A Study of Phonological Processes in Wassa

University of Education,Winneba http://ir.uew.edu.
gh
UNIVERSITY OF EDUCATION, WINNEBA
A STUDY OF PHONOLOGICAL PROCESSES IN WASSA
CLEMENT KWAKU KUMI
2016
University of Education,Winneba http://ir.uew.edu.gh
UNIVERSITY OF EDUCATION, WINNEBA
A STUDY OF PHONOLOGICAL PROCESSES IN WASSA
CLEMENT KWAKU KUMI
(8140080011)
A THESIS IN THE DEPARTMENT OF APPLIED LINGUISTICS,
FACULTY OF FOREIGN LANGUAGES EDUCATION AND
COMMUNICATION, SUBMITTED TO THE SCHOOL OF GRADUATE
STUDIES, UNIVERSITY OF EDUCATION, WINNEBA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR AWARD OF THE
MASTER OF PHILOSOPHY (APPLIED LINGUISTICS) DEGREE
OCTOBER, 2016
DECLARATION
STUDENT’S DECLARATION
I, Clement Kwaku Kumi, declare that this thesis, with the exception of quotation
and references contained in published works which have all been identified and duly
acknowledged, is entirely my own original work, and it has not been submitted, either in
part or in whole, for another degree elsewhere.
SIGNATURE………………………………………………
DATE…………………………............................................
SUPERVISOR’S DECLARATION
I hereby declare that the preparation and presentation of this work was supervised in
accordance with the guidelines for supervision of thesis as laid down by the University of
Education, Winneba.
NAME OF SUPERVISOR…………………………………..
SIGNATURE…………………………………………………
DATE………………………………………………………...
ii
ACKNOWLEDGEMENTS
I am most grateful to God Almighty for the achievement of this academic feat. If
not by His abundant grace, I would not have come this far.
I express my sincere gratitude to my supervisor, Prof. Thomas Essilfie for his
comments and guidance that helped to improve upon the quality of this work. Despite his
busy schedule, anytime I approached him for assistance, he welcomed me with alacrity. I
am very appreciative of his contribution towards the success of this work. God richly bless
you. However, any errors or shortcomings in the work are solely mine.
I am also grateful to my consultants, Mr. Kofi Abakah and Mr. William Andoh for
their assistance, encouragement, and pieces of advice during my fieldwork. They did not
only cross-check my data for its authenticity, but they gave me numerous examples and
useful information which assisted me greatly in the writing of this work.
My next thanks go to my lecturers, Prof. Lawrence Boadi, Dr. Kwaku Ofori, Dr.
Rebecca Akpanglo-Nartey, Dr. Charles Owu-Ewie, Dr. Paul K. Agbedor and Dr. Fofo
Lomotey for their patience, love, and pieces of advice in the course of my study. Special
thanks go to my colleague graduate students.
I wish to take this opportunity to express my heartfelt thanks to my wife, Tawiah,
for her unflinching support, prayers, love, care and understanding in the course of my
study.
Most of the time, when she needed me most, I was not there for her. To my father,
my late mother, siblings and to the whole family.
Finally, I am thankful to all those who helped me in diverse ways in the course of
my study, but whose individual names have not been mentioned here for the sake of space.
iii
DEDICATION
I dedicate this thesis to my family especially, my wife and children.
iv
TABLE OF CONTENTS
CONTENTS PAGE
DECLARATION ii
ACKNOWLEDGEMENTS iii
DEDICATION iv
TABLE OF CONTENT v
LIST OF TABLES viii
LIST OF FIGURES ix
GLOSSARY AND ABBREVIATIONS xi
ABSTRACT xviii
CHAPTER ONE: INTRODUCTION 1
1.1 Overview 1
1.2 Background of the Study 1
1.3 The Wassa People and their Language 2
1.4 Statement of Problem 9
1.5 Objectives of the Study 9
1.6 Research Questions 10
1.7 Significance of the Study 10
1.8 Limitation 10
1.9 Delimitation of the Study 11
1.10 Organization of Thesis 11
v
CHAPTER TWO: LITERATURE REVIEW 12
2.1 Overview 12
2.2 Generative Phonology 12
2.3 Autosegmental Phonology 13
2.4 Basic Tenets of Autosegmental Phonology 15
2.4.1 Skeletal Tier 15
2.4.2 Universal Association Convention (UAC) 16
2.4.3 Linkage Condition 18
2.4.4 Obligatory Contour Principle (OCP 19
2.4.5 Well Formedness Condition (WFC) 20
2.5 Motivation for Autosegmental Phonology 21
2.6 Tiers Used in this Thesis 21
2.7 Distinctive Features 23
2.7.1 The Major Class Features 23
2.7.2 Cavity Features 24
2.7.3 Manner Feature 25
2.8 Literature Review 29
CHAPTER THREE: METHODOLOGY 37
3.1 Overview 37
3.2 Research Design 37
3.3 Selected Site 38
3.4 Population and Sampling 39
3.5 Data Collection 40
3.6 Data Analysis 40
vi
CHAPTER FOUR: RESULTS AND FINDINGS 41
4.1 Overview 41
4.2 Assimilation Process 41
4.2.1 Vowel Harmony 42
4.2.2 Consonant Nasalization 52
4.2.2.1 Nasalization of Voiced Plosives and Affricates 53
4.2.3 Consonant Mutation 55
4.2.4 Labialization of Consonants 57
4.2.5 Palatalization 61
4.3 Syllable Structure Process 63
4.3.1 Elision 63
4.3.3 Insertion 71
4.3.4 Compensatory Lengthening 73
CHAPTER FIVE: SUMMARY OF FINDINGS, CONCLUSIONS AND
RECOMMENDATONS 81
5.1 Overview 81
5.2 Summery of Findings 81
5.3 Conclusions Drawn from the Study 82
5.4 Contribution to Knowledge 82
5.5 Suggestions for Further Studies 83
REFERENCES 84
APPENDIX 86
vii
LIST OF TABLES
PAGE
1: Oral Vowel in Word Examples 4
2: Nasal Vowel in Word Examples 5
3: Consonant Sound in Word Examples 6
4: The V Syllable Type 7
5: The CV Structure. 8
6: The C Structure. 8
7(a): Presents the Distinctive Feature Matrix for Vowel Sounds of Wassa 27
8: Presents the Distinctive Feature Matrix of Consonant Sounds in Wassa 28
viii
LIST OF FIGURES
PAGE
1: Family Tree Diagram of the Volta-Comoe Language 3
2: Wassa Vowel Chart 4
3: Consonant Chart 5
4: Structural representation of V syllable shape. 7
5: CV Syllable Shape. 8
6: Linear representation of segmental and suprasegmental features. 13
7: Autosegmental Representation of segmental and suprasegmentals features. 15
8: Skeletal Tier. 16
9: Universal Association Convention. 17
10: Linkage Condition. 18
11: Obligatory Contour Principles. 19
12: Well Formedness Condition 20
13: [+ ATR] Vowels within Root Morpheme. 43
14: [- ATR] Vowels within Root Morpheme. 44
15: Vowel Harmony across Morpheme Boundaries. 47
16: Vowel Harmony across Word Boundaries. 48
17: Vowel Harmony and Raising in Amenfi. 50
18: Nasalization of Voiced Plosives. 53
19: Bialabial b/m Lenition. 55
20: Cosonant Mutation. 56
21: Consonant Labialization. 58
ix
22: Autosegmental representation of Consonant Labialize before Non-rounded Vowels 60
23: Palatalization of Consonants. 62
24: Deletion of High Vowels. 65
25: Negative Vowel Deletion. 66
26: Deletion of final mid vowels. 67
27: Deletion of Intervocalic Nasal 69
28: Future Consonant Affix Deletion 70
29: Insertion of High Vowels 72
30: Progressive Affix Lengthening 75
31: Perfect Affix Lengthening. 78
x
GLOSSARY AND ABBREVIATIONS
Affricate: A phonetic segment which consist of a phoneme. Each segment has different
realizations which are only partly distinguishable for speakers.
Alveolar: A classification of sounds which are formed at the alveolar ridge (the bone plate
behind the upper teeth). Alveolar sounds are formed with the tip or the blade of
the tongue.
Alveolo-palatal: A classification of sounds which are formed with the hard palate as
passive articulator and the blade of the tongue as active articulator.
Bilabial: Any sound produce using both lips.
Consonant: On of the two main classes of sounds. Consonants are formed by a constriction
in the supra-glottal trade.
