Musah 11 Aspects of Kusaa
Musah 11 Aspects of Kusaa
Musah 11 Aspects of Kusaa
BY
JULY, 2010
ii
DECLARATION
I certify that except for references to other works which have been duly
(CANDIDATE)
(SUPERVISOR)
(SUPERVISOR)
iii
ABSTRACT
In recent times, there has arisen an overbearing need to describe, analyse and
document the various languages of the world. This research investigates
aspects of the grammar of Kusaal – a less familiar, un-described and under-
studied Mabia (Gur) language spoken in the north-eastern corner of Ghana by
some 450,000 people. The study entails a description of the phonology of the
language and is concerned with uncovering the underlying sounds, the syllable
structures, the phonological processes and the tonal patterns inherent in Kusaal.
It is cast within the generative phonological traditions especially formalised by
Chomsky and Halle (1968), on Goldsmith‟s (1979, 1990) autosegmental
phonology and on moraic theory. Data for the research was collected with an
electronic recorder from twelve respondents on a field trip to Bawku and
Zebilla. The data was then used in corroboration with other sources found in a
few study materials. Native speaker intuition and the Praat acoustics software
were important resources in the transcription, description and analyses of the
data. The study reveals that Kusaal has twenty-four consonants, nine short
vowels which have long correlates, five nasal vowels which are all [-ATR] and
several sequences of vowels. The study also shows that the various syllable
structures of the language are collapsible into four predominant types: the peak
only, the VC, CV and the CVC. Kusaal is morphologically isolating but words
can take up to four syllables of different structures and types in the language.
Two phonological processes are identified in the language: assimilation and
syllable structure processes. In the former, nasalisation, homorganic nasal
assimilation, labialisation, palatalisation and vowel harmony are prevalent
while the latter manifests in syllable deletion and truncation, aspiration,
glottalisation and loanword re-syllabification. Kusaal is a register tone
language which distinguishes three level tones – high, mid and low. It also
distinguishes a downstepped high from an underlying high tone. The study
identifies the mora as the tone bearing unit and discusses the role of tone in the
morphosyntax of the language.
iv
DEDICATION
To my mother,
and my father,
ACKNOWLEDGMENT
I have been truly blessed! My cup now overflows and my gratitude is endless.
Blessed be the name of the Lord God Almighty. His will be done.
To my parents for blessing me with life and education and for believing in
what I chose to do; for being such powerful beacons of hope and inspiration
and especially to my mum for bracing the storm and bringing me this far even
after the Old Boy passed on. To Bapiam, Gladys, Kiran-May, Mati and Clare
for all the sacrifices you‟ve always had to make for me. You are the greatest!
for their many insightful comments, painstaking editing and prompt delivery
that have greatly directed the course of this work and made it what it is today.
great wealth of resources at my disposal. But for his support and magnanimity,
my progress would certainly have stalled. May the Good Lord never tire in
of Linguistics under whose wings I have grown and am now poised to soar. To
Profs Duthie, Osam, Anyidoho, and Chris Collins and to Drs Saah, Amfo,
My auntie Mrs Aboase and her family at Madina also deserve my unrestrained
gratefulness for their kindness and hospitality whenever I got back to Accra
vi
from Bawku. You have always given me a real home away from home. While
of the pressure that often came with this work. I can never say thank you
Sarah, Joan, Akua, Jessie, Francis, Mmra, Xevi-Cole, Shehu,Suraj and Umar,
in reminiscence of all the up times and down times. We have overcome! It was
To all the people who are constantly making tremendous inputs into my life, I
have cause to be thankful. To Bador, Ike, Asangba, Agnesa and Mba, Linus
endeavour, and to all those I have inadvertently left out, I say the battle is won
Agoswin A. Musah
TABLE OF CONTENTS
TABLE OF CONTENTS
BIBLIOGRAPHY … … … … … … … 148
xi
LIST OF MAPS/FIGURES
CHAPTER ONE
Fig. 1.1 - Map of Upper East Region showing the three Kusaal
speaking districts: Bawku West, Bawku Municipal and
Garu-Tempane … … … … … … 5
Fig. 1.2 - Genetic affiliation of Kusaal … … … … … 7
LIST OF TABLES
CHAPTER TWO
Table 1 - Kusaal consonants … … … … … … 31
Table 2 - Distribution of consonants … ... … … … 35
Table 3 - Kusaal vowels … … … … … … 37
Table 4 - Feature matrix of Kusaal consonant sounds ... ... 51
Table 5 - Feature matrix of Kusaal vowel sounds ... ... ... 52
CHAPTER THREE
Table 1 – Aspirated and un-aspirated sounds… … … … 98
xii
LIST OF ABBREVIATIONS
1 First person
2 Second person
3 Third person
A. Agole
T. Tonde
DEF Definite
DET Determiner
FUT Future
H High
L Low
M Mid
HAB Habitual
HTL High tone lowering
LTR Low tone rising
LTS Low tone spread
MAX. P Maximum Pitch
NEG Negative marker
PERF Perfective
PL Plural
POSS Possessive
PROG Progressive
PST Past Time
HOD Hodiernal
REC Recent
REM Remote
SG Singular
SUB Subject
TAM Tense, Aspect and Mood
UR Underlying representation
.n noun
.v verb
1
CHAPTER ONE
1.1 INTRODUCTION
This section of the study sets the tone for the thesis. It makes the introductory
statements about the subject and area of study including the motivations for
undertaking the research. It includes brief write-ups on the language and its
linguistic facts as well as the methodology that is employed for the study. The
In recent times, there has arisen an imperative need to document the various
languages of the world. Linguists, evangelists and archivists the world over,
have stressed the importance and benefits that come with a detailed description
people. This study is therefore a modest effort at adding to the meagre stock of
linguistic data on the Kusaal language which is one of the many under-studied
so our focus will be on describing and analysing the underlying sound systems
Kusaal, as intimated earlier, is one of the many under described and under
studied languages in Ghana. Though some research has been carried out on the
language, most of the works are very old and date as far back as 1957
(Melancon), 1968 and 1972 (Spratt and Spratt). Even these, considering the
time lapse, have become quite outdated in relation to the current state of affairs.
The consequences of this acute shortage of literature on the language are low
school enrolment and literacy because adults do not see the importance of
Even with the change in the educational language policy which allows for the
years of primary education, schools in the Bawku area can not take advantage
1.4 DEMOGRAPHY
The expanse of land within which Kusaal is spoken is called Kusaog meaning
3
„the country/land of the Kusaas‟. The Kusaas are found mainly in the north-
eastern corner of Ghana. As a result, the language users share boundaries to the
North with the Republic of Burkina Faso, to the South with the area near the
to the East by the Republic of Togo and to the West by the Red Volta after
which follows the land of the Talensis and Nabdams. There are therefore
people, especially on the outskirts of the major towns, are conically shaped and
built with mud. The roofs are mostly made with thatch not only because it is
the readily available material but also to regulate room temperature as the sun
Three districts of the Upper East Region are inhabited by the Kusaas. These are
the Garu-Tempane and Bawku-West Districts and the Bawku Municipality (see
Fig. 1.1 below). The three districts are further divided into five constituencies
country.
Incidentally, Bawku is the only town in all of Ghana which shares borders with
two countries at the same time; these being Burkina Faso to the north and the
Republic of Togo to the east. Its strategic positioning makes it a very important
4
location to the indigenes, businesses and the government alike. It is thus one of
the very viable commercial areas in the country and one of the biggest
Unfortunately however, the town has often received a lot of bad media as it has
been the scene of sporadic skirmishes and civil unrests between the indigenous
Kusaas and migrant Mamprusis over issues pertaining to land and politics and
Aside these two ethnic groups, there are large numbers of other migrant
Dagaaba, Bulsa, Bimobas, Yorubas, Igbos and several others. Melancon (1957:
4) states that Kusaal is mutually intelligible with Mampruli and further, that it
is not difficult for the Kusaal speaker to communicate quite easily with
that linear order. At the extreme end of this cline are Bissa, Buli and Kasem
which are languages within close proximity of Kusaal but with which
1.4.1 Ecology
The areas inhabited by the Kusaas are generally low lying savannah plains with
intermittent ranges of mountains that are potential sites for tourism especially
pockets of trees and vast grasslands. There are also a number of important
forest reserves including the Tiile, Binduri and the Morago West reserves
5
which stretch for miles on end. In the general Kusaal speaking area, two
important rivers provide drainage: the White and Black Voltas. These rivers are
often sourced by the rains which come between May and October of every
year. Some of the men who live near the water bodies are therefore able to
majority of the people. These farmers depend a lot on the rains and produce a
variety of crops including maize and millet which are the staple food of the
people. Large quantities of onions are also produced in the dry season through
(www.ghanadistricts.com)
Fig. 1.1 - Map of Upper East Region showing the three Kusaal speaking
districts: Bawku West, Bawku Municipal and Garu-Tempane1
The name of the language is Kusaal, also called Kusaalle, and the speakers
refer to themselves as the Kusaas (PL) or Kusaa (SG). When other language
speakers make reference to the native Kusaal speaker, terms like „Kusasi‟ (SG)
450,000. There are two dialects: Agole which means „up‟ or „high‟ and Tonde
which means „front‟ or „ahead‟. The former is the more dominant variety. It is
Kusaal is found under the Mabia cluster of the larger Niger-Congo language
family which is spoken in most of the north of Ghana. Hall (1983) states that
the languages of the north have been assigned labels such as „Gur‟ and „langues
because of their relative location near the Volta rivers respectively. A third
label „Moore (Mole) – Dagbani‟ has gradually lost its appeal because of its
underlies a „sisterly‟ relation among all these languages. This study will
Kusaal is also captured under the Proto Oti-Volta sub-grouping of the Niger-
Congo family. In the literature it is classified under the Western Oti-Volta sub-
grouping (cf. Naden 1988, 1989, Williamson and Blench 2000). The figure 1.2
below puts the genetic affiliation of the language into greater perspective.
NIGER CONGO
Atlantic Congo
Volta Congo
Central Mabia
Northern Southern
The Kusaas, along much the same lines as other Ghanaians, are generally a
very religious group of people who believe in the existence of a supreme deity
called Wina’am or Win meaning God. There are also several smaller gods and
totems who are representative of the various clans and communities. While
there are large numbers of Christians and Muslims, the majority are traditional
often with the aid of a tendaan (sing.) „the land owner‟. He (the tendaan), acts
as the custodian of both the spiritual and physical needs of the people within
his immediate domain (usually a small community). Kusaog (the land of the
Kusaas) is further divided into twenty (22) traditional areas each headed by a
sub-chief (Na‟ab) who is the custodian of all the traditional systems of the
people as well as the first point of legal justice and redress. Together with the
tendaannam (plu.) and all their subjects, the sub-chiefs pay allegiance to the
majority are able to use some other language apart from Kusaal. In view of the
fact that the language is mutually intelligible with a host of others, inter ethnic
while at the same time the speakers of Kusaal may not discourage their wards
from using other languages such as Bissa, Hausa, Moore or English, they hold
a very strong affinity to the language, and are in fact, very proud of it.
9
Considering that the language is actively used across a wide range of domains -
from religion through commerce and business to the daily social interactive
stand the test of time. It will be essential in transmitting and preserving the
There have been several studies on some of the related languages in the Mabia
study of about fourteen languages in the group. His work also includes an
The German missionary and linguist, Westermann, in his study with Bryan
(1952: 65) writes a line or two on Kusaal. In their study, however, they classify
similarity they bear to the language (Dagbani). Though this may appear to be
the case, there is indeed a lot of information to prove that these „dialects‟ in
spite of their similarity to Dagbani and to each other, are actually separate
languages on their own albeit with a common ancestor (cf. Rattray 1932 and
Swadesh et al. (1966) attempts to trace the time depth since the various
(Mabia) languages diverged from each other with evidence from a range of
10
words while Naden (1988, 1989) establishes the various languages and their
Work has also been carried out by native speakers on cognate languages such
perspective. It studies the morphology and syntax of the Gurun verb and verb
phrase and also highlights a number of important facts on the phonology and
On Dagaare, Bodomo (1993 and 1997) are also very representative. In the
former, serialisation of verbs as obtains in the language and its near relations
(including Kusaal) is discussed. It is also the work which suggests the name
„Mabia‟ for the cluster of languages which, hitherto, had been called „Gur‟.
feature „tone‟ and the various syllable structures of the language. Some
phonological processes are also discussed in the study. It also describes aspects
number of important facts about the underlying sound systems of the language
and captures some of the recurring syllable structures together with some of the
11
phonological and syllable structure processes that are very important in the
language. The study also discusses the tonal patterns of words in Buli and
specifically, posits a two tone system (high and low) for the language. The
depth analysis of the tonal system of the language. In addition to the two tones
that are established in his 1994 study, this research identifies an emerging tone
(mid) which is an intermediate level between the high and low. The study also
identifies the syllable as the tone bearing unit and describes some of the tone
processes that are evident in the language. Tones, in the study, are described
so it concludes by discussing the role of tone in the morphology of Buli and the
All these researches shed a lot of light on some important aspects of the
grammar of Kusaal because there are several points on which they converge.
