misunderstandings about the real aims
of the program. The result was that Fridays tended to be
pretty chaotic, at least at the Junior Sec, since most
teachers there were in the NISTEP classes and many children
were left without direction for most of the day.
The Junior Sec was set up to offer both "academic" and
"practical" subjects to children who could not enter
secondary school. In 1982 the subjects offered were as
TABLE 1.6- SUBJECTS OFFERED 1982
Language Arts Art and Craft
Mathematics Physical Education
Health Science Electricity
General Science Typing
Agriculture Commercial Arts
Current Affairs Biology
When the school opened in 1971, an attempt was made to
offer all subjects to all children all the time. This meant
an extremely complicated timetable, in which children were
exposed to 15 or more subjects each week with often only one
period a week for some subjects. In 1972, we tried
developing a step by step curriculum by which the children
never took more than 5 or 6 subjects at a time, but were
exposed to all subjects by the time they completed their
Junior Sec career. By 1982 this format had been abandoned
and children were again taking as many as 15 subjects at the
same time, with only language arts and mathematics getting
more than two periods per week. The schedule for the 2B
class is typical:
TABLE 1.7- CLASS SCHEDULE: FORM 2B
MONDAY TUESDAY WEDNESDAY THURSDAY
Language Maths Maths Compreh
Maths Grammar Composition Geography
Geography History Reading Religion
Sewing Comm. Arts Spanish History
Agriculture Gen. Science Arts/Crafts Music
Health Sci. Phys. Ed. Ind. Arts Comm.Arts
The Treatment Group
The children that the Ministry of Education asked me to
work with entered the Junior Sec from their respective
primary schools in September 1982. They were grouped by the
Junior Sec teachers into a single form, "lC", on the basis
of their performance on the placement test which was
administered in the primary schools late in the previous
The following table summarizes the Treatment group by
sex and age as of December 1982:
TABLE 1.8- TREATMENT GROUP: AGE & SEX
BOYS GIRLS TOTAL
NUMBER 20 9 29
MEAN AGE 12.3 12.8 12.5
The children came from every region of Carriacou, as shown
in the following table (see map in Karyaku Wod Buk, Appendix
TABLE 1.9- TREATMENT GROUP: RESIDENCE
1. Hillsborough and surrounding areas..........14
2. L'Esterre, Harvey Vale, etc............... 3
3. Dover, Belvedere, etc.......................6
4. Top Hill, Mt. Royal, Belair, etc.............4
5. Grand Bay, Mt. Pleasant.....................1
6. Windward, Petit Carenage.....................1
Total ......................................... 29
In Carriacou children are commonly left with a
grandmother, aunt, or other relation while one or both
parents emigrate to find work. The following table
summarizes living arrangements for 19 children in the
Treatment group, as of July 1984:
TABLE 1.10- TREATMENT GROUP: HEADS
HEAD OF HOUSEHOLD %
BOTH PARENTS 26
Houses in Carriacou are classified in two kinds: /bod/
'frame' and /wal/ 'concrete block'. Frame houses are more
common and less likely to be wired for electricity than the
newer concrete block ones. Well over half the children in
the Treatment group lived in frame houses without
electricity, as shown in the following table:
TABLE 1.11- HOUSE TYPE AND ELECTRICITY
/BOD HOWS/ 5% 58%
/WAL HAWS/ 21% 16%
Another important consideration in Carriacou is provision
for water. As mentioned before, all water for human use is
rain water which is generally caught in the form of runoff
from the roofs of buildings. Most houses in Carriacou have
some setup for collecting rain water: a concrete cistern or
an oil drum (/kyan/). There are also a number of government
built public tanks. The following table shows how the 19
Treatment group households get their water:
TABLE 1.12- PROVISION OF FRESH WATER
PUBLIC TANK 5
Finally, the ways in which cooking is done in the
households surveyed are presented below. "Fireside"
(/faya-sayd/) implies the use of wood gathered from bush or
forest. "Coal pot" (/k61-pat/) is an iron pot used with
charcoal, which is made locally.
TABLE 1.13- COOKING METHODS
GAS STOVE 42
KEROSENE STOVE 10
COAL POT 10
On a typical school day, children in Carriacou take
care of a number of chores before going to school. These
include opening up the house2, sweeping, tending
animals, /dr6gin/ ('carrying') water, finding firewood,
washing wares, and sometimes cooking. Many walk to school,
although some live as much as three miles from Hillsborough.
Breakfast is usually some kind of hot drink or "tea" with
bread, fried bakes, rice, or left-over /kuku/.3
After the morning session at school, snack is usually a
"snow-ice" (a small bag of frozen Kool-Aid) or perhaps a
soft drink with bread and cheese; a few carry a lunch from
home, but most do not eat a meal until they return home in
the middle afternoon. From time to time, depending on the
availability of foodstuffs, lunch is prepared for children
in the school kitchen.
Back at home, there are usually more chores, such as
tending animals or perhaps weeding in the garden, and maybe
some free time to play bat or stones, or have a sea bath,
before dark begins to settle and the kerosine lamps are lit.
If there is moonlight children and adults will remain
outside, either in the yard or visiting, but if it is /dak
nayt/ 'dark night' people will be inside soon after full
dark. This is a time to read, talk quietly, perhaps write
letters or sew, before turning in for the night.
1. On the other hand, a poorly designed orthography can
create problems. Rural native speakers of Haitian Creole
are reported to have great difficulty reading with the
orthography designed by Albert Valdman (see Valdman 1970)
which is the one considered "official" oy the Haitian
2. Houses in Carriacou are usually closed up very tightly at
night to ward against night breezes, which are considered
dangerous, as well as such nighttime marauders as the
/sukuya/ and /lugaru/.
3. /Kuku/ is a baked dish made from corn meal, often with
/6kr6/ added. As a leftover, it is often fried.
ECOLOGY OF LANGUAGE IN CARRIACOU
The language used by most people in most kinds of
social interactions in Carriacou has generally been
considered a substandard variety of English, when any notice
of it is taken at all. Because so much of the lexicon is
shared with metropolitan varieties of English, it is easy
for British and North American visitors to the island to be
fooled into thinking they understand more of it than they
really do, since they are usually completely unaware of the
underlying grammatical differences. The U.S. Peace Corps,
in the early 1970s, told Volunteers headed for Carriacou
that the language was English; no special language training
was provided other than learning a few quaint words and
Carriacou people themselves refer to their language as
"Broken English" indicating that they are well aware that
their language has some relationship with English, and that
they have accepted (at least on the surface) the mythology
that the relationship is unequal: Carriacou speech is a
"corruption" of English. This attitude has been reinforced
by most visitors to the island over the years, sometimes in
very subtle ways (see Starbird 1979 for an example).