Context: A term referring to the environment in which an element (sound, word, phrase)
occurs.
Diachronic: Refers to language viewed over time and contrasts with synchronic which
refers to a point in time.
Economy: A principle of linguistic analysis with demands that rules and units are to be
kept to a minimum.
Fricative: A type of sound which characterized by air passing a constriction somewhere
between the glottis and the lips.
Generative: A reference to a type of linguistic analysis which relies heavily on formulation
of rules for the exhaustive description (generation) of the sentence of a
language.
xi
Glide: A sound which from the point of view of phonological classification lies between a
vowel and a consonant. It is formed with little friction and has a high degree of
sonority which accounts for why glides are formed near the undams of syllables,
sometimes called a semi-vowel.
Homorganic: Any set of sound which are articulated at the same point in the vocal tract.
IPA: A system of transcribing the sounds of the languages which consists of some Latin
and Greek letters and a variety of additional symbols and diacritics.
Labial: A reference to a sound which is formed at the lips. It encompasses both bilabials
and labio-dentals.
Labio-dental: Describes a consonant which is formed by the lower lips making contact
with the upper teeth.
Language: A system which consist of a set of symbols (sentences) – realized phonetically
by sounds – which are used in a regular order to convey a certain meaning.
Manner of Articulation: One of the three conventional parameters (the others are place of
articulation and voice) which are used to specific how a sound is
produced.
Morpheme: the minimal grammatical unit of meaning. It is an abstract form.
Nasal: A sound, vowel or consonant, which is produced by opening the nasal cavity
(through lowering of the velum).
Natural Class: A group of sounds which behave similarly.
Oral: Articulated in the mouth. The term usually implies that the nasal cavity is not
involved.
Palatal: A place of articulation of the hard palate in the centre of the roof of the mouth.
xii
Phoneme: In traditional phonology the smallest unit in language which a sound is
produced.
Phonetic representation: the pronunciation of a word in speech.
Phonetics: The study of human sounds.
Phonological representation: the underlying form of a word.
Phonology: The study of the sound system of a language or languages.
Phrase: A linguistic unit made up of a word or words. The unit has a word whose word
class determines the class of the phrase.
Place of Articulation: The point in the vocal tract at which a sound is produced.
Plosive: A sound which is produced with a complete blockage of the pulmonic airstream.
Productivity: A reference to the extent that a given process is not bound in its application
to a certain input.
Prosily: A term which refers to all the suprasegmental properties of language such as pitch,
loudness, tempo and rhythm.
Root: In grammar the unalterable core of a word to which all suffixes are added.
Segment: A unit of speech which is identifiable and separate from others. It contrasts with
the term suprasegmental which refers to those aspects of phonetic structure above
the level of individual sounds.
Speech: The production of sounds using the organs of speech; contrasts directly with
writing which is a secondary medium for communication via language.
Supra-segmental: A reference to phenomenon which does not belong to the sound
segments of language but which typically is spread over several
segments, e.g. intonation, stress, tempo etc.
xiii
Syllable: The most important structural unit in phonology. A syllable consists of a series of
sounds which are grouped around a nucleus of acoustic prominence (usually a
vowel).
Synchronic: A reference to one point of a time in language.
Underlying Representation: A representation of what is assumed by the linguist to be the
structure which lies behind or forms the initial stage in the
generation of a surface structure item.
Underlying Structure: an initial structure of a phrase (spoken or unspoken), from which
another sentence is derived.
Voiced: Spoken with simultaneous vibration of the vocal folds.
Voiceless: Spoken without the vocal folds vibrating, the folds can either be open (the
normal state) or closed with the compression of air between them and the supra-
glottal stop position producing sounds.
xiv
ABBREVIATIONS
1PL First person plural
2PL Second person plural
3SG Third person singular
Affr Affricate
Alv Alveolar
Am Amenfi
Ant Anterior
ATR Advanced tongue root
BA Bachelor of Art
Bil Bilabial
Bt Both dialects
C Consonant
Cons Consonantal
Cont Continuant
Cor Coronal
CV Consonant Vowel
CCV Consonant Consonant Vowel
CVCV Consonant Vowel Consonant Vowel
CVNV Consonant Vowel Nasal Vowel
Fs. Fiase
Fut Future
H High tone
xv
IPA International Phonetic Alphabets
L Low tone
Lab Labial
Lat Lateral
M. A. Manner of Articulation
MPhill Master of Philosophy
N Nasal
NEG Negation
NV Nasal Vowel
Obs Obstruent
OCP Obligatory Contour Principle
Pal Palatal
P. A. Place of Articulation
Perf Perfect
PhD Doctor of Philosophy
PL Plural
Prog Progressive
SG Singular
SIL WL Summer Institute of Linguistics African Area Word List
Son Sonorant
SVO Subject Verb Object
Syll Syllabic
TBU Tone Bearing Unit
xvi
UAC Universal Association Convention
V Vowel
WFC Well Formedness Condition
SYMBOLS
Low tone
High tone
Ó Syllable
// Phonological Brackets
[] Phonetic Brackets
() Brackets
# Word boundary
+ Morpheme boundary
→ Becomes
! Downstep
xvii
ABSTRACT
This thesis presents some phonological processes in Amenfi and Fiase dialects of Wassa.
Wassa is one of the unwritten and least studied languages in Ghana. The language belongs
to the Niger Congo language family. The study is cast within the theoretical framework of
autosegmental phonology. The approach applied in this dissertation is qualitative. Data for
the analyses are drawn from both primary and secondary sources. The phonological
processes discussed in this work include vowel harmony, labialization, palatalization,
elision, insertion, compensatory lengthening and consonant mutation. Both dialects of
Wassa operate a limited stem-controlled Advanced Tongue Root (ATR) vowel harmony
system. There is also evidence of raising of the low vowel /a/ to mid back vowel in the
same compound name in Amenfi. The phenomenon where plosives and affricates change
into a nasal is a similar process in Amenfi and Fiase. Also, the phenomenon of consonant
mutation where obstruents becomes voiced in the environment of nasals is very productive
in Wassa. Bilabial lenition where bilabial /b/ lenites to [m] is also a phonological process in
Amenfi dialect. Future affix deletion in the environment of first personal pronoun and
deletion of the future and progressive in the environment of the negative are all similar
phonological processes shared by both dialects. Finally, the past and progressive affixes are
lengthened in both dialects.
xviii
CHAPTER ONE
INTRODUCTION
1.1 Overview
The section gives; a background to the study; a brief history of the Wassa people
and their language; the statement of the problem; the objectives; the research questions and
the significance of the study. The limitation and delimitations to the study as well as the
organization of the study are also presented in this chapter.
1.2 Background of the Study
According to Dophyne (1988), changes and adjustment in sounds are characteristics
of the Akan language. However, not much work has been done on the description of
changes that those sounds go through in the various dialect of Wassa.
Afful (2006), compared, aspects, of the sounds and tonal system of the Wassa
Amenfi dialect to the Asante and Bono dialects of Akan. Her main aim was to find out
whether the Wassa language is more related to the Asante or the Bono in terms of sounds
and tonal structure. Her findings prove that Wassa shares some differences and similarities
with both dialects.
Afful‟s work gives a good general phonological overview of the Wassa Amenfi
which is very helpful to major study as this. But her work could not point out some of the
unique phonological processes in Wassa. According to Schane (1973), combination of
morphemes to form words differs from one language to the other. One cannot use the
phonological process of one language to represent the other. There is therefore the need to
1
investigate and verify these tentative since some premilary observation proves some
phonological processes in the dialects.
1.3 The Wassa People and their Language
Oral tradition indicates that Wassa takes it origin from the collapse of the Akwamu
Kingdom. According Bassaw (1994), the families of Wassa migrated from Takyiman in the
17th century but other setters from Asante, Ahanta, Nzema and Fante later settled in the
area.
Geographically, Wassa is spoken in the western region of the republic of Ghana.
According to GSS (2014), their population is about six hundred and eighteen thousand
from hundred and twenty one (618,421).There are five districts in the area. There are also
two subdialects of the language namely, the Amenfi and Fiase subdialects. The Wassa
Fiase traditional area cut across two districts namely, Wassa Fiase West and Wassa
Mpohor East District. The traditional capital is Benso and that is where we the stool of the
Omanhene. The Wassa Amenfi traditional area is also designated Wassa Amenfi Central,
Wassa Amenfi East and Wassa Amenfi west district. The traditional capital of the Amenfi
is Akropong.