Melancon (1957) identifies the major sounds of the language describing some
of its morphology and explicating the various kinds of prefixes, roots and
suffixes that are inherent in the language. He makes some notes on the syntax
12
of Kusaal pointing out some of the major grammatical categories. The study
Spratt and Spratt (1968) is a study on the phonology of Kusaal. It captures the
phonological hierarchy of the sounds and identifies the phoneme as the basic
level of the hierarchy, after which comes the „bar‟ (which corresponds to the
syllable). The study mentions tone but does not go into detail on its influence
orthography of Kusaal and includes a word list based on Swadesh‟s 100 item
syntax and does quite a good job of describing the language. All these works
have, conspicuously, been done by non-native speakers and even in the last few
Two primers which are used for literacy purposes in the non-formal education
sector of the Ghana Education are also available. They are written in a simple
and straight forward manner and talk to one social issue or the other and so
they will be relevant sources and appropriate reference material for this study.
The most recent work on Kusaal however, is, Abubakar (2007) which describes
some of the main points on which they either diverge or converge with what
Kusaal has seen relatively little description in the last forty or so years.
this current state of affairs leaves much to be desired. Further, the stock of
literature is too small and mostly too obsolete to allow for an instructive usage
education;
it will also contribute toward efforts being made to have the language
ii. an account of the various syllable types and structures that are
and
Three phonological theories are relevant to this study. The „Distinctive Feature‟
and „Moraic‟ theories are specific to Chapter Two while the „Autosegmental‟
theory is very crucial to the chapter on tones (Chapter Four). All of these are
In line with the views inherent in this theory, grammar is viewed as a system of
direct outcome of the underlying syntactic structures that are determined by the
not in terms of their acoustics (cf. Jakobson and Halle 1956) but rather, with
(http://www.sil.org/linguistics/GlossaryOfLinguisticTerms/WhatIsGenerativeP
representation. The former refers to the most basic form of a word or sound
representation refers to the form of the word that is heard or said - the actual
seek to explain the mechanisms that are on-going in the speech production
process. These rules map underlying representations onto phonetic forms and
among others.
(2001: 199), grows out of the work of Trubetzkoy (1890-1938), Jakobson et al.
(1951) and Jakobson and Halle (1956). The essence of “distinctive features” is
features. These features constitute the phonetic realities that are peculiar to the
segments, thus marking them as different, one from the other. It also considers
comprise the set of stages or processes which are used to generate the phonetic
production processes.
(Kenstowicz and Kisseberth 1979). The trend initially, was to view the process
of sound production as occurring in a linear fashion, one after the other. On this
17
issue however, there were a lot of opposing views because then the framework
failed to account for those items that operated beyond the level of the segment.
appear to be able to operate on their own. They are, to a very large extent,
tones.
The figures in (1a and b) below succinctly make the point about the crucial
(1)a.
b.
Observe that the various components of the derivational process in (1a) occur
sequentially, the one after the other. This represents the linearity that
at the terminal end of the cline is produced after the underlying representation
rules).
Over the years however, generative phonology has seen a lot of rethinking and
and labial are now viewed not as binary sets (Chomsky and Halle 1968: 65) but
as unary ones. These unary features draw their labels only from the active
articulators and are either specified for a segment or are considered irrelevant
descriptions. Also, the thinking has greatly changed on the initial sentiments by
Chomsky and others on the linear model that languages were assumed to take.
especially suprasegmentals.
In adopting the „generative phonology theory‟ and the related philosophies for
this study, it is hoped that our attention will veer more toward seeing grammar
competence speakers have when they use their language rather than their
ability to simply perform in it. This approach is also useful to the present study
19
segments (cf. Chomsky and Halle 1968: 65), Goldsmith (1976, 1990) proposes
(Goldsmith 1990: 8). In this regard, the phonological features that are specified
recognised as “extending over units which can encompass more than one
phonological representation. These tiers differ from each other and consist, on
20
their own, of a “string of segments”. Further, the features that are specified on a
particular tier are not allowed to appear on different tiers and so this makes for
The features that are specified on each tier are linked to the segments on the
between segments to show that they do not occur haphazardly but in tandem
with each other. They are used to indicate the “simultaneity in time” in the
of mapping between these segments is not always a „one to one‟ affair (as in
2a). It is therefore possible to observe that some features on the different tiers
are „multiply associated‟, i.e., a single segment on a particular tier may have
two association lines emanating from two features on a different tier being
(2) Kusaal
a. L H b. H
H implying low and high tones respectively. The segmental tier also relates to
while the unbroken lines connecting the two are the lines of association.
21
Further, features on the various tiers are mapped unto each other by a
convention determines that all other vowels are linked to some tone in an
outward fashion once there is some congruence between them. In light of this,
tones as well as the tone bearing units (TBUs) are matched, one to one, from
left to right. TBUs that are left-over are then associated with the last tone while
the left-over tones are equally associated with the last TBUs (Example 3a).
(3) Kusaal
a. L H LH
draw association lines across each other. This is in keeping with a formal
b. * L H H
The asterisk (*) in (b) above, shows that the association between the elements
on the two tiers is ill-formed. The goal of the „UAC‟ and the „No-crossing
constraint‟ is such that they aim at explaining what is ongoing in speech which
involves phenomena that operate beyond the level of the segment. A Well-
which differ from one tier to the other. The theory thus embodies a number of
ii. Separate tiers which establish the various manners and places of
assigned a tier which shows whether the underlying tone is high (H)
23
tone/tonal tier.
iv. Various tiers could also be specified for the respective phonological
the segments that are undergoing some change. Perhaps the most
is also referred to as the „CV tier‟ or the „timing tier‟. Rather than
then projects slots for the consonant and vowel segments. These
slots, which can be either V or C slots are the anchor points of all
consonants in such a chart must link unto the skeletal tier or they will
They include the following which are explained in relation to the figure in (4):
24
tier which provides the slots unto which the segmental features are
which spells out the exact consonant and vowel segments in the
process.
ii. Straight lines represent „association lines‟ which link one feature to
another while
iv. Two short lines drawn across an association line indicate that there
both of them.
The example in figure (4) below puts all of the above into perspective. It
highlights the processes that derive the French word /b/ from bon „good‟.
25
(4) French
Nasal Tier -N +N -N +N -N +N -N +N
Skeletal Tier x x x x x x x x x x x x
Segmental Tier b n b n b n b n
-N +N -N +N
x x x x x
b n b
In the derivation, two phonological processes which are strictly ordered are
represented. First is a spread of nasality from the word final nasal segment /n/
to the syllabic segment //. The second process involves a complete delinking
(Chapter Four) as it allows us to show how tones (together with other prosodic
units such as stress, vowel harmony, nasalisation, etc.) operate on their own. It
also allows for us to see how prosodic features are independent of other sound
(1932), but some of the contemporary research has been carried out by such
scholars as Hyman (1984, 1985) and van der Hulst (1984) after a lot of
26
theory‟, unlike the onset-rhyme theory, divides the syllable into „weight units‟
The approach views each mora as the segment which contributes to the weight
consonants) that may not add to the weight of the syllable. Light syllables
according to the mora theory are monomoraic, i.e. they contain only one mora
while heavy syllables are bimoraic, i.e., they contain at least two moras (Ewen
and van der Hulst 2005). The two moras may be only vowels (such as VV or
CVV) or they may include at least one syllabic element and a syllable final
consonant (eg. VC). The symbol for the „mora‟ is „µ‟ and (5) is the moraic
(5)
µ µ
(C) V (V)(C)
Following from the above, initial consonants and any prevocalic material are
characterised as descending directly from the syllable node while the weight
The motivation for using the moraic approach in analysing Kusaal syllable
vowels and vowel lengthening which rely more on the weights of syllables
1.11 METHODOLOGY
accurate and conclusive results, such data was often tested with other native
speakers. Language experts with insight into the Mabia languages especially on
In view of the fact that this thesis was subsumed in another project that was
work (approximately 10 days) was carried out in October 2009 in two of the
Kusaal speaking areas, precisely in Bawku and Zebilla. Data was elicited from
12 respondents in all; six from each of the dialects mainly represented in the
two towns and this was captured on a digital recorder. Transcriptions were then
The 12 respondents comprised six males and females respectively. The ages of
the respondents were also greatly varied so as to capture the current state of
obtains. Of the total number, three were below 17 years of age while four were
nearly middle-aged (between 25 and 35 years). The remaining five were people
aged 55 years and above. In all of these cases, the respondents showed a lot of
interest in helping towards the efforts being made at documenting the language
and at having their voices on the recorder. When the respondent was literate, it
facilitated the recording session as they could take charge of the wordlist and
28
provide the answers without any subtle attempts at leading them as was the
Respondents were also made to recount short stories and these were recorded.
The Ibadan wordlist was also a very useful resource in the data collection
process because it allowed for the researcher to gather data that is at once both
realistic and specific to the purposes of the study. The responses gathered were
include wordlists and previous studies. Some of the sources of this study
therefore come from some of the scant existing literature on the language.
In order to determine the nature of tone in the language and how they patterned
up in both the morphological and syntactic systems of the language, the Praat
on what it is that the language users do with the range of pitches at their
disposal.
This thesis comprises five chapters. Chapter One makes the introductory
statements on the language under review and as well highlights the various
theories which are relevant to different sections of the study. Chapter Two
to the tonology of the language. The conclusion and summary of findings are
1.13 SUMMARY
In this chapter, we have sought to ground the basis of this study. We started by
stating the scope of the study as well as the motivations for undertaking the
research. The chapter also described the demography, sociology and ethno-
linguistic profile of the people under review. Relevant previous studies on the
language and its cognates were also briefly cited. The theories that underpin the
methodology that is adopted for the study and the organisation of the rest of the
work.
30
ENDNOTES
1
This map shows eight districts of the Upper East Region but actually there are nine.
The Kasena-Nankana District was in October 2007, divided into the Kasena-
Nankana East and West districts with their capitals in Navrongo and Paga
respectively.
2
This is the acronym for the Ghana Institute of Linguistics, Literacy and Bible
Translation (GILLBT) located at Tamale in the Northern Region of Ghana.
31
CHAPTER TWO
2.1 INTRODUCTION
This section of the study entails a detailed description of the sound systems of
Kusaal. The description will be carried out within the larger „generative
and Halle (1968 hereafter SPE). Phonemically, Kusaal has twenty four (24)
The consonant sounds that occur in the language relate to seven places and five
on Table (1) below. Among them are four double articulated sounds which are
Approximant j w
Lateral l
Trill r
32
i. Plosives
All the plosives in Kusaal have voiced and voiceless counterparts except the
velar glottal // which is naturally voiceless. The glottal usually occurs at
orthography. The double articulated sounds /kp/ and /gb/ form one unit and are
thus inseparable.
The plosives occur in the following words: the first sets under the phonemes
represent occurrences at syllable initial position while the second are syllable
syllable position. In this study, long vowels that have the same tone are marked
/p/ /b/
pa „strength‟ ba „ring‟
/t/ /d/
tam „forget‟ dam „shake‟
/k/ /g/
ks „cough‟ gs „look‟
/kp/ /gb/
kp „enter‟ gb „sleep‟
- -
// /h/
- ah exclamation
ii. Nasals
All the nasals are naturally voiced. An allophonic variation of the bilabial nasal
fricative /f/. The palatal // is written (ny) while the velar // is left as such in
/m/ /n/
mua „suck‟ nua „fowl‟
nim „meat‟ min „also‟
// //
nyin „tooth‟ -
- tl „to send‟
/w/ /w/
wiak „wring‟ wiak „hatch‟
- -
iii. Fricatives
The labio-dental and alveolar fricatives have both voiced and voiceless
counterparts. While the palatal fricative // is voiced its glottal counterpart /h/ is
confused with the palatal approximant /j/ which is written as (y) in the
orthography.