Until recently, a systematic analysis of Carriacou
speech did not exist. Don Hill, in his exemplary
ethnography of Carriacou, recognized Carriacou speech as a
"variant of West Indian Creole English" that is "nearly
sufficiently differentiated from standard English to be
classified as an independent language" (Hill 1977:195). My
own work (Kephart 1980) showed where the major differences
between Carriacou speech and English lie and how those
differences fit into the total grammar of Carriacou speech.
At that time, I labelled Carriacou speech as a "variety of
Lesser Antillean Creole English" following Hancock (1977).
Many arguments have raged over the years as to just how
the Afro-American languages should be classified. Is
Haitian Creole essentially a variety of French (Hall 1966)
or is it a variety of West African (Sylvain 1936)? Based on
the demonstration that modern creoles can be compared with
each other to construct a proto-creole, but can not be so
compared with English or French (Alleyne 1980), I would
suggest that a new classification such as "Afro-diasporan"
be used (Kephart 1984).
Mythology of Afro-American Language
Over the centuries a number of myths about the origin
of Afro-American language have been developed, sometimes for
the purpose of justifying slavery and/or the treatment of
Africans and Afro-Americans. One such myth holds that the
speech of Afro-Americans is the result of "primitive" or
even "savage" minds trying to learn a "modern" European
language and unable to do so. Sometimes this form of the
myth deals with physical characteristics as well, e.g.
Africans can't pronounce the interdental fricatives because
they have "thick tongues."
Another form of origin myth holds that creole languages
grew out of the "grunts and gestures" used by masters to
communicate with slaves on the plantations. This particular
origin myth is very persistent: I have had
university-educated native speakers of Caribbean Creole tell
me this is how their language originated. Related to this
is the notion that European masters on the plantations
deliberately simplified their speech when talking to slaves,
and that this caretaker speech or "baby talk" was what the
A third myth on the origin of Afro-American language
holds that all traces of African language and culture were
destroyed by the process of enslavement, leaving the West
Africans involved as empty vessels into which European
language and culture could be poured, no doubt with a lot of
spillage, resulting in modern-day West Indian language and
As with origins, there are a number of myths regarding
the structure of Afro-American speech which heavily
influence attitudes toward them.
One such myth holds that creole languages are
"ungrammatical" i.e. not rule-governed behavior. Although
Creole French was described in the form of a grammar as
early as 1869 by John Jacob Thomas (Thomas 1969), it has
only been fairly recently that linguists have written whole
grammars of creole languages; most often, they are concerned
with small pieces of the total grammar. But even the
grammars that have been written have often not been
accessible to the non-specialist West Indian population. As
a result, the myth persists.
As an illustration of a small piece of rule-governed
behavior in Afro-American, note the "passive" rule as I have
formulated it for Carriacou Creole:
Verb(active) --> Verb(passive) /_#
/Shi sel di fish./ 'She sold the fish'
/Di fish sel./ 'The fish have been sold'
Another structural myth about creole languages is that
they are "imprecise," i.e. people can not talk "accurately"
in Creole, as they can in "standard" English. But note the
/Shi g6 in di shap./
/Shi g6 in di shap./'She went into the store'
/Shi di-g6 in di shap./
In this case, a single sentence in English has been
translated by two different sentences in Creole. The
difference is important to Creole, but not to English: is
the person still inside the store? In the first Creole
sentence she is; in the second, she is not. Here, in the
domain of verbal categories regarding completed vs.
non-completed aspect, Creole is clearly more "precise" than
This does not exhaust the mythology about Afro-American
speech. One university-educated North American, a person
very familiar with Carriacou and the West Indies in general,
told me in 1979 that people who speak creole language all
their lives suffer from deterioration of the brain cells!
Unfortunately, people who are in positions of power over
creole speakers are nearly always holders of these
pathological illusions about the nature of human language.
The work of students of Afro-American language such as
Alleyne, Bailey, Farquhar, Bickerton, Hancock, Allsopp,
Carrington, Craig, and Rickford, has provided the raw
material for the demystification of West Indians regarding
their language. Afro-American language has been shown to be
rule-governed and fully "natural" human language. The
difficulty lies in getting the information to the West
Indian public in a non-technical form.
Origins of Afro-American Language
An Origin Model
The following synthesis owes most to Alleyne (1980)
although some factors from Bickerton (1975; 1982) also enter
In West Africa, around the slave-trading departure
areas, populations of multilingual West Africans were
becoming concentrated in the 16th and 17th centuries. These
were mostly speakers of languages belonging to the
Niger-Congo family, in particular the Mande and Akan groups
(Alleyne 1980). This population needed to be able to
interact linguistically both with the Europeans and with
themselves. To accomplish this, they took the most
accessible part of the dominating European language, namely
the lexicon, and mapped it onto the broad phonological and
morpho-syntactic patterns shared by the languages of the
Mande and Akan groups. At this point, the language was a
pidgin, i.e. there were probably no native speakers and
people continued to speak their own African languages.
After transportation to the Caribbean, members of a
single ethnolinguistic group were often separated from each
other (whether by chance or design). This resulted in
"speech communities" which were too small for the
maintenance of the native West African languages. The only
common language, in many cases, would have been the Pidgin.
The first generation of children born under these
circumstances might have learned their mothers' native
language, but for interaction with their peers they would
have had to go to the Pidgin. By the next generation the
Pidgin would have acquired native speakers, i.e. become
This model leaves open the question of whether English
and French were pidginized and then creolized separately, or
whether both were later relexifications of an earlier
Portuguese Pidgin/Creole. The important point is that the
model stresses the active part which West Africans played in
the development of Afro-American language; it was they who
took the language of the European exploiters and processed
it through the structure of their own shared systems,
thereby creating an essentially new language and culture.
The Life Cycle
The label "creole" or "creolized" language implies that
the language so labeled has passed or is passing through a
particular historical process, not all parts of which are
fully understood even today. This "life-cycle" can be
summarized as follows:
TABLE 2.1- PIDGIN-CREOLE LIFE CYCLE
Contact situation (A B C) + X -->
Pidgin X ->
Removal of (A B C) -->
Creole X ->
Decreolization (X still present) or
Stabilization of Creole X
Here A, B, and C represent native languages of the
population brought together, in this case members of the
Mande and Akan groups; X represents a politically dominant
language, such as English. "Stabilization" does not mean
that the creole will not undergo any change; the changes
will be those of natural language change, rather than
prestige attraction (decreolization).