There are three main types of religions in the area Traditional, Islamic and Christian
religion. Traditional religion is the oldest but Christianity and Islamic is widely practice in
the area. There are a number of economic activities that take place but Agriculture is the
main productive activity engaged in by the majority of the people. Surface mining also
known as „galamsey‟ is another economic activity which gives employment particularly the
youth.
2
Linguistically, Wassa is an umbrella term from two dialects: Amenfi and Fiase.
Speakers of the dialects understand each other. The wassa language forms a sub-group of
the Volta-Comoe. Boadi (2009) indicate that the languages that form the groups are
categorized into three: Western, Eastern and Central Comoe. Based on linguistics evidence
Wassa could be classified with group B of the central Comoe. Figure (1) below from Boadi
(2009) shows the classification of Volta-Comoe with few modifications.
Figure 1: Family Tree Diagram of the Volta-Comoe Language

Volta – Comoe
Western Central Eastern
Group A Group B
1. Ahanta, Nzema 1. Akwapem, Akwamu 1. Akyem, Asante, Guan: Kyerepong
2. Aowin, Sehwi 2. Fante Wassa Larteh
3. Chakosi 3. Agona 2. Asen, Kwahu, Denkyira, etc
(Source: Boadi, 2009): Adapted and modified
Phonologically, ten (10) different vowel phonemes have been observed: /i, ɪ, u, ʋ,
e, ɛ, o, ɔ, a, æ/. The chart in figure two shows the Wassa vowel chart.
3
Figure 2: Wassa Vowel chart

Central
Front l ose u ack lose
ɪ ʋ
Close-mid e o Close-mid
Open-mid ɛ ɔ Open-mid
Open Open
The following words in table 1 below show examples of words in which oral vowels occur.
Table 1 shows word examples of the oral vowels in Wassa.
Table 1: Oral vowel in word examples
Vowel Example Gloss

/i/ t l head
/I/ ɕ to smoke, to burn
/u/ t uproot
/ʋ/ t throw
/e/ d d noise
/ɛ/ palm
/o/ dl medicine
/ɔ/ w cold
/æ/ d name
/a/ bʋ animal
Apart from the ten oral vowels, there are five nasal vowels in Wassa. These vowels are / ɪ,
u, ʋ and a/. Table 2 shows word examples of nasal vowels.
4
Table 2: Nasal vowel in word examples
Vowel Example Gloss

f dirt
/ɪ/ ɛs tooth
u/ fu belly
/ʋ/ k war
a/ m a bed
There are twenty six (26) consonant phonemes /p, b, t, d, k, g, kw, ɕ, ɕɥ, ʨ, ʨɥ, ʥ, ʥɥ, j, f,
s, h, m, n, ɲ, ɲw, ŋw ŋ Ɩ r w/.
Table 3: Consonant chart
Labio- Pre-
Bilabial Alveolar Velar Glottal
dental Palatal
k g
Plosive P b t d
kw
ʨ ʥ
Affricate f s
ʨɥ ʥɥ
Fricative ɕ ɕɥ h
Nasal m n ɲ ɲw ŋ ŋw
Lateral Ɩ
Approximate w r j
5
Table 4: Consonant sound in word examples
Vowel Example Gloss

/p/ p break
/b/ bɪpɔ mountain
/t/ t tear
/d/ d no
/k/ kaɪ read
/g/ ŋg t dw knee
/kw/ ɛkwaɪ road
/f/ f home
/s/ sl dɛ fat
/r/ fʋfʋrɔ new
/Ɩ/ ɲ ɲ ʥ lɛ scorpion
/w/ w l mary
/j/ j lɛ yam
/ʨ/ tɕ tɕl tie
/ʥ/ ʥ t lion
/ʨɥ/ ɛnwɔtɕɥ eight
/ʥɥ/ eʥɥ louse
/ɕ/ ɕ to smoke
/ɕɥ/ ɕɥ to drain
/h/ h here
/m/ h ʋma thread
/n/ m ɪna hole
/ɲ/ ɲinaa all
/ɲw/ ɲwɔŋgʋɔ friend
ŋw aŋwa cooking oil
/ŋ/ ɛŋgʋnʋ nine
6
The language is tonal and use two basic tones of High and Low (dá „sleep‟ and d „never‟ .
Each syllable bears its own tone which is realized on vowels and syllabic consonants.
Wassa belongs basically to the language group of V, CV and C syllable typology.
The syllable nucleus may stand alone as syllable as vowel (V) in Wassa. Every vowel
constitutes a syllable. It occurrence is at all word position (initial, medial and final).
Examples of this shape of syllables are given in table 4 below.
Table 5: The V syllable type
Word Gloss
t listen
sʋ car
p lizard
ɛwʋɔ Honey
sɔ hoe
The structural representation of V is shown below in figure 4 below:
Figure 3: Structural representation of V syllable shape
V Syllable Shape
Syllable Tier
CV Tier v c v
Segmental Tier s ɔ
„ ‟ in „ sɔ‟ constitute the V syllable structure in Wassa.
The CV structure constitutes a consonant and a vowel. The occurrence of the
structure is also at word initial, medial and final position. Examples of words of these
syllable shape attest in Wassa are given in the tables below.
7
Table 6: The CV structure
Word Gloss
tʋ shoot
nʋ Cook (v)
jɛ do
s fʋɔ leader
hn employ
We should note that in the CV structure the syllable is placed on the vowel
Figure 4: CV syllable shape
Syllable Tier
CV Tier c v
Segment Tier t ʋ
This structure usually occurs when there is a deletion of a word (Boadi, 2009)
Consonants that are syllabic are liquids and nasals /l, r, n, m, ɱ ŋ . They also occur at word
initial and media position.
Table 7: The C structure

Word Gloss
ɛjƖi wife
ŋgƖ ma Am worm
ɱvƖ ma wind
ŋgɔ don‟t go
8
The phenomenon of consonant mutation rule in which obstruent become voiced is very
common in all the language speaking area. The lenition of r to Ɩ is also common to ot h
dialects. The second and third person plural is all Fiase speaking communities is [bɛ] but
[mʋ] and [wɔʋ] are used in some part of the Amenfi dialect respectively. It is an SVO
language.
Asante and Fante are widely spoken as a second language in all wassa speaking
communities. They are used as a medium of instruction in the school.
1.4 Statement of the Problem
So far the only linguistic work on Wassa I have come across is a B.A. long essay by
Afful (2006) on the phonology of Wassa in the Linguistics Department of the University of
Ghana. Her main aim was to find out which of the two dialects of Akan, Asante and Bono
is related to Wassa. Her findings indicated that Wassa shares some similarities and
differences with both dialects in terms of sound system and tonal pattern.
Phonological processes in Wassa may have unique features that could contribute
immensely to linguistic knowledge. However, the problem that exists is that these
interesting features of Wassa have not been scientifically studied. We are unable to
ascertain whether these tentative features actually exist in the Wassa language or not. It is
therefore certainly motivating to carry out phonological analyses in both dialects of Wassa.
1.5 Objectives of the Study
Having taken cognizance of the problem identified above, this study would seek to:
1. Expose the reader to the phonological processes that exists in Amenfi and Fiase
dialects of Wassa.
9
2. Find out the variations that exist in the two dialects.
3. Identify the unique phonological processes in Wassa.
1.6 Research Questions
In order to achieve the objectives set above, the following research questions are
posed to guide the study:
1. What are the phonological processes in the Amenfi and Fiase dialects of Wassa?
2. What variations exist in the two dialects?
3. Does Wassa exhibit any unique phonological processes?
1.7 Significance of the Study
There is the need to document and preserve the grammar of a language so that it
would be possible for everyone to read or engage in a discussion. Agyekum (2012) posits
that every language should be will kept and saved for proper use and serve as a heritage for
posterity. Ofori (2014) is also of the view that the documentation of a language goes a long
way to ensure the sustenance of the language for posterity and for linguistic argumentation
even if the language is no more spoken.
1. This study would be added to the body of literature on Wassa and Akan
linguistic study in general.
2. Aside providing additional literature, it will also pave the way for a more
thorough analysis of the phonology of Wassa.
1.8 Limitation
Getting the data from the two dialects constituted a great challenge considering the
time available. This gave me much work to do when it came to transcription and glossing.
10
As the data collection process involved a lot of movements, much time was also needed for
the study.
1.9 Delimitation of the Study
As a general rule, a research put boundaries to demarcate the extent of the study.
This study limits itself precisely to the confines of the phonological process and not
everything in phonology to be able to manage the data. Tone, for example, is not given
much attention in this thesis due to time. Morphology is also excluded though some of
these processes are related to morphology.