/f/ /v/
fu:g „shirt‟ vu:g „pull‟
nf „eye‟ -
/s/ /z/
siak „agree‟ ziak „to variegate‟
ni:mi:s „birds‟ -
// /j/
jauk „type of dance‟ ya:g „grave‟
- -
/h/ //
hali „until‟ bia’ar „mud‟
ah „exclamation‟
iv. Glides
and trills. All glides in Kusaal are intrinsically voiced. The palatal approximant
/j/ is written as (y) while the labial-velar approximant /w/ occurs together with
the labial-velar nasal /w/. The lateral /l/ is more pronounced in the language
than the trill /r/ which is prominent only in idiolectal or dialectal variations.
/j/ /w/
yalg „wide‟ walg „shower‟
- -
/l/ /r/
flg „sheath‟ frg „to sip‟
fa:l „space‟ fa:r „to save‟
35
medial and final positions. Shaded areas indicate that the consonant does not
Based on the distribution of consonants on the above table, we can make the
following generalisations:
i. Initial positions
At word and/or syllable initial position, all the consonants can occur except the
velar nasal //. The behaviour of the glottal stop // in the above regard is quite
The glottal fricative /h/ and the labial-velar nasal /w/ are the only consonants
syllable medial positions, no consonants obtain. Such positions are the absolute
reserve of vocoids and syllabic nasals which qualify to go into syllable nuclei.
sounds /kp/ and /gb/, and the glottal // do not occur. Two nasals, the palatal //
and the labial-velar /w/ are also restricted from occurring at such positions.
Together with the approximants /j/ and /w/, all the voiced alveolar fricatives
/v/, /z/, and // are barred from occurring at syllable or word final positions.
37
Spratt and Spratt‟s (1968: 29) pioneering work identifies a nine oral vowel
system for Kusaal [i, , e, , a, u, , o,] and five nasal vowels [, , a, , ]
which are all produced with an advanced tongue root. These vowels are still
very important in the language. All the nine oral vowels also have long vowel
correlates2. Sequences of vowels are also very prolific in the language. Table
language:
LOW
a
All the vowels of Kusaal are produced at either the front [i, , e, ], central [a]
or back [u, , o, ] of the tongue and also at the high [i, , u, ], mid [e, , o, ]
or low [a] position of the tongue body. The vowel sounds on Table 3 above
represent only the phonemic short vowels in the language. Examples of their
All the short vowels in the Table 3 above can also be produced with a
perceived degree of lengthening in their quality. The evidence for long vowels
is given by the existence of many lexical items in the language within which
vowels are produced over a longer duration of time than the short ones. Long
vowels are usually represented by the short vowel sound followed by a colon
(). The sets of words below are illustrative of the distribution of short and long
vowels in Kusaal. While the set to the left exemplify short vowel occurrences,
In the production of vowel sounds, the velum is usually raised above the
normal position in order to check the airstream from escaping through the nasal
cavity. This rids the vowel of any nasal qualities thus making it a bona fide oral
lowered to ensure substantial airflow through the nasal cavities‟ (Clark and
Yallop 1995: 31-2, 63). Kusaal has five of such nasal vowels and they are
drawn only from the unadvanced/retracted tongue root [-ATR] sounds. They
within the same syllable. Such articulations involve vowel glides from one
height position in the oral aperture to another and their number and occurrence
the first parts usually take a stronger stress than the second. The following
make the foregoing clearer. However, they do not on their own, make up an
SPE (Chomsky and Halle 1968) is the basis of this section of the study. It
(1951) and Jakobson and Halle (1956) and presents the view that phonological
building blocks of phonological analyses and as such they are the properties
that mark one segment as discrete from the other. They are thus regarded as the
features with a set of features that have, in most cases, articulatory correlates‟
perspective. This makes for a lot of clarity and precision in the description and
The landmark study specifies a larger range of unique universal features in the
(1956) which posits only twelve DFs. It also assigns to each of these features
the binary values of Jakobson and Halle (ibid). Binarity specifies that a feature
former case, it attracts a plus (+) value while in the latter it attracts a minus (-)
value. In light of stronger evidence and current thinking since the time of the
SPE, some features, namely, [labial], [dorsal], and [coronal] have come to be
41
production. They are labelled unary features because, unlike their binary
counterparts which are specified with regard to their passive articulators, they
(unary features) are either relevant to the segments under consideration or they
Roca and Johnson (1999: 97-8) point out that though the parameters used in the
„distinctive feature theory‟ are naturally binary, it is not all of them which
behave as such. The point they make is that the presence of some articulatory
gestures does not imply the absence of another in some situations. For instance,
the presence of labial consonants does not mean the absence of other
express mutually exclusive groups - where the presence of one precludes the
mean the absence of the other. In essence, the features [labial], [coronal] and
[dorsal] “can co-occur since the gesture each feature represents is not
other,
segments,
phonological processes,
(http://nickclements.free.fr/featuretheory.html)
These functions set out some of the crucial roles a feature theory plays in cross-
drawing largely on SPE (1968: 299 ff). An attempt has been made in this study
to minimise redundancy from the start and so features that are similar in
terminology or function have been collapsed into each other. The following are
The major class features differentiate those sounds that are produced with
vocal tract. Consequently, they make the key distinctions between which
2.4.1.1 [Consonantal]
region of the vocal tract; non-consonantal sounds are produced without such an
obstruction” (SPE: 302). [+cons] sounds will therefore include all those ones in
the production of which the airstream is manipulated. These are all the
consonants in Table (1) above; plosives, fricatives, nasals and lateral sounds.
2.4.1.2 [Sonorant]
The extremes in this feature are used to explain the voicing quality inherent in
any two sounds. While sonorant sounds are “produced with a vocal tract cavity
impossible” (SPE: ibid). Obstruents are the same as [-son] sounds. Sounds
which are naturally voiced in Kusaal are the vowels, nasals, palatals, and all the
Manner features specify the stricture types that are used as the airstream flows
through the vocal tract. Stricture refers to the various modifications of the
2.4.2.1 [Continuant]
Sounds which are produced with an absolute obstruction of the airstream are
considered [-cont] while those sounds in whose production the airstream is not
impeded are [+cont]. This feature thus groups the fricatives, approximants and
laterals into the [+cont] category while plosives and nasals are classified as
2.4.2.2 [Nasal]
the velum from those which are produced with a raising of the velum. The
process of either lowering or raising the velum allows for a larger component
of the gust of pulmonic egressive airstream to exit through the nasal or oral
cavities respectively. Sounds produced through the nasal cavity are given a
[+nas] attribute. They include all nasal consonants and nasalised vowels. All
2.4.2.3 [Lateral]
Only one lateral segmental feature is specifiable in Kusaal; the alveolar lateral
[l] which is produced by lowering the mid-section of the tongue at the sides
allowing air to flow out near the molars. The lateral sound is voiced in Kusaal.
2.4.2.4 [Strident]
This feature makes a difference between the „acoustically‟ more noisy sounds
and the less noisy ones. The SPE states that the surface over which the
airstream flows, the rate or even the angle at which it flows may result in the
45
This class of DFs puts sounds into groups based on the places in the vocal tract
where they are articulated. A total of seven such features are relevant for the
purposes of this study. The discussion further incorporates some of the current
trends in the „distinctive feature theory‟ such as viewing the coronal primary
stricture of Chomsky and Halle (1968: 304) - in unary rather than in binary
terms. Two other features which are described as unary - [dorsal] and [labial],
are also found under this group. Referring to such features as unary is more
certain feature is part of the character of a specific segment from the point of
view of the passive articulators. In the case of the unary features, active
as the blade or back of the tongue or the lips makes a precise movement when
superfluous to specify such an action when the active articulator has not moved
2.4.3.1 [Coronal]
While Chomsky and Halle assign binary values to the feature „coronal‟ (SPE:
304), this thesis makes reference only to the privative coronal for the reasons
stated above. Therefore, sounds which are coronal are produced with the blade
46
tongue is raised above its normal position in the oral cavity and is very critical
to the nature of the sound that is to be produced. Coronal sounds in Kusaal are
all the alveolar and palatal sounds. The rest are not coronal.
2.4.3.2 [Dorsal]
Dorsal sounds are produced with the back (dorsum) of the tongue as the active
articulator. It is a unary feature and refers to the [+back] feature of the SPE (pg.
305) which is “produced by retracting the body of the tongue from the neutral
position.” The velars [k, g and ] and the glottals [ and h] together with all
back vowels have a dorsal constriction. The labial-velars [kp, gb w, and w]
also take part of their nature from this feature and from the feature [labial]. It is
therefore possible to say that such segments are, at once, both [labial] and
[dorsal] rather than labelling them as [+back] [-back] at the same time.
the binary [back] feature. The binary specification is still important because the
plus (+) or minus (–) oppositions are indispensable in apt descriptions of some
nature of vowels because aside front and back vowels, there are also central
members.
2.4.3.3 [Labial]
Labial sounds are produced with the lips as active articulators. The flow of the
airstream is thus modulated by the two lips or the lower lip in conjunction with
47
some other articulator such as the teeth. It thus captures all the bilabial and
feature labial is also used to describe the labial-velars [kp, gb and w]. All
2.4.3.4 [Anterior]
“Anterior sounds are produced with an obstruction that is located in front of the
palato-alveolar region of the mouth; non anterior sounds are produced without
such an obstruction” (SPE: 304). All sounds which are produced at any point
from the alveolar ridge to the lips are [+ant], i.e. all the alveolars, the labio-
dentals and the bilabials. The [-ant] sounds are produced at some point after the
alveolar ridge beginning from the palatal region through the velum to the
pharynx. The labial-velar plosives [kp] and [gb] are however produced
spontaneously at the two extremes and so they may not be properly described
impression this gives is that they are produced sequentially: [k] first then [p].