The Post-Creole Continuum
When the language "X" is present, it is possible to
find an area of variability between the Creole X and X in
which features of X are adopted, first as surface forms with
underlying creole semantics, then with the semantics of X as
well (Bickerton 1975). The resulting set of subsystems is
known as a post-creole continuum. Members of the
post-creole speech community, depending on their exposure to
the metropolitan forms of X through formal education,
travel, etc., tend to know both the creole system and one or
more of the subsystems within the continuum.
Some writers, following Bickerton (1975), refer to X as
the acrolect and to creole as the basilect; the area within
the continuum is then called the mesolect.
The continuum for Carriacou can be illustrated with the
following, extracted from a tape of 13-year-old children
guiding each other toward creole (see Appendix X):
TABLE 2.2- POST-CREOLE CONTINUUM
X My brother and I went crab-hunting.
1. My brother and I went torching.
2. Me and my brother went torching.
Cr Me and me brother di-going and torch.
(/Mi an mi broda di-g6in an toch./)
The children were instructed to talk about whatever they
wished, but to avoid "school language" and talk as they
would with friends on the road or at home. The text
occurred while a boy was trying to tell a story about crab
hunting. He started with sentence #1 ("X" represents the EE
form, which was not produced on the tape). The others
present, after admonishing him to "speak bad," proceeded to
guide him toward the Creole end of the spectrum. The first
shift was in the subject, where "my brother and I" became
"me and me brother." The second shift was.with the
predicate, from "went torching" to "di-going and torch"
(/di-g6in an toch/). This little piece of text showed
very nicely that some children (these were high academic
students) are aware of the differences between CC and EE and
know how to move consciously from one to the other; although
of course like most native speakers of a language they can
not articulate the grammatical rules involved.
It seems clear that Carriacou is a "post-creole"
community, in that the farthest people normally go from
Export English is usually the last step above. In order to
get a more "deep" creole form, I have had to specifically
ask older people to tell me how something used to be said.
For example, everyone uses the suffix /-in/.for nonpunctual
aspect nowadays. The older form was a prefix /a-/. My
consultants tell me that /A g6in./ 'I'm going' would
have been, in earlier times, /mi a-g6./. This form is
found today, so far as I can tell, only in stories where
earlier patterns of speech are being mimicked.
The earliest inhabitants of Carriacou were immigrants
from the South American mainland: the Siboney, the Arawaks,
the Caribs, and most recently the creolized "Island Carib"
language and culture. These peoples made more cultural than
purely linguistic contributions to modern Carriacou life,
especially in the form of foods such as corn (maize),
cassava, and pumpkins. They were unable to contribute much
genetic material because, in a small place like Carriacou,
with nowhere to hide, they were especially vulnerable to the
diseases carried by the European and West African
Nevertheless, a few lexical items from Amerindian
languages have survived into modern Carriacou speech,
including (Alleyne 1973):
There are a few other items, such as /zuti/ 'nettle' and
/kongori/ millepedee' which may or not be of Amerindian
West African Languages
While the phonology and morpho-syntax of CC reflect the
underlying postulates of the African base upon which
Afro-American was built, this section is concerned mainly
with surface features. A number of lexical items of West
African origin can be found in Carriacou Creole; the
following is merely a sample:
/saraka/ 'ritual feast' (Bambara)
/juk/ 'pierce, stab' (Fula)
/b6da/ 'anus' (Bambara)
The following Asante male day-names occur as surnames:
In addition, the following are recognized by Carriacou
people as African "nations" which their ancestors belonged
to and from which they are descended (Benjamin 1982):
Kromanti Ib6 Kong6
Mok6 Manding Temn6
Bong6 Tramba Kwelbe
Bula Arada Juba
There are some "nations" which are not (apparently) of West
African origin, such as the /Bele/ which appears to be
related to a place name (Belair) in Carriacou, and
/Haldkod/, Creole French for 'pull rope'. Some of the
clearly African ones are modified, such as /Skach-Ib6/
'Scotch Ibo' and /Ib6-Grenad/ 'Grenadian Ibo'.
Knowledge of the African nations from which Carriacou
people descend is enshrined in the Big Drum or Nation dances
(see Pearse 1956; Hill 1977). Each nation has a particular
drumming pattern and accompanying songs. While most of
these songs are in Creole French (see below) a few are,
according to the people who know them, in African languages.
The following example is a /Kromanti/ nation song (Appendix
Anansi 6, e,
Anansi 6 sari baba.
The only recognizable word in this song is /Anansi/, the
trickster spider. The meaning of the rest of the song is
Some Carriacou people told me they had learned in
primary school that Bogles (/b6glz/), the name of a
village, was an African word for "hell hole." The village
is on the site of a former estate owned by Robert Bogles
(Fenner 1784). Perhaps the estate was a "hell hole" in
terms of the treatment of slaves, and the folk etymology
preserves a bit of ethnohistory.
Lesser Antillean Creole French
The Creole French spoken in Carriacou is most closely
related to that of Martinique and Guadeloupe. This reflects
the fact that the first Europeans to settle Carriacou were
French who came from these areas in the 17th century,
presumably with their West African slaves (Hill 1977:205).
By 1750, there were about 200 people living in 9 major
settlement areas on the island. About half of these were
free and half slaves (Brinkley 1978). If the model of
creole language origin posited earlier is correct, certainly
by this time Creole French must have been the principal
means of interaction, except when French people were
speaking French with each other. Creole French remained the
main language of Carriacou until the last part of the 18th
century, when the British acquired Carriacou and most of the
island appears to be owned by people with English, rather
than French, surnames (Fenner 1784).
Despite the official status of English since that time,
as late as 1900 Creole French or /Patwa/ as it is called by
Carriacouans was being learned either as the first language
or along with Creole English, especially in the "French"
areas of L'Esterre and Belmont:
At those time we didn't speaking /Patwa/. Is only the
big people, parents and so on. . But they speak
English to us, but with their people and so on, the
surrounding, they speak /Patwa/. . We took it for
ourselves. Who knows it, take it for themselves, by
learning, you know, the surrounding, we take the
/Patwa/. (Benjamin 1983)
The following is a fragment from a story in Creole
French about a cat and a turtle, as told by Mr. Peter
Benjamin (Appendix X):
i ka-al6 deye pu misye Shat, i pa-we misye Shat, i
kwiy6 misye Shat. Misye Shat mem pa-w6pon,
pas, i la su d6 misye Toti ka-manigw6 ko-i.