1.10 Organization of Thesis
The whole research is divided into four chapters. Chapter one serves as the
introductory chapter. Chapter Two discusses the literature review, theoretical framework
and phonological features used in the description of the data. We proceed to Chapter Three
to look at the methodology. Under Chapter Four, we discuss the phonological processes
which include: vowel harmony, nasalization, consonant mutation, labialization,
palatalization, elision, insertion and compensatory lengthening. Chapter Five presents the
summary of work done and
draws general conclusion.
11
CHAPTER TWO
LITERATURE REVIEW
2.1 Overview
This chapter looks at existing works of relevance to the area of study. First, it
discusses the theoretical framework of some of the concepts and the motivation in relation
to Auto-segmental Phonology. Though the mode of phonological representation is within
Goldsmith‟s 1976 Auto-segmental Phonology a review of homsky and Halle‟s linear
generative approach to phonological representation is important since this theory generated
from Generation Phonology.
2.2 Generative Phonology
The Generative phonology theory was a dependable theory for phonetic and
phonological analysis before the emergence of autosegmental phonology. The theory
espoused in Chomsky and Halle (1968) Sound Pattern of English (SPE). It characterized
segment as unstructed homogenous matrices “where every segment has a specification for
each of the two dozen distinctive features orderly ound to gether as unit” Odden 2005 .
Phonological representations of segments in generation phonology, therefore,
consist of a string of segments arranged in a linear form like the serial arrangement of the
orthography. For instance for the Wassa and [ŋgá] ‘scent’, each segment is broken down
into their distributive features and arranged linearly with each representation of the entire
word. Figure 6 below shows the linear generative phonology representation of the word
ŋgá.
12
Figure 5: Linear representation of segmental and suprasegmental features

ŋ g a
+ cons + cons + sylla
+ nasal + voiced - ATR
+ back + back + high
In this representation, the high tone is lumped up with other features and considered
as part and parcel of the segment /a/. Viewing segments in this way implies that in the
event of segment deletion, every distinctive feature goes with the segment in question.
However sometimes, tone and other suprasegmental phonemes which are neither
vowels nor consonants resist the effect of deletion of a segment. Consequently, generation
phonology could not succeed in explaining such facts about sounds. It indeed fails woefully
in accounting for suprasegmental features of sounds such as tones in phonological analysis
(Odden 2005), thus the emergence of new phonological theories.
2.3 Autosegmental Phonology
Autosegmental phonology was developed within the tradition of classical
Generative phonological theory of Chomsky and Halle (1968), following the works of
Williams (1971) and Leben (1973) on tone systems in West African language such as
Margi, Igbo and Mende. But the principle and remarkable innovation of Autosegmental
phonology are exemplified in Goldsmith (1976) in his dissertation to Massachusetts
Institute of Technology (MIT). Autosegmental phonology is not a departure from the
principle of Generative phonology codified in homsky and Halle‟s Sound Pattern of
13
English (SPE) in 1968. It is a non-linear version of phonological analysis of Generative
Phonology while SPE is a linear version of phonological analysis.
Indeed Goldsmith 1990 confesses that Autosegmental phonology is a “direct
confirmation of the traditional works in generative phonology codified in Chomsky and
Halle‟s Sound Pattern of English in 1968”. The only difference e tween the SPE and
Autosegmental theory is “the development of a multi-linear phonological analysis in which
the various tiers are organized y the „association lines‟ in Autosegmental phonology
Goldsmith 1979 ”. It can therefore best be viewed as an attempt to supply a more adequate
understanding of the phonetic side of linguistic representation (Goldsmith, 1979).
Goldsmith 1979 further explained that autosegmental phonology is a “theory of
how various components of articulatory apparatus, i.e. the tongue, the lips, the larynx and
the velum are coordinated” in the process of sound production. It therefore implies that
phonological features lead their own independent lives and not grouped together in ordered
bundles (segments) as it is made to believe in generative phonology.
In this vein therefore, Autosegmental phonology places segments and
suprasegmentals (especially tones and other prosodic features) on different tiers parallel to
each other with the suprasegmentals being linked to the segments by association lines.
Every segment on each tier is specified for a set of features specific and unique to that tier
and segment on each tier are associated with segment on another tier by association lines.
The Autosegmental phonological representation of the Wassa word ŋgá will therefore take
the form as in figure 6 below.
14
Figure 6: Autosegmental representation of segmental and suprasegmentals
features
Tone tier L H
Skeletal tier X X X
Segmental tier ŋ g a
In figure 6 above, the segments are arranged on multiple separate and independent
tiers and each relates to the other tier by the association lines. The low tone and the
segment [ŋ] relate to each other via the skeletal tier signifying the simultaneous
coordination of the segment and the suprasegmental at the point of articulation. The same
applies to the high tone and the segment [á].
2.4 Basic Tenets of Autosegmental Phonology
The application of Autosegmental phonology is based on some fundamental
principles and conditions. These principles and conditions though originally meant for tone
analysis apply to all other phonological representations. They include the following
principles and conditions: the Skeletal Tier, Linkage Condition, Universal Principles of
Association, Obligatory Contour Principle and Well-Formedness Condition.
2.4.1 Skeletal tier
The skeletal tier play a crucial role in the organization of the entire phonological
structure.it is the mediating point on which the elements on the different tiers anchor. A
segment which is not linked to a position in the skeletal tier is not phonologically realized.
The skeletal tier hold units represented by consonants and vowels, hence they are
sometimes represented as C-slots, V-slots or X-slots. The skeletal tier can relate with other
15
tiers on one-to-one, one-to-many or many-to-many basic. Figure 8 below illustrate the
position or relation of the skeletal tier to other tiers in autosegmental representation of the
word kùsù „cat‟ and ɛ „palm tree‟.
Figure 7: Skeletal tier
a. Tonal tier L
Skeletal tier x x x x
Segmental tier k u s u
b. Tonal tier L H
Skeletal tier x x x
Segmental tier a b ɛ
In 7(a) above, the Low tone on the tonal tier relate to /u/ on the segmental tier via
the V on the skeletal tier on one-too-many fashion. In 7(b), the Low tone on the tonal tier
relates to /a/ on the segmental tier on one-to-one fashion so as the High tone also relates to
/ɛ/ on the segmental tier via the V on the skeletal tier on one-to-one fashion.
2.4.2 Universal association convention (UAC)
These principles indicate the relationship between tones and Tone Bearing Units
(TBUs) and guide the assignment of a tone to TBUs as outlined by Durand (1990) below:
Mapping: Associate vowels with tones in a one-to-one fashion from left to right until we
run out of tones or vowels.
16
Dumping: If after applying (mapping) some tones are still free (that is unassociated), link
the to the last vowel of the right.
Spreading: if after applying (mapping) some vowels are still free, link them to the last tone
on the right.
Line Crossing: Lines are not allowed to cross each other. The following data on Margi (a
Chadic language in Northern Nigeria) adopted from Oyebade (1998) best illustrate these
principles in Figure 8.
Figure 8: Universal Association Convention.
Tone tier H L HL H L
Segmental tier sa dla bdlu sa-na dla-na
a. Mapping Tone tier H L HL H L
b. Dumping Tone tier H L HL H L
c. Spreading Tone tier H L HL H L
17
2.4.3 Linkage condition
Though segments are represented on autonomous tier, they are not phonologically
realized if they are not linked to other tiers. This condition thus demands each tonal feature
and tone bearing segment on the separate parallel tiers be linked to each other via the
skeletal tier by association lines before it can be phonologically realizes (Goldsmith, 1990).
Association has represented concurrent articulations so that when tone is linked to
the segment thus it means that both tone and segment are articulated simultaneously. This
condition actually caters for the phenomenon of floating tone and tone stability.
Figure 9: Linkage condition
Tone tier L H L H L H
Skeletal tier x x x x x x x x x x x x
Segmental tier b υ l ͻ b υ l ͻ b l ͻ
Tone tier L H L H
Skeletal tier x x x x x x x
Segmental tier b l ͻ b l ͻ
In figure 10 a ov e the segment υ is linked to the skeletal tier in the underlying
form but it is delinked during fast speech and for that matter it is not realized phonetically
in the pronunciation of the word. The υ therefore remains silent while its Low tone docks
onto the following sound /l/ hence the resultant is the syllabic consonant formation in the
language. The word therefore suffices as [b֙lͻˋ] rather than [bύlͻˋ] during past speech.