2.4.3.5 [High]
The feature [+/-high] relates to the vertical positioning of the tongue in the oral
cavity (Catford 1988: 124). When at rest or neutral position, the body of the
tongue is usually raised and pointed towards the front part of the mouth. In the
production of [+high] sounds, the body of the tongue is further raised beyond
this neutral position while [-high] sounds are produced without any such
movement (SPE: 304). [+high] consonants are the palatals and velars. The high
48
vowels [i, , and u] are found in this group. All other sounds attract a [-high]
value. Nonetheless, this should not be taken to mean that [-high] sounds are
naturally [+low] but rather that they are produced without any considerable rise
2.4.3.6 [Low]
While [+low] is used to describe a lowering of the body of the tongue below
the normal rest position when articulating some sounds, [-low] indicates that in
such situations the tongue is not lowered below the normal. Along much the
same lines as the [+/-high] feature above, [-low] does not imply [+high]. Two
[+low] consonants are used in Kusaal: the glottal plosive [] and fricative [h] in
addition to one [+low] vowel [a]. Considering the few representatives of this
group, one can assert that the process of further lowering the tongue below the
language users often aim at using the most facile and readily available means to
achieve their communicative purposes, marked forms, of the like above, are
2.4.3.7 [Round]
Rounding relates to the posture of the lips during certain articulations. While
[+round] sounds are produced with a narrowing of the lips, [-round] sounds are
produced without any such narrowing (SPE: 309). This feature thus
distinguishes between the rounded vowels [u, , o, ] and the unrounded ones
[i, e, , a]. The labial-velar fricative and approximant [w, w] are also
2.4.3.8 [ATR]
Kenstowicz (1994: 14) explains that “in the pharynx, the root of the tongue
that are described in these terms because yet another parameter is needed to
distinguish one vowel from the other in languages such as Kusaal which have
four vowel height specifications. Aside the [+/-high], [+/-back] and [+/-round]
distinctions then, [i, e, u and o] are considered [+ATR] while [, , , and a]
2.4.3.9 [Voice]
This feature puts into context the action of the vocal cords as the airstream is
pushed out of the larynx. When the vocal folds vibrate during the process of
indicated by [+voi], while those produced without any vibration are voiceless
The state of the glottis is another parameter used to describe the nature of
that are „aspirated‟, i.e. produced by spreading the vocal cords or „glottalised‟ –
by constricting the vocal cords. When the sounds are aspirated they are given a
50
[+spread gl] feature while they are [-spread gl] when they are glottalised. Spratt
and Spratt (1968: 11) identify three [+spread gl] plosives in Kusaal [p, t, and
k]. [-spread gl] is wide spread especially intervocalically and degenerates from
The charts in (4) and (5) below are the feature matrices for the various
consonant and vowel sounds in the language. The norm, usually, is to put a
specifications without any overt marking. However, considering that the entire
the matrices, such an enterprise will prove futile chiefly because the charts will
appear a jumble of circles and double circles (if all the redundancies are
circled). Advertently, outright redundant features have been left blank. Also,
since unary features do not take a (+) or (–) value, they are specified with a tick
(√) when they are relevant to the segments under discussion. When they are
p b t d k g kp gb m n w F v s z h j w l r
Consonantal + + + + + + + + + + + + + + + + + + + + + + + +
Sonorant - - - - - - - - - + + + + + - - - - - - + + + +
Continuant - - - - - - - - - - - - - - + + + + + + + + + -
Nasal - - - - - - - - - + + + + + - - - - - - - - - -
Lateral - - - - - - - - - - - - - - - - - - - - - - + -
Strident - - - - - - - - - - - - - - + + + + + + - - - -
Coronal √ √ √ √ √ √ √ √ √ √
Dorsal √ √ √ √ √ √ √
Labial √ √ √ √ √ √ √ √ √
Anterior + + + + - - - + + - - + + + + - - - + +
High - - - - + + - - - + + - - - - + - + - -
Low - - - - - - + - - - - - + -
Round - - - - - - - - - - - - - + - - - - - - - + - -
Voicing - + - + - + - + - + + + + + - + - + + - + + + +
Spread glottis + - + - + - + - + - - - - - - - - - - - - - - -
52
i e a u o i : e a u o a
Consonantal - - - - - - - - - - - - - - - - - - - - - - -
Sonorant + + + + + + + + + + + + + + + + + + + + + + +
Continuant + + + + + + + + + + + + + + + + + + + + + + +
Nasal - - - - - - - - - - - - - - - - - - + + + + +
Front + + + + - - - - - + + + + - - - - - + + - - -
Back - - - - - + + + + - - - - - + + + + - - - + +
High + + - - - + + - - + + - - - + + - - + - - + -
Low - - - - + - - - - - - - - + - - - - - - + - -
Round - - - - - + + + + - - - - - + + + + - - - + +
ATR + - + - - + - + - + - + - - + - + - - - - - -
Voicing + + + + + + + + + + + + + + + + + + + + + + +
53
Having established the various features that are used in the description of the
other features in a matrix (Ewen and van der Hulst 2001). Kenstowicz (1994:
vis those ones that are unpredictable, contrastive or distinctive. From the charts
the features of the sounds. For instance, one can infer that a segment which has
the feature [-cons] also has the feature [+son]. In like manner, if a segment is
specified for [+nas] then it means that it is also [+son]. Statements of this sort
to (8), as the precedents and then proceed to extract some redundant features
which are specific to Kusaal in examples (9) to (13). The downward pointing
arrows are used to connote „implies‟ or „then‟. Statements that have (a) and (b)
counterparts are „“inversely” symmetrical‟ (Schane 1973: 36), i.e., when the
54
The notation in (4) makes a a prediction about nasal sounds. It states that once
voiced segment.
(4) If [+nas]
[+son]
In (5), one can infer that if a segment is [+son] then it is also naturally voiced
(5) If [+son]
[+voi]
From (6) below, the statement is to the effect that if a sound is a nasal, which
also makes it, a naturally voiced sound, then in its production there is a
(6) If [+nas]
[-cont]
If a segment is produced with the tongue body raised above the rest position,
then it cannot at the same instant be produced below the normal rest position of
vice versa is also true in the (b) statement of (7) below. These statements are
thus inversely symmetrical to each other. In (8), the formalism suggests that if
a segment is produced from the alveolar region onwards to the lips [+ant], then
[-low] [-high]
(8) If [+ant]
[-high]
[-low]
The statements in the above are attested across many of the languages of the
world. What follows in (9) to (13) is information about some of the predictable
patterning of features in Kusaal. Considering that there is only one low vowel
in the language, one is able to infer then that such a segment is automatically
unrounded and has other feature specifications for [-back] and [-ATR]. This is
(10) (a) If V
+nas
[-ATR]
Statement (10) situates the place of nasal vowels in the language. Kusaal has
five nasal vowels which are specified for all the features of vowels in addition
to nasalisation. The nasal vowels that occur in the language are restricted to
only five phonemic short vowels and so one can predict that if a vowel has a
nasal quality in Kusaal, then it means that it is produced with a retracted tongue
root.
The following (11a and b), are also inversely symmetrical to each other. A
[-strid] [+cont]
Unary features are the subject of (12) and (13). In (12) the statement points to
the fact that if a sound is coronal, i.e. produced with the blade of the tongue
(12) If [cor]
[-low]
[-round]
Along much the same lines as the preceding redundant statement above, the
with the dorsal feature. Precisely, one can deduce that if a segment has a dorsal
(13) If [dors]
[-ant]
[+high]
The point of all the discussions on the „distinctive feature theory‟ and thereafter
function. These points have already been made elsewhere in this chapter
(section 2.3) and pertain to the functions of the feature theory. Summarily,
describing the sound system of Kusaal along these lines is a very crucial step
achieve „economy‟ through the use of a small set of features to describe a vast
range of segments; they also enable us to group sounds into „natural classes‟
which share similar features. Features also perform phonetic and phonemic
instance and by creating lexical differences in the second (Urua 2009: 4).
58
Speakers generally pattern their languages along lines that will make the
phonological items in these languages specific to them while at the same time
conversations when they have to take a pause and then continue. This inherent
in such languages and so it is the case that the sounds in any language are
complex unit which comprises nuclear and marginal elements (Laver 1994:
114). These elements together make up the atomic or „internal structure‟ of the
occur before or after the nucleus. In the following discussion (mostly adapted
from http://www.personal.rdg.ac.uk/~llsroach/phon2/mitko/syllable.htm), we
cite some auditory, acoustic and articulatory motivations used to describe the
syllable.
pulsations in the chest. The „chest pulse theory‟ thus views each syllable as a
product of the activities of chest muscles and lungs in the speech production
process. Granted that the number and type of syllables can be determined by
59
experimenting with chest pulses and increases of air pressure in the lungs, this
approach however fails to account for especially peak only syllables which do
not involve chest pulses. In Kusaal, for instance, the word for „yes‟ is . This
syllable or not because it does not involve any evident situations of chest
syllable descriptions.
syllables in a word, one must identify the number of prominent sounds in that
the number of vowels that occur in that word. Thus, a word containing three or
However, this theory does not help much in discussions of syllable division
because what may be heard as one peak may actually be more than one and
vice versa.
To buttress the foregoing, we will draw on the Kusaal word for „goat‟ buug.
represent two peaks of prominence. However, this is not the case in the
buug which contains two syllables. The glottal stop thus becomes the onset of
60
the second syllable. In the second instance, the velar final consonant is deleted
to derive buu which has a long vowel occurrence and only one peak of
syllable depictions.
The last approach to describing syllables, which is not unlike the foregoing
another. Each syllable then corresponds to a peak in the flow rate of pulmonic
air. In the natural scheme of things, vowels are the most sonorous sounds
order.
Two main theories that support the importance of organising sounds into
syllables are cited in the literature: the „onset-rhyme theory‟ and the „mora‟
theory.
Ewen and van der Hulst state that “in the Onset-Rhyme theory, the syllable is
consonants preceding the vowel, and the rhyme which contains the vowel and
whatever follows it” (2001: 120). In this regard, the onset and the rhyme are
higher level categories that come immediately after the syllable. The rhyme
61
Onset Rhyme
Nucleus Coda
The terminal nodes that come after the onset are often consonants that indicate
syllable initial positions while the nucleus could branch further in nucleus
weight languages. Syllabic consonants and vowels are the only elements that
occur in this position whereas the coda is the preserve of any other occurring
consonants.
draw on moraic theory which divides syllables into „weight units‟ rather than
into immediate constituents called onset and rhyme. The approach views each
mora as the precise segment which contributes to the weight of a syllable while
there may be other occurring segments, especially initial consonants that may
We shall draw on „moraic theory‟ for the analyses of the syllable structures that
are evident in Kusaal because this system is more representative and is able to
vowel sequencing as well as long vowel occurrences (see also §1.10.3 above
As has already been stated, the syllable segments that occur in the language
will be accounted for in terms of the number of moras in the syllable rather
The following are some preliminary statements about the structure of syllables
in Kusaal. To begin, only vowels and syllabic consonants (which in this case
refer only to the nasal sounds) can occur in the nucleus of the syllable. Further,
initial consonant clusters of any kind are barred from occurring in Kusaal
with all the other consonants in the language with the exception of the velar
nasal // and the glottal stop //. The diagraphs „ny‟ in words such as nyan
„shame‟ and „nw‟ in nw „hit‟ are the palatal [] and labial-velar [w] nasals
respectively.
At word medial position, the clusters that occur are actually boundaries
between compound words, e.g., tenban „shrine of the land‟: ten „land‟ and
ban „shrine‟, while there are no consonant clusters of any kind occurring in a
syllable. In addition, Kusaal may not simplify word final consonant segments
„child‟ is biya in Guren but bi: in Kusaal. The following syllable types
63
obtain in the language: the peak only (V/N), the CV, the VC and the CVC
syllable types.
also be found as the only elements in such syllable types. These forms are
light-weight syllables as they involve only one single node from the syllable to
the mora (cf. with the graphical representation in 17 below). Examples of such
(b)
2.6.3.2 VC syllables
There are a few syllables which take the VC pattern. Such syllables are heavy
syllables because they have two moras occurring as separate branching nodes
within the syllable. There are also a few syllables that are patterned on two
subsumed under this categorisation chiefly because these vowels are often long
b. (i) ii)
µ µ µ µ
They also point out that the syllables are heavy weight syllables because there
are two moras in each case. Observe also that the long vowel in (16b) :n „dry
2.6.3.3 CV syllables
world and also in most of the indigenous Ghanaian languages which have
syllable types that are structured along similar lines. This structure is preferred
Table (1) above combine with one vowel or the other to create different lexical
or functional items in the language. These syllables, in line with the views
inherent in the mora theory, are also light-weight syllables. Examples include
65
b. i. ii.
µ µ
p „swear‟ kp i „die‟
An extended form of the CV syllable structure is the type which has an onset
occurring before two obligatory vowels (CVV). The vowels that occur after the
consonant may either be long or they may be a sequence of vowels. The range
of consonants that can occur at syllable initial position is also not constrained.
lexical items (18a). The structural representations of these forms are in (18b):
b. i. ii.
µ µ µ µ
p a „woman‟ l a „bowl‟
66
The structural representations above show that the two syllables are heavy-
weight syllables because they are bimoraic, i.e., they contain two moras. In
the number of vowel segments present in it. Short vowels take only one mora
Perhaps, one of the most important syllable structures in Kusaal is the CVC
syllable structure which is quite unlike the forms which pertain in the other
Mabia languages. Usually, most of the languages in the Mabia subgroup prefer
The case in Kusaal is different with syllables taking the CVC structure, often,
compulsory onset and coda consonant. Except the velar nasal //, all the other
consonants can occur at syllable initial position. All the consonants can also
occur syllable finally with the exception of the labial-velar consonants /kp/,
/b/, /w/ and /w/, the palatal nasal // and the voiced labio-dental and alveolar
fricatives /v/ and /z/. Examples of this syllable type include the following in
b. i. ii.
µ µ µ µ
k k „stool‟ W i n „God‟
The vowel in a CVC syllable type could also be two obligatory vowel sounds
which may either be sequences of vowels or long vowels. The vowels in the
CVC syllable could also be two nasalised vowels. Consider the following
(20)a. CV:C
fa:r „property‟ ma:l „make‟
b. CVVC
piab „speaking‟ luak „search‟
c. i. ii.
µ µ µ µ
d a m „pito‟ p i a b „speaking‟
68
weight and so they are moraic. Such consonants may take a mora if there is no
vowel which goes into the second mora slot. When a vowel enters into the
second mora slot, syllable final consonants are associated with such vowels
(Hayes 1989).