Misye Toti pa-kong si misye Shat la su d6-i. Le
i wive su Lantikwi, i tini 5 sakwi ta ki
pas6: tone, zekle, lapli, tut sa. M6 misye
Shat ka-manigw6 ko-i2la su d6 misye
While /Patwa/ is not learned as a native language in
Carriacou today, it is pervasive in the language and culture
of Carriacou. It apparently influenced the phonological and
morpho-syntactic patterns of Carriacou Creole English from
the time Creole English was introduced into Carriacou. The
two languages can be written with essentially the same
orthography and many syntactic constructions of CC are
morpheme-for-morpheme replacements from /Patwa/:
TABLE 2.3- CREOLE FRENCH & ENGLISH
CREOLE FORMS EE GLOSS
Patwa: /i ni/
CC: /i av/
Patwa: /fe sa ba-mw6/
'do that for me'
CC: /du dat gi-mi/
Patwa: /ka-fe sh6/
CC: /mekin hat/
Patwa: /kot6 u ye?/
'where are you?'
CC: /we yu d6?/
Patwa: /mi mwe isi-a/
'here I am'
CC: /luk mi ye/
Patwa: /kuma u y6?/
'how are you?'
CC: /how yu d6?/
Patwa: /mwe la/
CC: /a d6/
Many lexical items, including words for local plants
and animals, toponyms, and surnames, are in /Patwa/:
/tet-she/ 'tree boa'
/gwagozh6/ 'brown pelican'
/gwanambafey/ 'seeds-under-leaves (plant)'
/soley/ 'bigeye fish'
/Shap6 Kar6/ 'Chapeau Carre'(toponym)
/Mon Jalu/ 'Morne Jaloux' toponymm)
/Bid6/ 'Bedeau' (surname)
It is, however, mainly in folk and Big Drum songs that
actual texts in /Patwa/ have been preserved. Quite often,
the singers of these songs do not know the meaning of what
they are singing. The following song, a /Gwa Bele/, tells
of a vessel wrecked in an area of rough sea between
Carriacou and Grenada (Appendix X):
U pa-tan Venja kule?
Venja ney6 Kikamjeni-6!
Kapten Dezbat, ay-6!3
U pa-tan Venja kule?
Creole English was brought into Carriacou in the late
1700s with the arrival of English-speaking planters and
their African slaves. They found a society of French and
Creole French-speaking people already in place. Presumably,
the Creole English which arrived at that time would have
resembled, to some degree, the more conservative varieties
of Creole still existing in parts of Jamaica and Guyana.
Evidence for this is found in that older persons in
Carriacou, when asked how people used to speak, cite forms
such as the following:
/mi a-g6/ 'I'm going'
/a bway/ 'a boy'
Further evidence for the nature of Old Creole English in
Carriacou can be found in some of the Big Drum songs:
/yu n6-yeri?/ 'don't you hear?'
The form /yeri/ 'hear' reflects the canonical syllable shape
of early forms of Afro-American; it survives in Jamaican
Creole, though in modern CC the form is /ye/ (Alleyne
In one folk tale told by children, a man realizes that
the child he has found on the road is the child of a
/lajables/.4 The tipoff is the linguistic interchange
between the /lajables/ and the child; /lajables/ tend to
speak in Creole French or in Old Creole:
/Deziw6, we yu a-g6?/
(Desiree, where are you going?)
/D6 s6 d6 a-bring mi a dakta-6!/
(They say they're taking me to the doctor!)
Modern Creole English of Carriacou belongs to what
Alleyne (1980) calls "Intermediate Forms of Afro-American."
That is, it is closer to Export English than the more
conservative forms, such as rural Jamaican Creole. A
detailed sketch of the grammar of modern CC is provided in
Chapter Three. Following is a brief story in CC (Appendix
Wan ivnin, wen wi waz in D6va Skul, wi bin g6in
h6m man. Wi s6 wel wi g6in an red som
tambran. S6 wen wi d6 g6in, al ov os
klaym-op di tambran tri man. Neks ting, wi si dem
tichaz komin. Wi en-n6 wat tu-du boy. Wi plan wi
s6 wi wil-klaym dong di tri, an hayd in som gini
gras. Kom dong di tri an hayd in di gini gras. I ad
som oda children komin-op, s6 wi tel dem tu-hayd.
D6 neva hayd man! D6 stand-op. Wen dem tichaz
mit dem, man, d6 s6 "wat glyu duin in pipl
gyadn?"(St. Hilaire 1979)
The other language which has had a great effect on the
historical development of language in Carriacou is Export
English (see Preface). Export English is in a sense the
ultimate target at which the decreolization process is
aimed. This process has not gone as far in Carriacou as it
has in, say, Barbados, because of the shorter time period
during which English has been official and the relative
isolation of the island.
The local oral realization of Export English is what is
sometimes referred to as Standard West Indian English. This
is the language produced on formal occasions by the literate
elite. A brief sample of Standard West Indian is found in
Context of Language Use
"African" forms of language, as found in the Big Drum
songs, clearly occur only in a highly-specialized,
ritualistic context: the performance of a Big Drum dance. A
Big Drum may be held for the purpose of calling the
ancestors to a feast and asking their blessing on some human
activity, such as a boat launching, the opening of a new
house, or a funeral. They may also be held to thank the
ancestors for good fortune, either personal or communal.
Increasingly, /Patwa/ is also being confined to a
ritual context: Big Drum and folk songs in particular.
There are, however, large communities of Creole French
speakers on the islands of St. Lucia, Dominica, Martinique,
and Guadeloupe with whom Carriacou people have frequent
contact. Creole French is the native language of most
people in these areas and shows no signs of disappearing
there; thus there is good reason to suppose that it will
continue to exist, at least marginally, in Carriacou.
Whether it will ever make a comeback, as was attempted on a
modest scale under the People's Revolutionary Government
(1979-83) remains to be seen, but my impression is that many
young people are interested in their cultural heritage and
will continue to learn it from older people in the
community, while others, especially those who travel in the
Eastern Caribbean, will continue to learn it for more
Context and The Continuum
Some structural aspects of the Export <-> Creole
continuum have already been discussed (pages 7-8). In
general, all Carriacou people speak CC (Carriacou Creole
English). Many, but not all, can in addition operate within
one or more of the subsystems which connect CC, ultimately,
with Export English. The exact nature of these subsystems
as they exist in Carriacou has not yet been investigated as
thoroughly as has the system for Guyana (Bickerton 1975).
Carriacou people exploit this continuum in terms of the
context in which speech occurs. The following shows how
this can happen in a number of selected domains:
TABLE 2.4- CONTEXT OF LANGUAGE USE IN CARRIACOU
EXPORT ENGLISH CREOLE
Unequal status Equal status
Metropolitan culture Folk culture
Formal schooling Enculturation
Bible Folk religion
Radio, TV, Books, etc. Local Oral Tradition
It is important to keep in mind that this does not represent
a dichotomy of speech forms, but rather a series of
interlocking subsystems within which people operate,
depending on their ability to do so. The arrows represent a
pulling effect; people are "pulled" toward one end of the
spectrum or the other depending on the context in which they
find themselves. There are some who always operate at or
near the Creole end of the spectrum; there are others who in
effect operate at any point. Only foreigners operate only
toward the Export end of the spectrum.