18
2.4.4 Obligatory contour principle (OCP)
This principle spells out categorically that concentration of two or more adjacent,
identical tones are fused into a single tone e fore they are “mapped onto” their
corresponding vowels. What this means is that, a word with more than one High tone
vowel such as ὲpύnύ „door‟ have to e mapped onto one High tone that is one-to-many)
instead of a one-to-one mapping of the two high tones. The principle also applied perfectly
to other segments on the other tiers. The representation in Figure 11a is in consonance with
the OCP whilst 11b is a violation of the principles as shown below.
Figure 10: Obligatory contour principles
a. Tone tier L H
Skeletal tier x x x x x
Segmental tier ɛ p υ n υ
b. Tone tier * L H H
Skeletal tier x x x x x
Segmental tier ɛ p υ n υ
From the illustration above in Fig 11a, the OCP permits the two High tones to be
merged into just a single Low tone (one-to-many) representation.
19
2.4.5 Well formedness condition (WFC)
Well formedness conditions are the universal principles which govern the multi-tier
structure of autosegmental representation. These conditions actually determine the way
tone and vowels are associated on the parallel tiers and allow the addition or deletion of
association lines at any point throughout a phonological derivation. These are stipulated in
Goldsmith (1979:207) as follows:
1. Each vowel must be associated with (at least) one toneme.
2. Each toneme must be associated with (at least) one vowel.
3. Association lines may not cross.
By the Well Formedness Condition, all the tiers remain independent throughout the
derivation with association lines linking them. At no point should the derivation with
association lines cross in the course of derivation.
The representation in figure 12a is a violation of the Well Formed Condition
because the High tone has crossed the Low tone to map onto the last tone bearing unit via
the skeletal tier, even though the two are not adjacent. The current configuration which
conforms to the Well Formedness Condition is 12b.
Figure 11: Well formedness condition
a. Tone tier * L L H
Skeletal tier x x x x x x
Segmental tier ͻ t a ɱ v o
20
b. Tone tier L H L H
Skeletal tier x x x x x x
Segmental tier ͻ t a ɱ v o
2.5 Motivation for Autosegmental Phonology
The motivation for the use of autosegmental framework in this thesis is based on its
ability and flexibility in representing complex segments such as Contour tone, Tone
stability, Floating tones among others and also in providing explanations for their
occurrences which were otherwise problematic to deal with in linear generative phonology.
Since Wassa is a tonal language and also has floating tones, autosegmental phonology is
the best option for the phonological representation in this analysis.
2.6 Tiers Used in this Thesis
Tiers are the multi-levels at which the different phonological features are arranged
in non-linear generative phonology. They usually consist of sequences of features which
differ from one tier from one tier to the other. There are various kinds of tiers that can be
specified for various phonological analyses. In this thesis the following tiers are applicable
to the phonological representation:
i. Skeletal / CV Tier
It is the mediating point on which the elements on the different tiers anchor.
The skeletal / CV tier holds unit representation by consonants and vowels; hence
they are sometimes represented as C-slots and V-slots or X-slots. The skeletal / CV
tier can relate with other tiers on one-to-one or one-to-many basis.
21
ii. Segmental Tier
This is the level at which phonetic representation relevant to the
description are arranged. In other words, the underlying phonetic segments occupy
this tier.
iii. Place / Manner of Articulation Tier (P.A / M.A Tier)
This tier specifies the Place or Manner of articulation features of the
segment that are significant to the phonological analysis. This implies in the
formalization of phonological process such as Homorganic Nasal Assimilation,
Labialization and Palatalization.
iv. Tonal Tier
The tonal tier specifies whether the underlying tone on the segment is high
(H) or low (L).
v. Tongue Root Tier
This tier spells out the tongue root position features of the phonological
segments whether it is an Advanced Tongue Root features [+ATR] or Un-advanced
Tongue Root features [-ATR]. This tier is very relevant in the discussion of vowel
harmony as a phonological process in Wassa.
vi. Syllable Tier

This tier indicates the units of sounds that make up a word. It is applicable
in the discussion of the syllable structure processes such as elision, insertion and
compensatory lengthening.
22
2.7 Distinctive Features
Distinctive features are a set of articulatory and acoustic features sufficient to define
and distinguish, one from the other, the great majority of the speech sounds used in the
languages of the world (Halle & Clements, 1983).
The search for which distinctive features are really significant linguistically and
how these features combine is a sub theory in self. This theory of phonological features is
concerned with the discovery and explanation of the generalization about the phonological
behavior of phonological segments both in isolation, in sequence or as a member of a
system.
Generative Phonology proposes twenty features which are loosely classified in
terms of major class features, place of articulation features, manner of articulation features,
tongue body features, lip features, acoustic features, state of the glottis features and other.
2.7.1 The major class features
2.7.1.1 Syllabic / Non-syllabic: [+Syll]; [-Syll]
Syllabic sounds constitute a syllabic peak. The vowels are [+Syll]. All other sounds
are non-syllabic [-Syll . Some contextual variants of nasals and laterals are [ Syll
especially where a vowel would normally occur. For example the velar nasal [ŋ in [ŋgʋkɔ]
„fowls‟ is sylla i c.
2.7.1.2 Consonantal / Non-consonantal: [+Cons]; [-Cons]
Consonantal segments are produced with a radical obstruction in the mid-saggital
region of the vocal tract (Chomsky & Halle, 1968). Segments implemented in this way in
23
the Wassa languages are the plosives, affricates, nasal consonants and liquids. The vowels
and glides are non-consonantal.
2.7.1.3 Sonorant: [+Son]; [-Son]
Obstruent or non-sonorants [-Son] are sounds produce with a cavity configuration
that makes spontaneous voicing impossible (Chomsky & Halle, 1968). Plosives, affricates
and fricative are [+Obs] or [-Son]. Non-obstruent or sonorants are sounds produced with a
vocal tract shape which makes spontaneous voicing possible. The vowels, liquids, nasals
and glides are sonorant or non-obstruent.
2.7.2 Cavity features
2.7.2.1 Anterior / Non-anterior: [+Ant; -Ant]
According to Chomsky and Halle (1968), a consonant is anterior if it is produced
with an obstruction located in front of the palate-alveolar region of the mouth. No vowel is
[+Ant] on this definition since none involves obstruction. The [+Ant] consonants in Wassa
are the labials and alveolars. The palatals and velars are [Ant].
2.7.2.2 Coronal / Non-coronal: [+Cor]; [-Cor]
We shall use the term Coronal to refer to Wassa underlying alveolars /t d s n/, their
reflexes as well as their palatals and palatalized consonant. The velars and labials are non-
coronal.
2.7.2.3 Back / Non-back: [+Back]; [-Back]
Back sounds are produce with the back of the tongue retracted. Non-back sounds
are produces with no such retraction of the tongue. In Wassa [+Back] segments include the
24
back vowels [u ʋ o ɔ], and velars [k g kw gw h hw w]. The front vowels are [-Back], and so
are the palatal and palatalized consonants, the alveolars and labials.
2.7.2.4 Rounded / Non-rounded: [+Round]; [-Round]
Rounded segments are produced with a protrusion of the lips. Non-rounded ones
are produced with lip-spread. In Wassa the back vowels are redundantly [+Round].
2.7.2.5 High / Non-high: [+High]; [-High]
High segments are those which are produced with body of the tongue raised
towards the roof of the mouth above the neutral position. In Wassa, they include the high
vowels, [i u ɪ ʋ], the glides, the velars, palatal and palatalized consonants. The non-high
segments are mid and low vowels [e ɛ o ɔ], [a] and [æ] as well as anterior consonants like
labials and alveolars.
2.7.2.6 Advanced and non-advanced tongue root: [+ATR]; [-ATR]
Advanced tongue-root vowels are produced by drawing the root of the tongue
forward and raising the body of the tongue with a resulting expansion of the pharyngeal
cavity. The [+ATR] vowels are [i u e o æ], Non-Advanced Tongue Root ([-ATR]) vowels
are produced by retracting the tongue root and narrowing the pharynx. The Wassa [-ATR]
vowels are [ɪ ʋ ɛ ɔ a].
2.7.3 Manner feature
2.7.3.1 Continuant / Non-continuant: [+Cont]; [-Cont]
Continuant refers to the vowels, glides, liquids and fricatives. Non-continuant
sounds are the oral and nasal stops (plosives, nasal and affricates). In producing non-
25
continuants, the air-flow through the mouth is effectively blocked. This is true of all nasal
consonants and oral stops, but not of laterals, fricatives, glides and vowels.