It is important to also state that the number of syllables in a Kusaal word could
range from one to four. Often the longer words are compounds of some already
These are words which are made up of only one syllable which function as
complete morphemes. Examples include the following: m „I, me‟ which is peak
only. All the other syllable structures can also constitute single syllabic words.
ti „our‟
These sets of words arise from a combination of two syllables which take
(24) below:
2.8 SUMMARY
An inventory of the number and nature of Kusaal sounds is the major goal of
this chapter of the study. We began by establishing twenty four (24) consonants
and nine (9) phonemic and vowels. Nine long and five nasalised vowels are
also widely used in the language while a number of sequences of vowels also
prevail. The chapter also establishes the various distinct groups or classes into
which the sounds of the language are found and proceeds to make some
ENDNOTES
1
In both dialects, /d/ and /r/ are free variants of each other usually at syllable
initial position in bi-syllabic words. Free variation also occurs in mono-
syllabic words but usually in rapid speech. Sometimes they mark dialectal
differences as in the example for „room‟ on the table (2). Other examples
include the following which are however not dialectal variations:
ta:ra/ta:da sandals sra/sda truth
2
Consultations with some representatives of the Community Literacy
Development Programme (COLDEP) and GILLBT (Messrs. Awimbilla and
Alalbilla) indicate a lot of agreement on this matter. The existing translation of
the New Testament was done using only five phonemic vowels: i, e, a, u, o.
This poses a number of problems in the daily usage of the language in and out
of church.
3
We indicate syllable boundaries with a period within examples or words but
with the dollar sign when we formulate phonological rules (See Chap 3)
72
CHAPTER THREE
PHONOLOGICAL PROCESSES
3.1 INTRODUCTION
This section of the study is concerned with establishing some of the recurring
whereby one sound copies or behaves like another within a word or phrase
3.2 ASSIMILATION
influence of one sound segment on the articulation of another such that the two
change may either be partially or completely assimilated. In the first case, the
major features of the assimilated segment may linger on while in the second
they are completely lost. The assimilatory process may also relate to the
qualities or a segment that comes after another may condition the change. In
they may influence each other from a distance across intervening segments
(Lass 1984: 171). The sets of data in (1) representing consonant nasalisation in
(1) BULI
(2) KUSAAL
From the data above, a number of facts come up. First, all of the data exhibit
assimilation between contiguous segments. i.e., the conditioning factors for the
across morphemes or word boundaries (see 3.1.6 below). Second, while the
consonant segments in the nouns assimilate the nasal qualities of the definite
74
determiner mu) the Kusaal data exhibit a regressive assimilatory process where
the definite determiner la assimilates the features of the syllable final
consonant. A final observation is that all the examples in both sets are cases of
completely lose all of their inherent properties in order to take on new ones.
Rather the segments /b, l, and r/ which undergo mutation in the Buli data, lose
only their oral nature while retaining their voicing, labial, sonorant and anterior
result of assimilation.
In the sections that follow some of the assimilatory processes in the language
3.2.1 Nasalisation
The feature [+nasal] is the bona fide „property‟ of segments such as /m/, /n/,
//, // and /w/ because of the nature of their articulation – with a lowering in
the position of the velum so as to allow the mass of airstream to exit through
In this regard then, all [-ATR] vowels are candidates for nasalisation in Kusaal
especially when they occur before or after nasal segments in the language.
75
The following data sets in (3) through (8) give examples of such nasalisation
processes in the realisation of hitherto oral segments. The oral vowel segments
spread from the adjoining nasal segments onto them. In this study, we will
marks will get in the way of the markings on tone. The following are
illustrative:
(3) N____N
Observe that all the vowels occurring in the data are [-ATR], i.e., they are
produced with a retracted tongue root. This is the case in all environments
and // are also mutually exclusive of each other because they are in
76
initial position, the latter occurs only syllable finally. For the above then, the
The rule states explicitly that, a vowel (V) becomes nasalised when it occurs
between two nasal segments. In the set of words in (5), the vowel segments
occurring immediately after the nasal consonants are also realised with some
nasal quality in their nature. In all the cases of nasalisation evident in Kusaal,
(5) N____
ma „mother‟ ny „see‟
m „know‟
The data in the set above highlight the influence of the adjoining nasal
segments on the vowels. Here, the nasalising effect is from left to right and is
restricted, as stated earlier, to the syllable level. For instance, the last example,
an „to be‟, consists of two syllables and so the impact of nasalisation is not
realised on the vowel in the first syllable but in the second. We formulate the
rule below (6) to capture the foregoing. The statement in the rule is generally,
A third point at which oral segments may become nasalised in Kusaal is when
they precede nasal consonants but occur in the same syllable. Instances of this
process are varied and many and are also restricted, primarily, to the [-ATR]
(7) ____N
pa „strength‟ bn „buttocks‟
bn „end‟
In the above, nasalisation of the oral segments is instantiated from right to left
process take on some of the qualities of the adjoining nasal consonants and are
thus also produced with some degree of nasality. We formulate the rule in (8)
to explain this process. The rule states that a vowel acquires nasality if it is
V
(8) Rule 3: V [+nasal] / ____ [+nasal]
environments that have no nasal consonants. In such cases, the nasal vowel is
segment. These nasalised forms are probably the result of a diachronic process
of nasal assimilation whereby the hitherto overt nasal consonants are lost
through time. Many examples can be found in the language and, as in all the
cases of nasalisation identified in this study, the [-ATR] vowels are the ones
which are usually influenced by this process. The following exemplify Kusaal
nasal vowels.
(9) __
t „vomit‟ :l „vein‟
While we may state that the above may once have had nasal consonants
operating in the underlying environment, this fact does not come up readily in
the above data except in the example for „rings‟ baas. Evidence for the deleted
nasal in this case is adduced from the singular form ba „ring‟. The process
thus highlights an instance of first nasal spreading, then nasal deletion and
finally vowel lengthening to make up for the deleted consonant. In the rest, the
two consonants become homorganic: they agree in articulation (Lass 1984: 48).
segments in the first syllable assimilate the nature of the next consonants so
that they make for ease and economy in the process of communication.
Perhaps, the most important point at which HNA is very pervasive in the
language is when two morphemes or roots are put together to create a new
In the last example tentambs „fine sand‟ for instance, HNA assimilation is
observed at the first two syllable levels. The first nasal is homorganic with the
voiceless alveolar /t/ while the second conforms to the voiced bilabial
Buli too, the possessive pronoun mam „my‟ becomes the bilabial nasal m in
80
most natural occurring speech. This nasal is also the first person pronominal
form and in all instances it assimilates the place of articulation of the initial
consonant of the following noun (in the case of possessives) or verb (in the
case of the pronoun). The examples in (11) and (12) illustrate this
phenomenon.
b. m fu: [fu:]
c. m di:b [ndi:b]
d. m nbr [nnbr]
e. m kpal [kpal]
b. m keya [keija]
set of data (11), possessive nasal consonants consistently assimilate to the place
of articulation of the following consonants and the same is true for (12) which
exemplify various usages of the first person pronominal form. The particle that
is used to express future time - na „will‟ - also influences the nature of the
In Kusaal, much like the case in many Bantu languages such as Swahili,
past time. Three different particles are used to express the hodiernal/immediate
past (indicated with pa:), the recent past (indicated with sa:) and the remote
past (indicated with da:). In all of these instances too, the first person
In the above, the particles pa:, sa: and da: are used to mark various degrees of
past time reference in Kusaal. In (a), the meaning conveyed by the particle pa:
is that of an action which occurred in the course of today prior to the point of
speaking but not yesterday or the day before. The particle sa: in (b) indicates
that the act of killing was carried out only yesterday while da: in (c) refers to
evident only with the first person pronominal form in all cases because it is
3.2.3 Labialisation
Labialisation refers to the rounding of the lips when a sound is being produced
vowels, which are all [+back] or by the labial-velar approximant /w/ which
lend their rounding nature to the segments they co-occur with in syllables. All
consonants occurring before the rounded vowels in syllable initial position are
realised with some degree of lip-rounding. Syllables beginning with the labial-
83
velar approximant are also rounded. The following examples in (14) are used to
The above set of data highlights two cases of rounding. In sets (a and b), there
are instances of inner rounding in the syllables because the initial consonant
that come immediately after them. The last two sets (c and d) however
back vowel which is also [+round] in Kusaal. These examples involve very
consonants can occur. We formulate the rule in (15) to capture the facts above:
(15) Rule 4: C C V
[-round] [+round] / [+round]
Stated simply, (15) means that an unrounded consonant sound (C) becomes
Spratt and Spratt (1968: 37) note that labialisation is one important level at
which difference is evident in the Agole and Tonde dialects of the language.
The fact remains that presently, while speakers of the former dialect may
labialise certain syllables, the speakers of the latter do not add any such
84
features to syllables or words. Consider the following which are adapted from
In the data above, two separate means of labialising segments in Kusaal are
the high round vowel /u/ operating generally after initial consonant segments of
first, primary palatalization, results from a process whereby the body of the
tongue is raised toward the hard palate in the production of some consonants or
(Ladefoged 1982: 210) The second, also secondary palatalization, is the effect
that front vowels or the palatal approximant /j/ have on adjoining consonants,
i.e., the effect that the “addition of a high front tongue position has on another
respectively.
fricative // is quite restricted too, occurring in only a few lexical items. In the
language, the process is related only to the secondary form whereby consonant
sounds occurring before the high front vowel, are produced with the body of
the tongue rising toward the hard palate. These consonants are often alveolar or
velar consonants. We provide the examples in (17) to underlie the effect of the
Observe that, the recurring vowel in the data is the advanced high front vowel.
The contiguity between the vowel and the relatively low consonants, results in
a raising of the tongue body towards the roof of the oral cavity. The statement
in (18) below serves to simplify the above: It states that a non-high consonant
86
V
(18) Rule 5: C C +high
[-high] [+high] / -round
vowels occurring after consonant onsets. In such sequences, the first vowel, if
Situations of pervasive glide formation in one dialect and none in the other is
another point at which Agole and Tonde diverge a little from each other. A
comparison of the following forms adapted from Spratt and Spratt (1968: 38)
dealing with vowel sequences of the type HFV+V (high front vowel + vowel).
more open vowels. In Tonde, however, this is not the case. The dis-preference
for glides in Tonde manifests in the selection of the low central vowel segment
in environments where Agole selects for the high front vowel with another
vowel. Sequences of vowels are also simplified regularly to vowels or, less
Though the underlying representations above appear different from each other,
their respective phonologically depictions are the same: they constitute heavy
bimoraic syllables. The point at which they diverge from each other is that
while the coda consonant in Agole (20a) is adjoined to the preceding mora, in
keeping with Hayes (1989: 257), it constitutes a moraic segment on its own in
(20) a. b.
µ µ µ µ
refers to the phenomenon whereby the vowels in a word are observed to be co-
occurring with each other. This underlying patterning of vowels is noted both
88
which the vowels in a given domain harmonise for a particular feature.” To this
language such as Kusaal, which has four height specifications for the various
heights and is rather tied to the position of the root of the tongue which is not
with evidence drawn from an Akan corpus of data. The approach he adopts in
that study is to describe the harmony between the vowels in Akan words with
respect to the position of the tongue root, i.e., whether the vowels are produced
tongue root [+ATR] implies the vowel sound in question is produced with the
root of the tongue pushed forward while the un-advanced (retracted) tongue
89
root [-ATR] vowels are produced with the root of the tongue drawn further
The nine vowel system of Kusaal can also be divided into the [+ATR] and [-
ATR] oppositions. Based on this distinction, the vowels which occur in a word
will usually be selected almost exclusively from only one of the sets and not
from the two sets at the same time. The sets are shown in (21):
(21) [+ATR]: i e o u
[-ATR]: a
systematic patterning of vowels. It is thus the case that stems or words which
appear on the surface, not to follow this [+/-ATR] opposition, can actually be
examples that can be used to highlight this include the ones in (22) below:
vowels can occur with each other in the organisation of the various words. As
confined to the relative position of the tongue root with regard to the vowels.