As an example, teachers in school are "pulled" by the
context of formal education toward the Export end of the
continuum. They do not always operate at their maximum
distance from Creole, however. At times, they choose to
speak at the Creole end, especially when talking to children
about an emotionally loaded subject such as conditions under
slavery or sex education. One teacher calls it "going back
to roots." They may also move to Creole when it s obvious
that children have not understood the point as presented in
EE. Disciplinary language is more likely to be in Creole.
Another contextual factor influencing teachers' choice
of language in Carriacou is the relative prestige of the
school. While the five primary schools are pretty much
equal in prestige, there is a great difference between the
prestigious church-supported high school and the Junior
Secondary school. The two schools have pupils age 12 to 16,
but the Junior Sec was opened in 1971 specifically for
children who did not qualify for admission to high school.
Teachers at the high school are under much more pressure to
perform in Export English, since their students will be
taking the external exams. The Junior Sec teachers appear
more willing to accommodate since their charges are for the
most part not destined to be externally evaluated. As one
teacher puts it:
I continue using standard English but I try to
correlate it with the Creole. The same meaning I try
to give them in standard English I try to show them in
the same Creole. That mean I am doing basically two
classes, one in standard English, but one side by side
in Creole so that they would understand. This would
also help them understand the difference between the
standard English and Creole.(Adams 1984)
The following is a short sample of teacher-pupil
interaction in the classroom. The teacher is a young woman
with in-service training. The pupils are 13-14 years old
and among the top academics at the Junior Secondary School.
CC forms, whether grammatical or lexical, are underlined
when they contrast with EE forms.
T: Now suppose, this morning this child came in,
Melvin, and he didn't knock at the door he came
straight inside. And what did he, uh, what
Vaira did? What did Vaira did?
P: She complain. She tell him go back.
T: What did she do? She what?
P: She complain.
T: She complain. What did she say, can anybody
P: You just walk in the class and you en-knockin.
T: "You just walked into the class and you en-knockin."
That's what? A sentence, a question, a comment,
P: A comment.
T: She made a comment. He went outside...
P: ...and he stood up.
T: And he stood up. And what he did after?
P: He knock on the door.
T: He knocked on the door. By knocking at the door
what what what, what did he, what did you get from
that by knocking? When somebody knock at the door
what are they doing?
P: They /aksin/ to come in.
T: They /ak-/ asking to come in. And when you
asking something you are doing what?
You asking a...
In this sample of classroom speech, the only lexical
contrast is "stood up" which in CC means 'stand in one
place'. Interestingly, the children used the EE past tense
form for the CC form /stan-op/ which in this context need
not be marked for tense (see Chapter Three). The other
underlined verb phrases contrast with EE because they are
marked correctly for CC tense and aspect, but they would be
considered grammatically incorrect if judged as though they
were EE forms. The one exception is "What did Vaira did?"
in the teacher's first speech, which is a performance error
that fits neither EE nor CC, although a good case could be
made that it was caused by interference between EE and CC
The following table provides a summary of teacher and
pupil performance in this discourse sample:
TABLE 2.5- ANALYSIS OF VERB PHRASES
EE VP'S 14 U
CC VP'S 10 8
NEITHER 1 0
---- ---TOTALS 2 --5 8
TOTALS 25 8
It is worth noting that this teacher is relatively
self-conscious about language and, as the teacher of the
most "advanced" 13-year-olds, tries to speak EE as much as
possible: "If I am doing English language with my class, I
try to speak correctly so that the children would adopt that
style" (Simon 1984). Still, 40% of her verb phrases are in
CC. The data on the pupils speak for themselves.
Glottopolitics: "Nostalgia as Repression"
Language and race have been, historically, the two most
visible markers by which people in the West Indies have been
sorted for exploitation by the metropole. The culture of
the West Indies is pervaded with the mythology about these
two domains of human existence, and the parallels between
the two sets of mythologies are striking. The notion of
"standard language" as a real entity parallels that of "pure
race." Like "intelligence," these are reifications which do
not exist apart from the ideology about them, and yet vast
social, political, economic, and ideological systems have
been built on the assumption that they are in fact real
entities (Gould 1981).
While "race" can not be used openly as a criterion for
discrimination any more, "language" is so used.
Specifically, mastery of "standard English" as defined by
performance on exams sent out from the Metropole, such as
the O-level exams set in London and Cambridge, is the single
most important criterion for educational, economic, and
social advancement in the anglophone Caribbean. Until very
recently, these exams contained little or nothing of West
Indian cultural content; now that similar exams are being
set in Jamaica by the Caribbean Examinations Council, there
is more West Indian cultural focus. However, the exams
continue to stress prescriptivist "standard English"
grammatical patterns, making no allowance for the fact that
these patterns often contrast sharply with Afro-American
grammatical patterns. The exams presume that Export
English, i.e. the language of the exams, is the native
language of the population taking the exam. Sometimes, the
questions are likely to be confusing, especially to children
under the stress of the testing situation. Note the
following question from the CXC basic level English Language
sample examination paper. Students have to choose which
replacement for the underlined phrase is correct:
TABLE 2.6- SAMPLE QUESTION FROM CXC PRACTICE EXAM
Some of the great writers of which we talk today,
had no idea they would be important to us.
(b) who we talking about
(c) about which we talk
(d) about whom we talk
(Caribbean Examinations Council 1983?)
Any test-taker who happens to be thinking in Creole when
they hit this question may well choose (b), since that is in
fact the correct answer for Afro-American speech in general.
The "correct" answer from the prescriptivist viewpoint, (d),
would hardly ever be uttered, even by native speakers of
varieties of English closest to EE. This question, then,
hits Afro-American speakers doubly hard since it not only
offers them an answer which is correct according to their
speech patterns, it also tests them on a speech pattern few
native speakers of English would ever use, except under
duress. Put another way, "prescriptive grammar in our day
is nostalgia as repression" (Pattison 1984:162).
It is important to keep in mind that at no time in
their school career have the people taking this exam been
given instruction in English as a second language. It is
nearly always assumed that deviation in their speech from
Export English represents a lack of grammatical knowledge
about EE, rather than use of a native-language system.
Hence language arts in school is largely a matter of filling
an empty reservoir, a reflex of the deficit model of
Afro-American language and culture. In most of the West
Indies, little effort is made to meet children where they
are and capitalize on their already-acquired knowledge, even
when the home language and school language are as different
as Creole French and Export English.