2.7.3.2 Lateral / Non-lateral: [_Lat]; [-Lat]
Lateral are produced with partial closure at some point in the mouth with the result
that the air-stream is allowed to escape on one or both side of the contact. In Wassa, the
only lateral sound, [l], is non-fricative and is alveolar, [+Cor, +Ant].
2.7.3.3 Nasal / Non-nasal: [+Nas]; [-Nas]
Nasal segments are produced with a lowered velum to allowed air to escape through
the nasal cavity. They include all the nasalized vowels, the nasalized glides as well as the
nasal consonants. The non-nasals are the oral vowels, oral liquids, glides and consonants
and produced with a velic closure or raised velum.
26
Table 8 (a): Presents the distinctive feature matrix for vowel sounds of Wassa
i ɪ u ʋ e ɛ o ɔ æ a ἶ ɪ u ʋ a
round - - + + - - + + - - - - + + -
back - - + + - - + + - - - - + + -
high + + + + - - - - - - + + + + -
low - - - - - - - - + + - - - - +
ATR + - + - + - + - + - + - + - -
Son + + + + + + + + + + + + + + +
Cont + + + + + + + + + + + + + + +
Syll + + + + + + + + + + + + + + +
nasal - - - - - - - - - - + + + + +
i ɪ u ʋ e ɛ o ɔ æ a ἶ ɪ u ʋ a
27
Table 8(b): Presents the distinctive feature matrix of consonant sounds in Wassa
p t k kw b d g ɕ f s ʨ ʥ h ʨɥ ʥɥ m n ɲ ŋw ɲw r Ɩ w j ɕɥ ŋ
Syll - - - - - - - - - - - - - - - - - - - - - - - - - -
Cons + + + + + + + + + + + + + + + + + + + + + + + + + +
Son - - - - - - - - - - - - - - - + + + + + + + + + - +
Ant + + - - + + - + + + + + - + + + + + + + + + + + + -
Cor - + - - - + - + - + - - - - - - + - - - + + - - - -
labial + + - - + + - + + + + + - + + + + + - - - - + - - -
- -
round - - - + - - - - - - - - - - - - - - - - + - - -
back - - + + - - + - - - - - + - - - - - + - - - - - - +
high - - + + - - + - - - - - - - - - - - + - - - - - - +
cont - - - - - - - + + + + + + + + - - - - - + + + + + -
lateral - - - - - - - - - - - - - - - - - - - - - - - + - -
nasal - - - - - - - - - - - - - - - + + + + + - - - - - +
del.rel. - - - - - - - - - - - - - - - - - - - - - - - - - -
28
2.8 Literature Review
The Wassa languages have not seen a lot of developmental or research work. The
only study we have come across so far has e en done y Afful 2006 . “A loser look at
the Phonology of Wassa” – A B.A long essay in the linguistics Department of the
University of Ghana. Afful (2006), in this work gives a good insight into the sound system
and some aspects of tones. Her main aim was to find out whether the Wassa language is
more related to the Asante or the Bono dialects of Akan. Her findings indicate that Wassa
shares some differences and similarities with both dialects.
The work identifies ten oral vowels, five nasal vowels and fourteen. She noted that
the Wassa sound system is not different from the other Akan dialects. It is the same as
Asante but has a slight difference with the Jaman subdialect of Bono and this difference is
the fact the /f, s/ have allophones, which are [v, z] respectively.
Afful further discusses types of tones, functions of tones and tonal patterns of
verbal forms. She identifies two basic tones (high and low). She identifies lexical and
grammatical as the two basic functions of tones in Wassa. Also in discussing tonal patterns
of verbal forms, she looks at progressive, past, perfect and future. She finds that the tonal
pattern of the progressive and past tense verbal forms in Wassa and Bono are similar.
Asante differs from Wassa in this area. Wassa and Bono have high tone on the progressive
marker but Asante has a low tone on it. The TBU of monosyllabic verb in Wassa and Bono
also carries a high tone whereas that of Asante carries a low tone.
Afful‟s work didn‟t do much work on the phonological processes which is very
unique to both dialects.
29
Bota (2002) discusses aspects of Bono phonology within the autosegmental
phonology. She looks at the sound system, some of it phonological processes and tonal
structure. She looks at homorganic nasal assimilation, voicing assimilation, consonant
mutation, vowel harmony, compensatory lengthening, deletion and some reduplication
forms are looked at as the main phonological processes.
According to her, a nasal consonant adopts the same place of articulation of the
following consonant. She also points out voicing assimilation where the voicing on the
nasal also spreads to the preceeding consonant. Bota (2002), also points out both labial
toungue root harmony in the Atebubu sub-dialect of Bono. She identifies a situation where
the perfect affix is lengthened after a pronoun. According to her study, front vowels are
deleted in compound words where they preceed the second word.
This study is very useful to the present one since one of the processes and theory is
similar.
Owusu (2002) investigated some phonological processes in Asante, Akuapem and
Fante within the framework of autosegmental phonology. He discusses some assimilatory
processes, syllable structure.
Under assimilatory processes, he looks at vowel harmony, nasalization and
palatalization. He describes the vowel harmony in Akan as regressive assimilation where
advanced vowel assimilates un-advanced once that precedes them. According to him, Akan
has a type of vowel harmony in which the ten vowels fall into two set of five, each
distinguish by the advancement verses reaction of the tongue root where the two set of
vowel do normally occur together in the same word. He also claims that vowel harmony
rule also apply between words in sentences and in compound. He finds that in addition to
30
the ATR vowel harmony in all the three dialects, Fante and Asante dialects also have an
additional type of vowel harmony called rounding harmony.
His study further probes five nasal vowels which occur independently nasalized
after non-nasal consonants. From the data he presented these vowels contrast with their oral
counterparts. The high vowel becomes nasalized when they occur before the nasal
consonant in all the dialects. In the Fante dialect, the non-high vowels become nasalized
when they occur before nasal consonants which do not, however, occur in the Akuapem
and Asante dialects. Owusu claims that glide in Akan, [w] and [j], can occur only with oral
vowels but due to this phenomenon in the Fante dialect, the nasalized vowel assimilates the
preceding glide into a nasal. Asante and Akuapem will, however, have glide in the position.
Also, in Asante and Akuapem dialects, when a voiced plosive or affricate is
preceded by a nasal that has the same place of articulation as the voiced plosive or
affricate, the oral consonant becomes nasal. This assimilation of a voiced plosive and
affricates into nasal does not, however, affect the vowels that follow the assimilated
consonant. He holds that homorganic nasal assimilation is a regular process in Akan.
His discussion on palatalization came across two types of palatalization in Akan,
one which is superimposition of high front tongue position on the initial non-palatal
consonant, and the other in which a non-palatal consonant becomes a complete palatal
consonant. The first type of palatalization (plural) occurs mainly in the Fante dialect, where
labial and alveolar consonants are palatalized when they occur before front vowels. Total
palatalization is when the non-palatal consonant that is affected by this assimilatory process
becomes a complete palatal consonant. The non-labialized back consonants in Akan
31
become completely palatalized when they occur before front vowels. These consonants
occur mainly before back vowels and [a].
In Akan, when a word that ends in a vowel is followed by another that begins with
the mid-vowels such as [e], [ɛ], [o] or [ɔ] (all dialects) or [i], [ɪ] (Fante), the initial mid-
vowel or [i], [ɪ] (Fante) of the second word is generally deleted. High vowels after nasals in
word final position in Akan also get deleted. When this happens, the final nasal becomes
syllabic and, therefore, carries the tone of the deleted high vowel. This vowel does not,
however, get deleted in the Asante dialect when the speaker is speaking slowly or
emphatically.
The work therefore provides information on some phonological processes and
theoretical framework in my discussion.
Ofori (2014) investigates a synchronic study of the grammar of Anum, one of the
Hill-Guan languages in Ghana in a PhD thesis in the Applied Linquistics Department in the
University of Education, Winneba. He combines the basic linguistic theory, autosegmental
and functional-typological theories to provide explanations for the structure of the
language.
The phonology of Anum identifies the language as having some resemblance with
some related Kwa languages in Ghana like Akan, Ga, Gonja and Larteh. The phonological
processes discussed include vowel harmony, vowel mutation, vowel deletion, consonant
deletion, compensatory vowel lengthening and labialization.