Observe that in some of the examples, the vowels are drawn from different
tongue heights. This highlights the cross height selective properties of vowels
in the language. For instance, the mid-low front vowel [] selects for the high
back vowel [] in the word for „beauty‟ vng. What is common to the two
vowels [] and [] is not their heights but the position of the tongue root in their
Harmony between the vowels is also evident in the plural forms of words in the
up readily because, such forms may involve only one vowel segment or one
examples in (24):
91
spread of [+/-ATR] from the first syllable onto the next. This situation is
observed with evidence from many idiolectal variations in the language. For
instance, the sets of data in (25) are possible variants of each other in the
From the above, we can explain that the [ATR] feature is usually very
widespread in the language. Any disparities which may arise will result from
the nature of the low central vowel /a/ which shows up in both the advanced
and retracted tongue root oppositions as in the last example in (25) above.
92
The low central vowel /a/ is usually produced with a retraction of the root of
find this vowel co-occurring with all the [+ATR] vowel segments in the
language resort to producing words with more vowels drawn from the un-
advanced tongue root classification than from the advanced class. In the data
collated on the current state of language use among the people, the propensity
of the respondents to produce more words from the retracted tongue root
light on the „missing‟ [+ATR] low central vowel and point further to a process
establishes the extent of merging between the two sets of vowels. The arrows
i e o u - a
93
The point of the above is to state that the process of merging is on-going
(represented by the broken arrows) from the [+ATR] to the [-ATR] except in
the case of the low central vowel which has been completely coalesced (shown
by the unbroken arrow). The straight lines map the overt representations of the
vowels to the process of merging. We cite the examples in (25) again to explain
that though it is possible to articulate any of the words from the different sets of
vowels, it is more widespread to hear the forms produced with the [-ATR]
vowels. Over time then, it is possible that the nine vowel system of Kusaal may
further be reduced to fewer than seven with the [-ATR] counterparts surviving
the process.
These concatenated stems usually generate new forms after they have
undergone some phonological processes so that they can harmonise with each
other and surface as one homogeneous lexical item. In the resulting output
forms, some of the features of the combined stems are often truncated or even
vowels are usually shortened some consonants are also deleted especially when
In Kusaal, when a syllable containing two vowels combines with the stem of
another syllable, there is usually a shortening of the vowel segments in the first
The candidate which is shortened in order for the compounded form to arise is
always the second vowel segment. The examples in (27) below are illustrative:
In all the examples, the second vowel segments are truncated thereby
shortening the lengths of the first syllables and subsequently reducing their
weights from heavy bimoraic stems to light monomoraic stems. This system of
sequence of vowels [V1] and [V2], syllable truncation eliminates [V2] leaving
behind [V1] at syllable final position of the first stem of compounded forms.
95
In compounded forms, all consonants at syllable final position in the first stem
usually get deleted while initial and final consonants in the second syllable are
left behind. The behaviour of syllable final nasal consonants in first stems
however deserves comment. When they occur after long vowels in first stems
they are often deleted in compounded forms. However, when they occur after
short vowels they become homorganic with the initial consonants of adjoining
(second) stems thereby evading deletion (c.f. 3.2.2 above on HNA). The
In the examples above, the final consonant segments in the first stems are
truncated to allow for a smooth transition in the novel usages the language has
been put by means of putting these original stems together. We formulate the
rule in (30) below to explain consonant deletion in Kusaal. The rule states that
words:
The phonology of compound forms in the language is also very systematic and
proceeds always in a specific pattern. For instance, when two existing stems
both vowel and consonant segments in order to allow for an agglutination that
fits into the syllable structure of the language. In the following examples,
syllable truncation proceeds in a feeding order that specifies first that, final
consonant segments in first stems be deleted and then long vocalic segments be
(31) we establish the order of the process of truncation while in (32) we cite
(32) Stem 1 2
Aside the deletion of final consonants in first stems of compound words, some
nominal stems which have the voiced velar stop // occurring at syllable final
position, the stop can usually be deleted if two special conditions are met:
These facts are captured in the examples in (33a and b) which are acceptable
variants of the same lexical items. The forms in (34) are however not
admissible in Kusaal because they do not pass the second criterion (ii) above:
(33) A B Gloss
bi: bi: „child‟
(34) A B Gloss
pao *pa: „bark‟
3.3.2 Aspiration
associated with voiceless stops. As Bodomo (1997: 12) reports for Dagaare, in
Kusaal too, “voiceless plosives are usually aspirated when they occur in
primary syllable initial position.” In particular, the consonants [p, t, and k] are
the candidates which are produced with a greater degree of force when they
state of the glottis. Examples are provided on the table (1) below. The labial-
velar plosive /kp/ is not specified on the table because it occurs only at syllable
ASPIRATED UN-ASPIRATED
Aspirated syllables are indicated by the superscripted //. While the set to the
right highlights the environments in which the voiceless plosives are aspirated
The phonological rule in (34) captures the facts on the table above. It states that
99
initially.
3.3.3 Glottalisation
many of the related Mabia languages. It refers to a “process where the primary
glottal level” (Laver 1994: 330). This process occurs with reference to the
realisation of the glottal stop // in certain positions within syllables. The
glottal stop is produced with a closure of the vocal folds thereby impeding
briefly the flow of pulmonic airstream out of the vocal tract. It is realised only
as a voiceless consonant.
such as the voiceless alveolar /t/ in words like „city‟ [sʔt]. At other times, oral
stops may be completely replaced by the glottal stops as in the example for
"Britain" pronounced [bɹɪʔən]. The case of the glottal stop in Kusaal is unlike
either of the foregoing. Specifically, it does not co-occur with any consonants
noted by Ladefoged and Maddieson (1996: 75) a complete glottal closure is not
achieved but rather “some compressed form of creaky voice or some less
100
glottal stop is usually inserted to simplify the length of the vowel segments.
vowels. By inserting the glottal stop, the sounds may be pronounced as though
they were two syllables with the glottal occurring at syllable initial position in
the second syllable. The glottal stop /ʔ/ is indicated by an apostrophe (‟) in the
orthography and this is shown in the data in (36) below. Glottalisation is more
exemplify the way the lexical items are re-syllabified in the language after
glottal insertion has occurred in keeping with moraic theory (see Hayes 1989).
µ µ µ µ µ
„chest‟
Across languages, lexical items are constantly being borrowed, i.e. they are
Kusaal has a large number of linguistic items that are borrowed from English
and Hausa which are two of the most dominant languages in the Kusaas
traditional area. These loanwords are configured so that they fit snugly into the
linguistic systems of the indigenes without deviating from what is the case in
deleted or replaced in order to simplify them so that they fit into the borrowing
clusters or vowel sounds in the source language are reconfigured such that they
compensate for such deleted segments, original consonants are replaced or new
which generates either long vowels or diphthongs to make up for the deleted
sounds. The English borrowings in (38) and (39) below provide examples of
The examples in (39) also illustrate another process of commuting lexical items
from other languages into Kusaal. This involves replacing consonant systems
in the source language with the nearest consonants in the target language as a
often without some hilarious renderings. Forms like these are „shortcuts‟ that
speakers of Kusaal including some of the literate English users, adopt when
they attempt to give renditions of words that contain the post-alveolar fricatives
// and // which are conspicuously missing from the Kusaal consonant chart.
3.3.4.2 Epenthesis
A fourth process that is probably one of the most important in Kusaal loanword
initial, medial or final positions are often simplified through the insertion of
one vowel sound or the other. Examples of such occurrences are widespread in
below in (41):
borrowed lexical items are subjected to the assimilatory and syllable structure
(38 – 41) for the discussions in this and the subsequent sections. In (42) for
of the low back and the mid-low central vowels in the first and second
examples respectively. The process results in vowel glides from high to low
positions. In the third example, the schewa vowel // which features as a
variation in the second syllable of table [teibl], is replaced by the high round
vowel. This results in the rounding of the voiced bilabial and alveolar plosives
stop insertion is realised between the long vowels. This leads to a restructuring
of the syllable with the result that the glottal stop goes into onset position of the
second (new) syllable. In the first three examples, the process is much more
In the last set of examples in (43), voiceless plosives in borrowed items are
initial position. Syllable final plosives do not acquire aspiration. This trend is
3.4 SUMMARY
segments and/or words assimilate each other in the language through processes
the processes that are on-going in the structuring of syllables in the language.
CHAPTER FOUR
TONOLOGY
4.1 INTRODUCTION
In this chapter, we explore the nature and types of tones in Kusaal within an
tones and the ways in which these tones are combined in the language together
with their distribution. In this chapter too, we establish the tone bearing unit
(TBU) in Kusaal and go on to discuss the nature of tones in the nominal and
verbal morphology of the language. We also describe and analyse the role of
tone in the syntax of the language by examining other linguistic structures such
contrastive, but relative pitch on each syllable” (Pike 1948:3). Pitch refers to
the rate of vibration of the vocal folds. Varying forms of this pitch can be
decreasing the flow of air out of the lungs (Ladefoged 1982: 226, 2001: 233).
These varying pitches are subsequently used to create lexical contrasts, i.e., to
differentiate one lexical item from the other. Pitch is also relative because
circumstances.
107
Further, the pitches which are used to create such meaning differentials are
phonemes do. Tone is not only crucial in creating lexical differences, it is also
Though relative pitch is a property of the languages of the world, the disparity
the former property are characterised as “intonational” (see for e.g. Ladd
1997), those which use pitch for the latter are called “tonal”. English, together
with a great many European languages, are intonational, while Chinese and
Distinctions are also made in the linguistic literature between register tone and
contour tone languages. Languages which have register tone use level pitches
are usually specific to one height position: high, low, or mid, occurring on the
tone bearing unit. In addition to these level pitches, register tone languages also
have gliding pitches occurring on the relevant tone bearing unit. These glides
are usually sequences of level pitches that are contiguous to each other and
could therefore be rising (i.e. a sequence of low – high tones) or falling (from
high –low tones). Contour tone languages on the other hand are not level or
fixed to specific heights but are characterised by pitch glides from one height
position to another on the same tone bearer. The languages of Africa usually
108
draw on level pitches to create tonal differences while most Asian languages
use tones that are contour (Pike 1968, Hyman 1972, Yip 1995, 2002).
such as Mampruli, Dagbani and Gurun (Naden 1988), Moore and Lama
have shown conclusively that the majority of these languages have two
tonemes. On Buli too, Akanlig-Pare (1994 inter alia) posits a two tone system:
the high and the low that characterise the languages in the cluster. However,
quite a different phenomenon from what obtains in the cluster where the trend
is not only for languages to have a two tonal system but to also distinguish a
The tonal system of Kusaal is similar to that of Buli in one regard, it contrasts
three underlying tones which are high (H), mid (M) and low (L) tones. It
underlying high tone. Generally, high toned syllables are produced with higher
pitches than mid tones which are in turn also higher in pitch than the low toned
ones. With the aid of the Praat acoustics software, we are able to determine that
the fundamental frequency (f0) of Kusaal high tones ( ) are initiated at 150Hz
109
and above. Mid tones ( ) oscillate between 120 and 150Hz, and low tones ( )
few lexical items are differentiated in the language based only on their relative
triples in all cases. We draw on both nominals and verbs to make the contrasts
in (1). In (2), we further exemplify the tonal differences by showing the pitch
tracings of three lexical items (the dotted lines go through the highest pitch
(1) HIGH
ba „father‟ ku „kill‟
MID
bi: „child‟ ya „houses‟
LOW
ba 3PL. ya 2PL
) z
)
z
z
(
H
H (
( h
h
c
h
t
c
i
c t
i
t
i P
P
P
75 75 75
0 0.310635 0 0.603719 0 0.368141
Time (s) Time (s) Time (s)
The high, mid and low tones are combined in the language to produce nine
separate sequences of tones on disyllabic words. These are the HH, HM, HL,
MH, MM, ML, LH, LM and LL. On polysyllabic words too, the tonal
occurrences are not restricted and thus a combination of these tones is evident
(3) HH
pi:a „ten‟ kurug „shorts‟
HM / MH
Wina:m „God‟ bp „daughter‟
HL / LH
krw:k „trousers‟ na’anam „chiefs‟
MM
bana: smock tampi: „rock‟
ML / LM
ylm „tell!‟ anu „five‟
LL
kr old zlm „tongue‟
Determining the elements or structural features which carry the tonal melodies
1995: 448). The dicey nature of this phenomenon is also articulated by Yip
(2002: 73) who affirms that “it‟s not always clear whether tones associate to
In some of the approaches to tone studies, the element on which tones dock is
autosegmental framework all vowels should be associated with tones and all
have proceeded to equate the tone bearing unit (TBU) to the vowel or even to
consonants. In Fox (2000), for instance, studies such as Gandour (1974) and
Schachter and Fromkin (1968) are cited as establishing consonant sounds and
vowel segments as the units which carry the tonal melodies of Thai and Akan
respectively.