Even well-meaning educators from the metropole often
accept the deficit mythology and regard teaching children
about their own language as likely to "confuse" them,
despite the fact that native language patterns continuously
turn up in attempts to write or speak Export English. These
people, in my opinion, underestimate the intelligence of the
children they claim to be so concerned about. My experience
in Carriacou has been that even specialists in language arts
usually do not know anything about the linguistic system of
Afro-American speech. A visiting teacher from Great Britain
was genuinely surprised to find that "They go in the shop"
as spoken by Carriacouan high school students means that the
party is still inside and is not simply an "ungrammatical"
utterance. The same teacher was also surprised to find
children in my project not "confused" by learning to read CC
in a phonemic orthography.
By denying children access to knowledge about and
respect for their native language and culture, West Indian
educational systems deprive them of the psychological base
they need to build the rest of their educational experience.
Hence they tend to be limited to mechanical and functional
literacy. This is beneficial to the metropole in at least
two ways. First, a population is maintained which can read
and write just well enough to sell their labor at cheaper
than the going rate in the metropole. Second, a society is
maintained which looks to the metropole for linguistic and
cultural norms at all levels. In this way (among others)
the exploitation of the former colonies by the metropole
1. Anthropology graduate student Kofi Akwabe Ameyaw, an
Asante speaker from Ghana, listened to the tape and was able
to identify only /anansi/.
2. "He looked for Mr. Cat, he didn't see Mr. Cat, he called
Mr. Cat. Mr. Cat did not answer, because he was maintaining
himself on Mr. Turtle's back. Mr. Turtle didn't know that
Mr. Cat was on his back. When he arrived over the Atlantic,
there was a terrible weather passing over: thunder,
lightning, rain, everything. But Mr. Cat kept himself right
there on Mr. Turtle's back." (Told by Mr. Peter Benjamin in
3. Haven't you heard Venture has sunk?
Venture drowned in Kikamjeni!
Oh, Captain Desbart,
Haven't you heard Venture has sunk?
4. The /lajables/ is a beautiful woman always dressed in a
long dress to hide the fact that she has one human foot and
one cow's foot. She plays many tricks on people, such as
riding the mailboat and paying for her passage in money
which later turns into /kaka pul/ 'chicken manure' or
/g6t mes/ 'goat shit'. People who talk to her are led
away into the bush, where their minds are addled. She can
be frightened off by the light from a car or cigarette.
5. "One afternoon, when we were attending Dover School, we
were on the way home. We decided to steal some tamarind.
So while we were on the way, we all climbed the tambran
tree. Then, we saw.those teachers approaching. We didn't
know what to do. We decided we would climb down the tree
and hide in some guinea grass. Come out of the tree and
hide in the guinea grass. There were some other children
approaching, so we told them to hide. They didn't hide at
all! They stayed right there! When the teachers got to
them, they said 'what are you doing in these people's
garden?'" (Story told by Maria St. Hilaire in 1979)
SYSTEM AND CONTRAST
In previous sections, the claim has been made that
children in Carriacou speak a language which has been
labeled Carriacou Creole English or CC, which is
fundamentally different from Export English in its
underlying grammar. The purpose of this section is to
outline, as concisely as possible and in one place, the most
important patterns of CC grammar to provide a clearer idea
of the basis for the claim of underlying difference. It
should be kept in mind that this is intended as a sketch
only, and not to be confused with a complete grammar of the
Because the orthography used in the Project reflects as
closely as possible the phonological system of CC, some
discussion of how the orthography was designed will be
offered. At the end of the chapter, some comments on the
contrast between CC and EE will be given, especially in
reference to reading and writing.
CC has a basic vowel system consisting of seven vowel
phonemes as follows:
In addition, there are three glides, two front and one back:
ay 3Y 3W
Because of the close historical connection between CC
and Creole French (/Patwa/) there are many lexical items
shared by the two languages. CC speakers, even those who do
not speak /Patwa/, tend to retain the /Patwa/ pronunciation
of these items, some of which contain contrasts not found in
CC, in particular the nasalized vowels and a front glide
/y/. Therefore, a complete phonological description of CC
must include these:
The CC consonant phonemes are as follows:
p t k
b d 9 g
f s s h
The problem of writing down a creole language is not,
from a linguistic viewpoint, different from that of
providing a writing system for any previously unwritten
language. The complications arise from the fact that most
of the world is controlled by a superliterate elite with a
heavy emotional investment in whichever metropolitan or
export language happens to be the "official" language of the
area in question. This emotion is invested not just in the
language itself, but in the writing system as well.
Compounding this is the deficit mythology which generally
surrounds both creole and non-creole subordinate languages.
The end result is a tendency to metropolitanize the folk
language in writing, usually distorting it in the process
(see Layme 1980 for a discussion of the Aymara situation in
officially Spanish Bolivia; see also Valdman 1970 on the
struggle for an orthography for Haitian Creole).
A number of linguists have directly addressed the issue
of providing orthographies for unwritten languages (see
especially Pike 1976; Hall 1966; Hardman and Hamano 1984).
A common theme in these discussions is the need to balance
the practical limitations of typewriters and type fonts
available in the area with the desirability of a one phoneme
= one grapheme correspondence. Further discussion of the
advantage of phonemic orthographies will be given in Chapter
IV. For now, it is sufficient to state that a phonemic
approach was chosen for the Project.
In using a phonemic approach, the burning questions are
what symbols to use and how to assign phonological values to
them. The IPA symbols, which are recognized
internationally, are impractical for everyday use. Symbols
can be invented, as was done for UNIFON (Culkin 1981), but
this leaves open the question of ease of transfer to the
official language. For the Carriacou Project an
ethnophonemic system was used (Hall 1966:41) in which so far
as possible only symbols found on a standard English
typewriter were used. At the same time, an attempt was made
to assign values to the symbols which corresponded, when
such correspondence existed, to approximate values in EE, in
order to assist in transfer. However, I did not feel that
transfer was worth distorting the structure of CC,
especially as transfer of process rather than detail was the
goal (see Chapter IV).
It was most difficult to handle the vowels both because
they represent an extreme contrast with the EE system and
because there are only five vowel symbols on the standard
English typewriter. The problem is in the contrast between
mid and mid-low vowels, both front and back. I tried /ei,
ie, ee, ey/ for the phoneme /e/ and /ou, uo, oo, ow/ for
/o/. None of these was satisfactory, because all imply a
glide when what is represented is a simple vowel. I finally
settled on the acute accent to mark the higher member of
each pair of mid vowels. Of course, most typewriters do not
carry an acute accent (lingocentrism is inescapable) but
later, with the help of the children, I found that a bar
(raised hyphen) was just as good. This gives the following
alternative vowel representations:
TABLE 3.1- CC VOWELS: TWO SPELLINGS
FORMAT A FORMAT B
i u i u
6 6 e o
e o e o
For ease in printing, Format A will be used throughout this
dissertation, except when copies of materials produced with
Format B are being illustrated.