According to Ofori (2014), the language operates a limited stem-controlled
Advanced Tongue Root (ATR) vowel harmony system with limited infractions. In his
discussion on consonant deletion and vowel mutation which occurs mainly with the second
32
person singular pronoun subject wu/wʋ, he posits that the approximant /w/ is deleted and
the vowel mutates from the high back vowel to its variant mid vowel in completive
construction. Also, vowel deletion occurs with the first person singular subject pronoun
mi/mɪ where the vowel of the pronoun is deleted leaving the consonant which becomes
syllabic. Again he claims that vowel lengthening in Anum occurs when the vowel of an
aspect marker is deleted but its tone stabilizes and dumps on the vowel of the pronominal.
In his labialization process, a consonant is rounded when it is followed by a rounded
vowel. This type of labialization is mainly found in CV and some CVV stemss
The observations made by Ofori (2014) on Anum phonology have great
implications for the study of Wassa since they are all Kwa related languages.
Essien (2015) studies some word-formation processes that occur in Nzema. He
discusses derivation, compounding, reduplication and some phonological processes that
occurred during the word-formation processes.
The phonological processes discussed are vowel harmony, consonant alternation,
loss if final syllable, loss of vowels or nasal prefix, homorganic nasal assimilation and
vowel lengthening. From the data, Nzema operates on a stem controlled vowel harmony
system. Vowels and nasals occurring at morph initial position are deleted for the effortless
use of the organ of speech. Also, a nasal consonant adops the same place of articulation of
the following consonant and voicing on the nasal also spreads to the preceeding consonant.
From the data, it is observed that in Nzema, it is possible for two or more phonological
processes to occur in one morphological derivation.
The observations present quite interesting similarities and differences in respect to
the phonological processes in Wassa which needs attention and discussion.
33
Abakah (2004) describes elision in Fante using the autosegmental phonology as a
theoretical framework. His discussion focused on elision of vowels, consonants and
syllables in the three major subdialects of Fante namely. Iguae, Anee, and Boka. He also
compared them to the Akuapem and Asante and concluded that these subdialects share
some similarities and differences with some dialects of Akan. He demonstrated that vowel
elision involving the truncation of one of two contiguous vowels at word boundary is very
productive process in Fante. Where it has been difficult to determine which of the two
identical vowels in a sequence at a word or a morpheme boundary is deleted in the
combinative style, he resorted to tonology to resolve the difficulty. He holds that to
determine which one of the two identical vowels in a sequence is deleted at word or
morpheme boundary is a near impossibility. He therefore demonstrated that it is not an
uphill task to place linguistic fingers on which one of the two identical vowels in question
is the likely candidate for elision.
However, the elision of post sonorant word-final [High] V has been noted to
operate differently in the various subdialect of Fante. The subdialect differences as well as
the differences between the three major dialects of Akan have been duly noted.
As regards the elision of consonants, he observes that any consonant that is deleted
in any of the varieties of Fante has to occur intervocalically at the underlying level of
representation. He also explains the elision of syllables. According to him, syllable elision
takes place word-internally and in imperative clauses containing associative phrases.
O‟keefe 2003 looks at the vowel harmony of Asante Akuapem and Fante dialects
of Akan using the optimality theory. He argues that Akan has two varieties of harmony:
tongue root position and rounding harmony where both systems are well-attested and
34
highly consistent. He claims that all three dialects are clearly harmonic for the tongue root
position with infrequent and potentially explainable cases. Harmony for lip rounding does
not apply to stems but within affixes and is a persistent enough pattern that Asante and
Fante can be said to be harmonic for it as well. Akuapem, however, has no rounding
harmony.
O‟keefe looks at the pattern of the harmony and demonstrates that it is easily
understood when its different domains – stems, prefixes and suffixes are examined
independently. From the data he presented, there is certainly good cause to accept the
existence of tongue root harmony in stems. He examines two consistent violators of
harmony within stems. The first is the low vowel /a/ which occurs at the right edge of the
word and violates harmony with the first syllable. The other harmony violator is /ɛ/ which
only appears to occur after palatalized consonants. He however could not identify rounding
harmony in stems.
According to him, Akan stems are generally quite short, and words often have a
large number of affixes. For this reasons he consider some verbal and nominal affixes since
they will provide valuable information. Verbal prefixes include: pronominal, future,
progressive, perfect, ingressive and regressive. For the verbal suffixes, he considered the
past and nominalizing. He also looks at singular for the nominal prefix and Asante
nominal, person, diminutive plural and kingship plural for nominal suffixes. He noted a
clear harmonic behavior with the most of the domains with minor exception in all three
dialects. This paper is very helpful to my thesis since vowel harmony in these dialects is
related to Wassa.
35
Boadi (2009) does a comparative description of the phonology of the affixes that
co-occur with the finite in seven of the Central and Western dialects of Volta-Comoe using
generative phonology. He outlines their general phonological properties, the fact that they
are all vowel harmony and operate common processes like labialization and palatalization
under similar conditions.
oadi‟s description of the phenomenon of vowel harmony in all dialects shows that
of a cross-height type where only members of one set can co-occur in a root morpheme of
two or more syllables and across word boundary with minor but systematic exceptions.
According to him, Volta-Comoe had the following ten-vowel system [i, ɪ, e, ɛ, æ, a, ɔ, o, ʋ,
u].
He discusses labialization and to him consonants could be rounded before a back
rounded vowel. His discussion also shows that some consonants may be fronted if followed
by a high front vowel. He also discusses consonant mutation where obstruents become
voiced if followed by a nasal consonant.
A look at the literature review above indicates that no studies have been made on
the phonological processes in Wassa. The work of Afful (2006) concentrates mainly on
comparing aspects of the sounds and tonal system of the Wassa language to the Bono and
Asante. Other works reviewed all discuss aspects of the phonology of related Kwa
languages. A major gap is therefore left and this is what this work attempt fills.
36
CHAPTER THREE
METHODOLOGY
3.1 Overview
To conduct research on the phonological processes in Wassa and to ensure that the
results are generalized, scientific methodology based on verifiable data must be used. This
chapter entails the research design, selected site, population and sampling, data collection
and data analysis.
3.2 Research Design
The research adopts a qualitative approach. Under qualitative research there are
several approaches. This study will make use of the inductive approach to qualitative data
analysis. Thomas (2006) stated the characteristics of qualitative research more clearly.
According to him the inductive approach can e st e defined as “approaches that primarily
uses detailed readings of raw data to derive concepts, themes or a model through
interpretation made from the raw data y an evaluator or researcher”. He further explains
that the primary purpose of the inductive approach is to allow research findings to merge
from the frequent dominant or significant themes inherent in raw data, without the
restraints by structured methodologies. He identifies the following purpose of an inductive
approach.
a. To condense extensive and varied raw text data into brief summary format;
b. To establish clear links between the research objectives and the summary findings
derived from the raw data to ensure that these links are both transparent and
37
c. To develop a theme, model or theory about the underlying structure of experiences
or processes that is evident in the text data.
In the sense of the first purpose of inductive analysis, the language data represents
the extensive and varied raw data that is mentioned whilst the underlying language
structure that will be realized represents the condensed brief summary of the data. In the
context of this study, the research intends to seek the phonological processes that exist in
Wassa.
The second purpose of the inductive approach is to establish clear links between the
objectives and the summary of the findings gotten from the data. In this light the objectives
stated in the first chapter of this report are in direct correlation with the findings in the
chapter four (4). The findings represent the underlying patterns concerned with the
phonological processes exist in Wassa.
The third purpose of the induction approach is for the formulation of themes, model
and theories. In our context the current study seeks to contribute to the identification of the
grammar of Wassa.
3.3 Selected Site
Data for this study was collected through six trips to Wassa each lasting for three
days. This researcher visited two towns from each sub dialect. Data was elicited from
Damase and Anakom in the Amenfi Sub dialect. In the Fiase sub dialects, Nsuaem and
Benso were also selected. The reasons are that, these four villages have the indigenous
speakers, who speak the Wassa language. The trips also helped to ascertain the linguistic
differences that exist between the dialects and answer one of the research questions.
38
3.4 Population and Sampling
Purposive sampling was used for this research as the subjects were selected
according to the purpose of the study. Purposive sampling is a type of non-probability
sampling. In non-probability sampling, participants are not chosen randomly. They are
selected through special procedures. They were selected through special procedures
because they can speak the language, hence they can provide the requisite information
needed by the researcher.
Eight (8) native speakers were sampled to represent the whole population due to
time, how to time, how to manage the population and accuracy. Two (2) native speakers
from each dialect were chosen within the age range of 40-70. The people below forty (40)
were left out in order to limit the language influence from other languages. The people
above seventy (70) were also left out because people at that age might begin to lose their
proper articulation due to teeth loss or some physiological factors.