This approach though seamless at the onset is faulted in one important regard.
It fails to account for those instances in the phonology of tone languages where
features, this view of the TBU as a segmental feature is untenable (Bao 1999:
134, Fox 2000: 218). We replicate below an example by Yip (2002: 67) on a
representations, the tonal features that are linked to them do not move but
remain in the same position and are then associated with the new segments.
In the build up to the discussions on tone bearing ability, Clements and Ford
(1979: 181 in Ahoua 1986: 74) assert that “…tones are not directly associated
113
with vowels or segments, but rather with higher level units…such as the
similar stance is also taken by Pike (1948: 4) who regards the syllable as the
domain of tones in languages. In other works, too, the TBU is regarded not as
the syllable but as the mora. The following views are very instructive in
determining whether the syllable or the mora constitutes the tone bearer:
With insight from the above, it is not difficult to posit the mora as the tone
bearing unit in Kusaal. Evidence that may be used to back this claim is offered
by the fact that various perturbations of tone in the language are restricted to
whether the affected syllables are light weight or heavy weight ones. While
tones are also realised only on heavy bimoraic syllables. These contour tones
(LH, LM, MH, HM, HL, or HM), are a combination of two level tones realised
heavy bimoraic syllables are not marked for tone. The structural
representations in the following (5) and (6) are used to exemplify the foregoing
observations.
(5)ab.
L M
µ µ µ
l a „laugh‟ b i „child‟
The above configurations depict that the suprasegmental feature docks on the
mora in each instance as a low tone in (a) and as a mid tone in (b). In (6)
below, the reason for citing the mora as the TBU in Kusaal is made clearer by
the realisation of a falling tone on the heavy weight bimoraic syllable saa
(6)ab. *
H L H L
µ µ µ
s a „rain‟ b EXISTENTIAL
The asterisk in the (b) example above indicates that situations whereby two
tonemes map unto one light monomoraic segment are inadmissible in the
syllables become bimoraic to enable them carry the extra tone. For instance,
the light monomoraic syllable p „bury‟ carries only one tone while two
tonemes are mapped unto the verb pia „speak‟ because it is a heavy weight
syllable containing two vowels. Contour tones will usually occur only on heavy
bimoraic syllables.
The rest of this chapter takes as its basis, the autosegmental representation that
resulting discourse on the topic (see § 1.10.2 above for an in-depth discussion).
Different tiers will be used to show the various levels at which the phonology
of Kusaal operates with the various elements specified for each of these tiers.
Segmental features and tones on the separate tiers will also be mapped onto
vowels (Goldsmith 1990: 14) but to “tone bearing units (a) from left to right
Similarly, we will, in consonance with Yip (2002: 76), modify the initial
to rather read as follows; the point of departure from Goldsmith is the entity
section, we will discuss the role of tone in nominal morphology. The types of
tones which occur on noun roots in Kusaal are not constrained. It is possible
then, to have either high, low, rising or falling tones occurring on various
very difficult to predict which tones can be realised and on what nominal form.
To a very large extent however, nouns usually carry non-low tones in the
language.
Singular forms of Kusaal nominal items usually carry different tones. Lexical
items are thus not specified in relation to which tones can occur on them in the
singular stems are retained and replicated on the corresponding plural forms.
segments and so the number of tones that are found on singular roots are
usually also required in the plural. If a noun root with long vowels has a
uniform sequence of tones, we will mark only the first one. If the sequence of
tones on long vowels is different, we will however, mark both of them in this
Nominal items can normally be derived from other categories of speech. These
derived items usually take on many of the properties of nouns and are used as
such. Nominal forms can be derived from the lexical classes of words such as
which nouns are typically derived is the verb class. Several nominal items in
from suffixing a nominalising affix to existing verb stems. We will describe the
They are derived from action verbs. Such nominals, as stated by Appah (2003:
verb. He provides the examples below (8) on Akan (Fante) to clarify this
(8) Fante
a. kenkan a-kenkan
„read‟ SG-read
„the act of reading‟
b. dzidzi a-dzidzi
„eat‟ SG-eat
„the act of eating‟
obstruent sounds take the suffix -u. Regularly, these derivations transpose
recurring low or mid tones on bare verbs into high tones in the nominal while
high toned verbs retain their high tones in the derived form. Single short
b. pia pia-b
c. kua kua-b
d. ka: s ka:s-u
e. yu:m yu:m-u
refer to the agents that are depicted by the verb roots from which they are
derived. The derived forms are then understood as meaning “one who executes
the action of the verb” (cf. Appah 2003: 73). Agentive nominals are derived in
Kusaal by suffixing the morphemes –r to verbs for singular and –rib/-ib for the
plural forms. In this case too, the nominalising suffix changes recurring low or
falling tones on verbs to high in the derived form or it retains high tones on
verbs in the derivations. As in the above, short vowels in light weight syllables
b. nu nu:-r
„drink‟ nuu-AG.NOM
„one who drinks‟
c. ku ku:-rib
„kill‟ ku-AG.NOM
„people who kill‟
120
d. b:d b:d-b
„want‟ want-AG.NOM
„people who like/want‟
Derived agentive nominals may often collocate with other nouns or they may
be used as they stand in the above. When they collocate with other nouns, they
point exactly to the activity that is being referred to thereby making it more
noun such as da:m „pito‟ to nu:r „drinker‟ yields the form danu:r „pito
together two different nouns. Though nouns may bear any tone, they generally
low tones and have both singular and plural forms. They are listed in (11):
Earlier in this section, we stated that the majority of nominal items in the
Kusaal language generally bear non-low tones. When the possessive pronoun is
joined to a noun then, the low tone on the pronoun is raised through a process
of low tone raising (LTR) to become a mid tone on condition that the noun has
a non-low tone, i.e, it has either a high or mid tone. If the nominal item carries
a low tone, this tone is retained in the associative, likewise default low tones on
ba: [ba:]
prevalent phenomenon in many languages. In Kusaal, the two nouns that can
be placed adjacent each other in the associative must be related to some degree:
schematically as „the N2 of N1”, where N refers to noun and 1 and 2 are the
As these constructions comprise two nouns, the tones on the nominal items are
compared to their pitches in the bare token forms. The following (13) is
illustrative:
4.5.4 Compounding
The tonal processes that obtain in compound forms do not differ very much
syllable final segments in the first stem are truncated so that the new
123
combinations, tones on the root stems are usually retained in the new forms.
items which refer to nominal forms. This is usually achieved in some cases
with the affixation of a suffix –im which derives nominal forms from
adjectives. This suffix always carries a mid tone in compounded forms. In most
copied and then associated with all tone bearing units in the new forms. In
forms. Such compounds are simple combinations of noun and adjective stems.
Proper noun forms which may not be compounds also arise from a
concatenation of the prefix a- which is used to denote that the unit spoken
process of high tone lowering (HTL). These high tones are then realised as mid
b. L H L H L !H L M
x x x x x x x x
Tone plays a very critical role in the syntax of Kusaal. The syntactic functions
relate to the ways in which the various perturbations of tone in the verbal
In this section we will explore the way tone influences the tense, aspect and
In Buli (Kenstowicz 2004: 3), it is not possible to find verbs that can be
distinguished based only on tonal contrasts. Verbs in the language are therefore
with a caveat – in the bare infinitive, a great majority of verbs in the language
carry a non-high tone. This comes with a number of implications for the
happened or that are anticipated. Such references could also relate to the
expresses a temporal relation between the time of the described event and some
reference time” (Van Valin and LaPolla 1997: 40). Implicit in this statement is
to an event that has already happened (understood as the past); and to an event
Languages and cultures perceive time and therefore the locations within that
situations. The various approaches that languages will use in the descriptions of
the linguistic time frame will reflect their intrinsic knowledge or concept of
time. According to Botne and Kershner (2008: 150) “tense systems constitute
(1985: 9). It implies then that, tense as a grammatical category must be marked
morphological inflections may not be overt morphs but may be coded as is the
system (either overt or coded) that is inflected on verb stems to show that the
event or situation being spoken about happens in the past, present or future.
that occur in the past or present time. References to future time situations are
also achieved through the use of the particle na occurring before main verbs in
verb phrases. Various degrees of remoteness pertaining to past time and future
time references also obtain in the language and these are achieved with the use
of three particles – pa:, sa: and da: for variations of the past and by the particle
Kusaal verbal systems are therefore inherently aspectual in nature. Unlike tense
which relates to the ways in which language is used to “carve up” various
time. This view is in line with Comrie (1985: 6-7) who characterises aspect as
the temporal references made by tense and determine whether the action being
between the progressive and habitual aspects in present time references. In past
time references, too, the perfective and imperfective aspects are evident.
Mood, also mode/modality, refers to verbs that are used in connection with
sentiments of the speaker toward the issue being spoken about. It shows
whether the action in a sentence is a fact, an order, a wish or other and thus
128
(2003: 135) says of modality that it “is a cover term for devices which allow
the words of Palmer (2001: 1) modality is “concerned with the status of the
In the majority of tonal languages, the role of tone in tense, aspect and
which also precede objects in the syntactic structure of the language. The
Present time references in the language are often made in relation to either the
lengthened over time, or that occurs frequently within a period of time, the
latter refers to an activity that is ongoing within a specific time frame. To form
the habitual aspect, verb stems take various allomorphs of the suffix /-t/: which
is realised as /-d/ after verb stems which end in back vowels, as /–t/ in stems
which end in front vowels and as the suffix /-id/ in stems which end in
obstruent sounds. In habitual constructions, all the verbs bear mid tones with
the first and second person pronouns while the pronouns are all realised with a
129
default low tone. In the third person however, recurring mid tones on verbs
(17) below:
When the verb takes a direct object in the habitual, it copies the high tone of
the object in the first and second persons. Recurring high tones on the object
are subsequently realised as downstepped high tones or mid tones in all cases.
The first and second person pronouns in turn acquire mid tones through a
process of low tone rising which is triggered by the verb. Direct objects do not
influence the tonal properties of verbs with the third person pronominal
b. L M H L M H L M H
x x x x x x x x x x x x
M H M H M H !H
x x x x x x x x x x x x
In the above representation, it is observed that the mid tone of verb spreads in a
pronoun acquires a mid tone. Subsequently, the high tone of the nominal form
also spreads in a leftward direction causing the mid-tonal verb to become high
toned. The high tone on the noun is realised as a down stepped high tone in the
final derivation.