The glides appear as:
TABLE 3.2- CC GLIDES
ay oy ow
The phonemes from /Patwa/ can be written as follows.
Note that although few English typewriters have the tilde it
is rarely needed.
TABLE 3.3- /PATWA/ VOWELS
SHARED WITH CC
Some examples of words in which these occur are:
/tetshe/ 'tree boa'
/soley/ 'bigeye fish'
The consonants are treated in CC orthography as:
TABLE 3.4- SPELLING FOR CC CONSONANTS
p t ch k
b d j g
f s sh h
v z zh
m n ng
The only "new" symbol is /zh/; all others share at least
their (approximate) CC values with EE, although most have
additional values in EE as well.
Note that the ethnophonemic system, while based on the
structure of the language itself, uses symbols found in the
children's environment, with about 85% transferability to
EE. The vowels transfer less than the consonants, but they
are within the mainstream European and African orthographic
traditions, which is more than can be said for traditional
Allophonic variation, i.e. non-contrastive variation in
the realization of phonemes, is relatively limited in CC.
Nevertheless, some of the places where it occurs are
interesting from a contrastive standpoint.
One way in which CC differs from most other varieties
of Afro-American is in not having a short-long or tense-lax
contrast in the area of the vowels /i a u/. Thus, in CC,
the lexical pairs ship-sheep, shit-sheet, slip-sleep, and
full-fool are homophones. Likewise there is no front-back
contrast in the region of /a/, so that and-on is also a
homophonous pair. The reason for this appears to lie partly
in the fact that Creole English as brought into Carriacou in
the late 18th Century was influenced by Creole French, which
was already present, rather than by metropolitan varieties
of English as was the case in Jamaica and Barbados.
There is some variation in the areas of /i u/. The
variation which has been observed appears to be governed
partly by environment:
/i,u/ --> [I,U] /_C
However, not all such variation is conditioned by
phonological environment, since the same speaker may realize
'shit' as [sIt] or [sit], sometimes in the same sentence.
The major allophonic variation in consonants occurs
with /t/ and /d/ as follows:
/t,d/ --> [8,5] / r
This rule applies across word boundaries, e.g.:
/bat/ 'bath' + /rum/ 'room' --> [bacrum] 'bathroom'
A striking feature of CC phonology is the effect
/g/ and /6/ have on preceding stops. /6/ is
prepalatalized, while /6/ is prerounded; in some
speakers the result is almost glidelike:
/k[k/ [kyek] 'cake'
/tep/ [tyep] 'tape player'
/k6t/ [kwot] 'coat'
/b6t/ [bwot] 'boat'
One phoneme which exhibits considerable variation in CC
is /h/. Most speakers have it clearly in contrasts such as:
/han/ 'arm' vs. /an/ 'and'
/h61/ 'hole' vs. /61/ 'old'
/h/ is often realized, however, as a glottal stop ([?]) or
as 0; the exact conditioning factors are not known to me
at this time. Hence a word such as /han/ 'arm' may have the
following phonetic realizations:
/han/ --> [han]
Further complicating the issue is the fact that /h/ may
surface in vowel-initial words which are not "supposed" to
have it, such as /heg/ for 'egg'. People in Carriacou claim
that this is a feature of people from certain parts of the
island; it is labeled "bad" speech.
In materials for the children I spelled a word with /h/
if some people regularly realize either [h] or [?] in that
word. Thus most CC words that are cognate with an EE word
that has /h/ are also spelled with /h/. The major exception
is /av/ 'have' which never has initial /h/.
Syllables and Timing
The preferred syllable shape is CV, a general feature
of all varieties of Afro-American language. All the vowels,
glides, and nasal vowels, as well as the consonants /1/ and
/n/, may occur as syllable nuclei. The CV shape asserts
itself across morpheme boundaries, e.g.:
/if a kof im/ --> /i-fa-ko-fim/
'if I hit him'
Because vowels are not reduced in unstressed position, CC is
a syllable timed language, like Spanish or French, rather
than stress timed like English. Again, this is a feature of
most varieties of Afro-American and results in the
perception, for English speakers, that West Indians speak in
a "sing-songy" way.
The pitch contours of CC are very similar to those of
Creole French. High pitch marks stress; low pitch is
non-stress. Certain elements within a sentence, such as the
nucleus of a noun or verb phrase, must carry high pitch.
Other emelents, such as the preclitics for aspect, are
nearly always unstressed.
In a noun or noun phrase of more than one syllable,
stress tends to fall on the last syllable unless the nucleus
of the last syllable is /1/, which cannot be stressed..
Note the following (stressed sylllable is in CAPS):
Morpho-Syntax: Noun Phrase
Structure of Noun Phrase
NP --> (Prenucleic)Nucleus
Nucleus -> Noun(suffix)
Prenucleic --> Determiner
Unmarked nouns are generic, that is all members of the
class are represented:
/fig/ 'bananas (in general)'
Determiners mark a noun as non-generic. Indefinite and
definite articles and demonstratives may occur in the
TABLE 3.5- CC ARTICLES & DEMONSTRATIVES
INDEFINITE a som
NEAR dis diz
FAR dat dem
/a fig/ 'a banana'
/di uman/ 'the (specific) woman'
/dis maniku/ 'this opossum'
/dem kyatl/ 'those cows'
There can be only one determiner per NP. Since the person
markers cannot co-occur with other determiners, they are
classed with them. When person markers occur in the
determiner slot, they mark possession.
TABLE 3.6- PERSON MARKERS AS DETERMINER
FIRST mi wi
SECOND yu alyu
THIRD i, shi d6
Examples of the use of person markers as determiners:
/mi fig/ 'my banana'
/wi fig 'our banana'
/yu fig/ 'your banana'
/alyu fig/ 'your (pl) banana'
/shi fig/ 'her banana'
Plural is an optional category in CC; it need not be
marked inflectionally (i.e. by a suffix) if it is already
marked syntactically by a plural determiner or a quantifier.
Note the following:
/wan kyat/ 'one cat'
/dem kyat/ 'those cats'
/tu fowl/ 'two chickens'
/plenti ship/ 'lots of sheep'
The plural suffix /-andem/ carries the meaning 'the
thing marked plus other things which may or may not be the
same'. It may occur with both human and non-human nouns:
/wi ting-andem/ 'our things'
/Kwamina-andem/ 'K plus others'
The EE plural suffix I-S] also occurs in CC. I have no
evidence that it covers a different semantic range from
/-andem/, and the two can co-occur:
The morphophonemics of [-SI are the same as in EE:
The suffix /-6n/ marks a noun as the possessor or
owner of something. It is obligatory for possessives in an
exposed position, i.e. when the possessed noun or NP is not
/dat iz Shoma-6n/
'that's Sherma's (something)'
Complex Noun Phrase
Complex NP's are formed when more than one noun occur
in the nucleus slot. There are two types of complex NP.
The first, and most productive, consists of noun + noun. In
this structure the first noun describes, limits, specifies,
or otherwise modifies the second noun:
/tambran tri/ 'tamarind tree'
/wata not/ 'green coconut'
/si kyat/ 'octopus'
/wal hows/ 'block house'
/dakta bod/ 'hummingbird'
/Kwamina pen/ 'K's pen'
/kow bres/ 'cow teat'
Notice that possession, which forms a separate structural
category in EE, is included in this CC structure.
The other complex NP structure takes the form noun +
/a/ + noun. The particle /a/ relates the second noun to the
/wan grin a fig/ 'a finger of banana'
/tu set a children/ 'two groups of kids'
Noun Phrase Markers
These precede the NP and mark it for location,
direction, instrumentality, goal, or separation. The most
common NP markers are illustrated below:
/in Karyaku/ intoo Carriacou'
/bay Kwamina/ 'at/to K's place'
/antap di tri/ 'up in the tree'
/wit r6p/ 'with rope'
/fo mi broda/ 'for my brother'
/ . gi mi/ '. . for me'
/sens monin/ 'since morning'
Some other NP markers are: /insayd/ 'inside of'; /from/
'from'; /bil6/ 'under'; /an/ 'on'; /bihayn/ 'in back
Morpho-Syntax: Verb Phrase
The person markers which occur as determiners in the NP
have already been presented. There are two other syntactic
slots in which the person markers may occur: a preverb
("subject") slot and a postverb ("object") slot, which
includes the slot after NP markers. When they occur in
these other slots, the person markers show syntactically
conditioned allomorphs, only one of which is obligatory.
The following table summarizes the person markers in all
TABLE 3.7- CC PERSON MARKERS
SUBJECT DETERMINER OBJECT
wi (alwi) (os)
There are two fundamentally different sentence types in
CC. One is the equational in which two NP's are tied
together by a copula. The other is predicational which
contains a subject NP and a predicate VP. The two types can
be formulated as follows:
S(Equat) --> NP Cop NP
S(Pred) -> NPVP
In this sentence type, two NP's are connected by a
copula. In modern CC, the universal copula is /iz/:
/mi fada iz a fishaman/
'my father is/was a fisherman'
/al a wi iz wan/
'we are all one people'
There is no obligatory marking for tense; however,
equationals may be marked past by /waz/:
/dat waz a lajables chayl/
'that was a witch's child'
Interestingly, there is no way to mark an equational
"future." Instead, a process verb such as /ton/ 'become'
which can take future marking must be used:
/shi g6-ton ticha/
'she will be a teacher'
The copula /iz/ contracts to /i/ in rapid speech. This
contraction is interesting because it runs counter to EE in
which the vowel rather than the consonant of the copula is
deleted. But it parallels /Patwa/ in which the full copula
/se/ contracts to /6/:
EE: That's the banana.
CC: /da i di fig/
Patwa: /sa 6 fig-la/
The explanation for this probably has to do with the fact
that EE is stress timed and reduces or loses vowels in rapid
speech, while both CC and /Patwa/ are syllable timed and
more likely to reduce or lose consonants.
These sentences, unlike equationals, are marked for
tense, mode, and aspect. They contain a verb, which is
defined as the minimal occupier of the nucleus slot in the
S(pred) --> NPVP
VP -> (Preclitic)Nucleus(Compl)
Nucleus --> Verb(Suffix)
Preclitic --> Aspect
The co-occurence of elements of the VP is summarized in the
TABLE 3.8- CO-OCCURENCE CONSTRAINTS
TMA NEG VERB SUFFIX
ASPECT +/- + +/-
TENSE +/- + +/-
MODE +/- +
---------------------------------------------~~ ~ Z Z 9 Z
All CC verbs can be classified into two cryptotypes
(Whorf 1979) depending on whether or not they can occur with
the non-punctual aspect suffix /-in/. Those which cannot
occur with /-in/ are called statives; those which can take
/-in/ are non-statives (cf. Bickerton 1982). Statives are
inherently non-punctual in aspect:
/i d6 in di ka/ 'it's in the car'
/a veks/ 'I'm angry'
/yu n6 dat/ 'you know that'
Non-statives are inherently punctual in aspect:
/Kwamina skr6p fish/ 'K cleaned fish'
/Kwamina g6 in tong/ 'K went to town'
Punctuality and non-punctuality in CC refer to whether
an action, state, or process is viewed as a point in time
(punctual) or as spread out in time (non-punctual).
Non-stative verbs, which are inherently punctual, can be
marked non-punctual; the reverse is not true, as shown in
the following table:
TABLE 3.9- ASPECT MARKING OF VERBS IN CC
Examples of use of /-in/:
/Kwamina skr6pin fish/ 'K is/was cleaning fish'
/Kwamina g6in in tong/ 'K is/was going to town'
Apparently the suffix /-in/, which is clearly a
borrowing from EE, has replaced an earlier prefix /a-/ or
/da-/ which is attested now only in folk tales or examples
of the way people in Carriacou used to talk (see Chapter
At one point in the history of CC, all TMA markers were
preclitics. At some point, the earlier non-punctual markers
/a-/, /da-/ were replaced by the suffix /-in/, which is
clearly a borrowing from EE. The remaining preclitics
superficially resemble EE auxilliaries, but function very
differently. They can not occur in exposed position, and
with two exceptions they are all unstressed.
Iterative /doz-/ All verbs may be marked iterative,
i.e. the action, state, or process is viewed as occurring
regularly or habitually. With attributive verbs., /doz-/ -->
/a doz-d6 al di taym/ 'I'm regularly present'
/shi doz-g6 in tong/ 'she goes to town
/ayskrim dozbi-nays/ 'ice cream is tasty'
Iterative and non-punctual aspect marking may co-occur, as
in the following:
/Kwamina dozbi-skrepin fish/
'K is regularly engaged in cleaning fish'
The anterior iterative marker /yustu-/, presumably from
EE "used to", occurs in many people's speech although it is
not obligatory: /doz-/ and /dozbi-/ may also refer to past
/d6 yustu-sel dem/ 'they used to sell them'