Forms in the progressive aspect also differentiate between those that have
objects from those that do not. The progressive aspectual marker is a suffix -
affixed to present or habitual forms of verbs. The default tone on this suffix is a
non-high tone – either mid for the 1st and 2nd persons or low for the 3rd person
pronouns respectively. The suffix however becomes high when there is a direct
object in the sentence. The examples below exemplify null object (19a) and
Past tense forms in Kusaal are always explicated with an inherent reference to
describes an event that was carried out in the past and that has been
(20a) or with one of three particles: pa: (hodiernal past time), sa: (recent past
time) and da: (remote past time) which designate various degrees of past time
reference (20 b - d). In the following we illustrate how tone patterns out in
1SG bath.PERF
„I have bathed.‟
2SG bath.PERF
„You have bathed.‟
Name bath.PERF
„Adaam has bathed.‟
132
The examples in (20a and b) make reference to a situation occurring within the
same time frame. We explain that in (a), for instance, a high tone is borne by
the verb stem that inflects for time. When the particle pa: is introduced in (b) a
number of tonal processes are triggered. First, the particle copies the high tone
of the inflected verb. Second, the high tone on the verb is deleted and it
subsequently acquires a low tone through a process of low tone spread (LTS)
(20c) however, the tone on the preceding particle causes a lowering of tones in
the verb stems and suffixes while in (20d) all the segments copy the underlying
(21) Tone M L H L M L H L M L H L
Skeletal x x x x x x x x x x x x
Segment m pa so ja m pa so ja m pa so ja
M H L M H L M H L
x x x x x x x x x x x x
m pa so ja m pa so ja m pa so ja
134
In order to form the interrogative from the perfective, the short final vowel
segment in the affix acquires a long vowel quality. This creates a domain for
falling contour tones (a sequence of high - low tones) to dock on. Penultimate
Distinctions between future time marking and the future interrogative are made
based on only tone contrasts in the language. In much the same way as the
indicate that the event being spoken about occurs in a timeframe ahead of the
present moment of speaking. It is used pre-verbally and bears a high tone in 1st
and 2nd pronominal forms. This high tone conditions adjacent low tones on 1st
and 2nd pronouns to rise to mid tones. Verbs retain their default non-high tones.
degree of tone rising until the end of the statement where short final vowel
segments in the verb stem are lengthened. The second mora in the verb always
takes a low tone. Observe also that mid tones in the first and second person in
the statements above become high toned in the question while low tones on the
third person become mid-toned. The examples in (24) illustrate this strategy:
interrogative particle -:/e: is suffixed to the verb stem so that it can carry the
Kusaal perfective and future forms can both be used in a negative sense. This is
p/k. When a low tone is associated with the moraic segment in the
morpheme - p/k - the derived meaning is „will not‟. On the other hand, when
these morphemes carry high tones - p/k - a negative meaning „did not‟
arises. Verb roots which occur in both forms acquire an extra long quality if
they end in short vowels. If they end in consonants, verb stems are re-
syllabified by a clitic -. The first moraic elements in the verb stems also copy
the tone of the negation marker. We illustrate this phenomenon in the following
examples in (26):
Imperative constructions are used to make demands of other people over whom
the speaker has some degree of authority or power. The imperative suffixes are
–m/-im for singular constructions and -min for plural imperative constructions
137
respectively. A verb takes the suffix –m in the imperative if the stem ends in
vowel sounds which can be either a sequence of vowels or long vowels but
never short vowels. Short vowels are regularly transmuted to long ones so that
they can take the suffix. The –im suffix is conjoined to verb stems which end
of non-high tones. Final tones in the singular imperative are generally low
toned while in the plural, the imperative suffix bears a mid-tone. The examples
4.7 SUMMARY
identified that much like the case in the Buli language; Kusaal has a three tone
contrast - high, mid and low. Unlike Buli, however, there are instances of high
tones which are realised as downstepped highs because of adjacent high tones.
We identified that the tones of nominal items are regularly non-low, i.e., that
they bear either high or mid tones, while uninflected verb stems are regularly
mid-tonal or low. The chapter further discussed the role of tone in Kusaal
138
also discussed the distinctions that are created in the syntax of the language
marking the different tense, aspect and modal systems of the language.
139
CHAPTER FIVE
5.1 INTRODUCTION
research on the language. The study is divided into five broad sections -
5.2 SUMMARY
Chapter One set the basis for the study. Kusaal, we intimated, is one of the
Though some research has been carried out on the language, the scant material
various languages of the world, we stated that there was a need to change the
status quo and undertake an up-to-date description of the language. The scope
some 450,000 Kusaas who inhabit the north-eastern corner of Ghana. This area
is bounded to the north by Burkina Faso and to the east by Togo where there
from the White and Black Voltas which run across the three Kusaal speaking
The language is found in the Mabia cluster under the Proto Oti-Volta grouping
of the larger Niger-Congo language family. It has two dialects – Agole and
Tonde which are spoken in the east and west of the language area respectively.
The indigenes are typically traditional African religion adherents who believe
in small deities and in a supreme being called Winaam. Christianity and Islam
are also common among the people. Kusaog – the traditional name of the
people, at the same time, hold a strong cultural affinity to the language.
The significance of the study include updating and adding up to the obsolete
language policy makers and translators; setting the basis for subsequent and
In this chapter too, we highlighted the theories which form the basis of the
and Halle (1968) was drawn upon because of our belief that the linguistic
competence of a speaker, and not their ability to simply perform in it, directs
him/her to choose the optimal form from a huge list of potential forms
141
generated by the syntax of the language (Kenstowicz nd: 1). The autosegmental
of the short falls of the generative theory, especially with regard to the linearity
are able to posit that certain elements are autonomous, unlike segmental
elements such as stress and intonation, and in our own case, tone. With moraic
theory, we are able to break down the atomic composition of Kusaal syllables.
units rather than to their immediate constituents (such as onset and rhyme).
The methodology that was adopted for the study involved embarking on a 10
day field trip to the general Kusaal speaking area. Using an electronic recorder,
we interviewed 6 respondents from each dialect with the aid of a word list. We
also recorded spontaneous speeches in both religious and social settings. The
collected data was then transcribed and, using Praat, we determined the relative
pitches of some of the words and constructions. Native speaker intuition was
also an important asset in this endeavour. Other native speakers and a few
secondary sources were often consulted for clarity and accuracy of data
language. We identified 24 consonants: three bilabials - /p/, /b/, /m/; two labio-
dentals - /f/, /v/; seven alveolars - /t/, /d/, /n/, /s/, /z/, /l/, /r/; three palatals -//,
/j/, //; three velars - /k/, /,/; four labial-velars - /kp/, /b/, /w/, /w/ and two
142
glottal sounds - // and /h/. At word initial position, all except the glottal stop
// and the velar nasal // can occur while word medially only the glottal
fricative /h/ does not occur. At word final position the labial-velar and glottal
stops /kp/, /b/, and //; the labio-dental and velar nasals // and //; the voiced
labio-dental and alveolar fricatives /v/ and /z/ and the approximants /j/ and /w/
There are also 9 short oral vowels - /i, , e, , a, u, , , o,/ which have long
vowel correlates - /i:, :, e:, :, a:, u:, :, :, o:/. Five nasal vowels which are all
[-ATR] also obtain in the language - //, //, /a/, // and // and there are also
The distinct features which underlie these sounds are three: the major class
[ATR], [voice] and [spread glottis]. All these features have binary values
except [coronal], [dorsal] and [labial] which are unary in nature, i.e., they are
specified only if they contribute to the nature of the sound under consideration.
able to deduce that if a vowel sound has a nasal quality in the language then it
perspective. Four broad syllable types were identified: the peak only syllable
which can be either a vowel or nasal consonant; the heavy VC syllable which
also extends to become CV: or CVV in some cases and finally the CVC
syllable which can take up to three different forms: CV:C, CVVC and
CVVCC. Words in the language can thus be mono-, di- or tri- syllabic or they
each other, they behave like each other in the same environment. Assimilatory
assimilation (HNA) through which nasal sounds copy the place of articulation
word; palatalisation which raises the body of the tongue in the production of
observed that the vowels which occur in syllables or words always come from
144
the same category. The harmonisation cuts across the various vowel heights
and is tied to the position of the tongue root in speech production. It thus
syllables thus occur only with members of their group but not with others. The
[-ATR] a
It was also observed that the only [-ATR] vowel which co-occurs with [+ATR]
vowels is the low central vowel /a/. It was further observed that variations of
words produced with either set of the vowels are often possible in the language.
In most situations however, the language users are prone to select vowel forms
from the unadvanced tongue root than from the advanced class in speech. This
they fit into the language especially when words are compounded. Aspiration
language. Loanwords are also integrated into the language by processes such as
language. Tone is the use of relative pitch to create lexical and grammatical
145
which has only level pitches and thus contrasts three tonemes – high (H), mid
(M) and low (L). The language also differentiates between underlying high
tones and downstepped high tones. Gliding pitches which occur in the language
are considered as sequences of these level pitches which create rising or falling
HM, HL, MH, MM, ML, LH, LM and LL especially on disyllabic words.
We identified the mora as the tone bearing unit (TBU) in Kusaal because
the language.
singular and plural forms, underlying tones are often retained while action or
agentive nominals derived from verbs always acquire high tones. In associative
are raised to mid tones through LTR while nominal forms become high toned if
they do not bear high tones already. In noun plus noun associative
constructions, the underlying tones on these nouns are retained while the
Compound forms usually also retain their tones. Tone also plays a very crucial
role in the syntax of the language. The various aspectual and modal systems of
146
the language are also often differentiated based on the underlying patterning of
tone. The habitual, the progressive, the perfective and interrogative, the future
and interrogative, the perfective and future negation and the imperative have
underlying tonal perturbations that mark them as different, one from the other.
5.3 RECOMMENDATIONS
From the onset, we acknowledged that this study scratches only the surface. It
is only the icing on a rich bulk of linguistic repertoire waiting to be tapped and
savoured. There is therefore a lot more work that can be done on Kusaal. The
language offers a range of options for researchers who are willing to venture
the following:
sociolinguistics, etc.;
investigate into the oral traditions of the people cached in their folklore
determine how much they converge or diverge from each other and
final consonants;
5.4 CONCLUSION
To conclude, we will iterate that this study covers only the phonology of
Kusaal and that there is a lot more that can be done on the language. Areas that
could be researched into have been identified and listed. It is believed that a
work on other aspects of the language. It is also hoped that this study would
BIBLIOGRAPHY
Abdul-Karim, Azure and Thomas Mbum (2001). Kusaal primer: Gbau ayi
Abdul-Karim, Azure and Thomas Mbum (2001). Kusaal primer: Gbau nwa ni
University of Ghana.
more evidence for the tone mapping rule with reference to Baule. In:
Bogers, Koen, Harry van der Hulst and Maarten Mous (eds.), The
pp. 63 -78.
University of Ghana.
Akanlig-Pare, George (2005b). The tonal structure of Buli verbs. In: Kropp-
Dakubu, M.E. and E.K. Osam (eds.), Studies in the languages of the
Bao, Zhiming (1999). The structure of tone. Oxford: Oxford University Press.
Publications.
Botne, Robert and Tiffany L.Kershner (2008). Tense and cognitive space: On
Broselow, Ellen (1995). Skeletal positions and moras. In: Goldsmith John
Chomsky, Noam and Halle Morris (1968). The sound pattern of English. New
Clements, George N. and Ford, Kevin C. (1979). Kikuyu tone shift and its
Cohn, Abigail (2001). Phonology. In: Aronoff Mark and Rees Miller Jamie
Oxford: Blackwell.
Dolphyne, Florence Abena (1988). The Akan (Twi-Fante) language: Its sound
Ewen, Colin J and Harry van der Hulst (2001). The phonological structure of
Maddieson, Ian (ed.), The tone tome. Studies on tone from the UCLA
http://ghanadistricts.com/districts/?r=8&_=102&rlv=climate. 15/11/09
151
Ghananet.com (2010).
http://ghananet.com/documents/upper_east_ghana_districts.bmp. 20th
February 2010.
MIT.
Blackwell.
Hyman, Larry M. (1975). Phonology: Theory and analysis. New York: Holt,
Hague: Mouton.
Blackwell.
Studies. pdf
http://web.mit.edu/linguistics/people/faculty/kenstowicz/generative_
Description and theory. Orlando: Academic Press Inc. Also online at:
http://www.sil.org/linguistics/GlossaryOfLinguisticTerms/WhatIsGener
pp. 77-103.
Kropp-Dakubu, Mary Esther (2006). Tone and the Gurune verb. In: Kropp-
University Press.
Harcourt College.
153
Ladefoged, Peter and Ian Maddieson (1996). The sounds of the world’s
Press.
http://www.sil.org/linguistics/glossary_of_lingusitic_terms_html.
Naden, Tony (1988). The Gur languages. In: Kropp Dakubu, M.E. (ed.), The
Naden, Tony (1989). Gur. In: Bendor-Samuel, John and Hartell, Rhonda L.
family. Lanham: University press of America and SIL. pp. 141 - 168.
Odden, David (1995). Tone: African languages. In: Goldsmith John (ed.), The
444 – 475.
Palmer, F.R. (2001). Mood and modality (2nd edition). Cambridge: Cambridge
University Press.
Blackwell Publishing.
Inc.
Spratt, David and Nancy Spratt (1968). Collected field reports on the
Institute of Linguistics.
Spratt, David and Nancy Spratt (1972). Kusaal Syntax. In: Collected language
Timberlake, Alan (2007). Aspect, tense and mood. In Timothy Shopen (ed.),
Van der Hulst, Harry (1984). Syllable structure and stress in Dutch. Dordrecht:
Foris.
Van Valin, Robert D. Jr. and LaPolla J. Randy (1997). Syntax; structure,
Westermann, Dietrich and Bryan M.A (1952). The languages of West Africa.
2009.
Williamson, Kay and Roger Blench (2000). Niger Congo. In: Heine, Bernd and
Yip, Moira (1995). Tone in East Asian languages. In: Goldsmith John (ed.),