In this study, a language consultant from each sub-dialect was selected. All the two
language consultants are native speakers of the Wassa born of wassa parentage who have
stayed there for a long time. The first is Mr. Kofi Abakah, a forty year old graduate teacher
who, hold a first degree in Akan from the Kwame Nkrumah University of Science and
Technology and is resident in Nsuem. The second consultant is Mr. William Andoh, a
retired teacher of Wassa Dunkwa D/A Junior High School (JHS). He is sixty five years old
and is currently the society steward of the Calvary Methodist Church in Wassa Dunkwa
and resident in Wassa Dunkwa. The selection of the consultant helped in the transcribing
and glossing since they have some basic linguistics background. The choice was to ensure
credibility and reliability of the results since the data that was collected from both dialects
39
were cross-checked with consultants. These consultants also helped in the history of the
people and language since not much work has been done.
Further interactions were also made with two level 400 male students (native
speakers of Wassa Fiase dialect) of the University of Education, Winneba studying
English. These native speakers have some basic knowledge in linguistics and they helped
in some of the analysis.
3.5 Data Collection
The researcher took part in storytelling sessions. These were mostly short stories.
The storytelling was chosen so that the researcher will be able to pick the data from a
natural setting since it is a form of entertainment. Four (4) stories were recorded from
different story tellers. Two stories were transcribed from both dialects. Conversation were
also recorded and transcribed. Out of the four conversations recorded, two were transcribed
from both dialects. Also, elicitations were done using the summer institute of linguistics
West African Area Wordlist 1 (SILWL 1), through unstructured interviews with the native
speakers. The word list was also distributed to the individuals‟ consultants and other
assistants. The elicitations were recorded using an audio recorder.
3.6 Data Analysis
The data was analyzed with the inductive approach by focusing on rule
identification. The analysis was in line with the autosegmental phonology which gave the
researcher the chance to observe the data and summarize the observation into phonological
rules.
40
CHAPTER FOUR
RESULTS AND FINDINGS
4.1 Overview
Phonological processes refer to changes that take place in sounds when segment are
justaposed. According to Oyebade (1998) phonological processes are sound modifications
motivated by the need to maintain euphony in a language or to rectify violations of well-
formedness constraints in the production of an utterance. This section describes some
common phonological processes in the Wassa language. The phonological processes that
will be discusses in this dissertation are divided into assimilatory and syllable structure
processes.
4.2 Assimilation Process
A phonological process is called assimilation, if as a result of its application two or
more segments in form agree in their value for some phonological feature(s) or feature
class (es) (Bakovi, 2007:335). Katamba (1989: 80) further explains that, assimilation is the
modification of a sound in order to make it more similar to some other sound in its
neighbourhood, with the aim of making a smoother, more effortless, more economical
transition from one sound to another.
When a sound is modified to look more like the sound that precedes it, the
assimilation is in a progressive direction. On the other hand, when the sound is modified to
look more like the sound that follows it, the assimilation is in a regressive direction.
Assimilatory process will focus on vowel harmony, nasalization of consonants, consonant
mutation, labialization and palatalization.
41
4.2.1 Vowel Harmony
Wassa, like many Volta-Comoe languages apply the vowel harmony rule. It is one
of the most important rules to observe in the pronunciation of words. Okeefe (2003)
defines it as a set of systematic co-occurrence restrictions. The vowels are divided into two
main sub-groups with regard to the advance tongue root specification. This type of
harmony would have the following harmonizing sets:
SET I [+ ATR] i u e o æ
SET II [- ATR I υ ɛ ͻ a
4.2.1.1 Vowels sequence within root morphemes
The ATR rule in both dialects of Wassa requires that, only the members of one set
can co-occur in a root morpheme of two or more syllabus. This is in line with (Bota, 2002;
Owusu, 2002; Okeefe, 2003; Boadi, 2009; Ofori, 2014). The examples in (8) below
illustrate the harmony patterns in Wassa.
SET I [+ ATR]
[ѐmwú] „inside‟
[ѐkwúnwú] „hus a nd‟
[ѐkwƖwó] „sore‟
[ѐpwƖˊwó] „squirrel‟
[ dwƖˊwó] „medicine‟
[kw dwú] „ a nana‟
[ nἵ] „eye‟
[b ŋgwú] „a kind of food made with maize‟
[ zwúó] „water‟
42
[ētṹ] (fs) „ uttoc ks‟
The representation is in (13) below explains it:
Figure 12: [+ ATR] vowels within root morpheme
ATR Tier [+ ATR]
Skeletal Tier x x x
Segmental Tier è m ú
SET II [- ATR]
[ dɔkwá] „quarreling‟
[ŋgɔŋgɔ zã] „gossip‟
[mwɔwƖέ] „nails‟
[bɔƖàmmɔ] „ ott le‟
[έjƖÍ] „wife‟
[έtwƖá] (Am) „ utt ocks‟
[έkwƖá] „cala a sh‟
[ŋgátɪ] (Fs) „groundnut‟
43
[ásɪέ] „under‟
[ánwʋ] „mouth‟
Figure 13: [- ATR] vowels within root morpheme
ATR Tier [- ATR]
Skeletal Tier x x x
Segmental Tier έ m ʋ
As observed in the data and representation above [e] and [u] which are advanced
vowels occur in the same word whereas the un-advanced [ɛ] and [ʋ] also follow similar
rule. This means that there is a restriction on the distribution of these vowels in both
dialects of Wassa, which does not allow the vowels of Set I to occur in the same word with
the vowels of Set II.
4.2.1.2 Vowels sequence across morpheme boundaries
The occurrence of the vowels across morpheme boundaries is restricted in a similar
manner in both dialects. Reduced pronominal forms as well as noun and verb prefixes have
at least two alternates each, and their selection is condition by the feature specifications of
the vowels in the stem. This process also sometimes trigger vowel sequence where vowels
are sometime undefined. A very common context in which this occurs is when the
progressive aspect follows any subject pronoun in both dialects. The perfect vowel in the
Fiase dialect is also lengthened. The examples in (9) below attest to this.
44
9a. Amenfi
/mɪ/ + /bέ/ + /dí/ [médí]
„1SG‟ „Fut‟ „eat‟ „I will eat it‟
/wɔ/ + /rɪ/ + /bá/ [wɔ ɔbá]
„3PL‟ „Prog‟ „come‟ „They are coming‟
/jɛ/ + /rɪ/ + /tú/ [jѐѐtwú]
„1PL‟ „Prog‟ „fly‟ „We are flying‟
/mɪ/ + / w [m w
„1SG‟ „Perf‟ „die‟ „I have died‟
/ɔ/ + /bɛ/ + /tùm [ twùm
„3SG‟ „Fut‟ „ e a le ‟ „He will e a le to ‟
m t [ m tw ]
„1SG‟ „Perf‟ „uproot‟ „I have uprooted it‟
/ɔ r d [ d
„3SG‟ „Prog‟ „eat‟ „He is eating it‟
/wɔ/ + /bɛ f [wɔbɛf
„3PL‟ „Fut‟ „take‟ „They will take it‟
/wɔ/ + /bɛ h [w hù
„3PL‟ „Fut‟ „ low‟ „They will low it ‟
/jɛ/ + f [jɛfwʋ]
„1PL‟ „clim ‟ „We clim it ‟
/bɛ/ + w [ w
„3PL‟ „die‟ „They die‟
45
9b. Fiase
/mɪ/ + /bέ/ + /dí/ [médí]
„1SG‟ „Fut‟ „eat‟ „I will eat it‟
/bɛ/ + /rɪ/ + /bá/ [bɛɛbá]
„3PL‟ „Prog‟ „come‟ „They are coming‟
/jɛ/ + /rɪ/ + /t [ jѐ t wú]
„1PL‟ „Prog‟ „fly‟ „We are flying‟
/mɪ/ + / w [m w
„1SG‟ „Perf‟ „die‟ „I have died‟
/ɔ/ + /bɛ/ + /tùm [ twùm
„3SG‟ „Fut‟ „ e a le ‟ „He will e a le to ‟
m t [m tw ]
„1SG‟ „Perf‟ „uproot‟ „I have uprooted it‟
/ɔ/ r d [ d
„3SG‟ „Prog‟ „eat‟ „He is eating it‟
/bɛ/ + /bɛ f [ ɛbɛf
„3PL‟ „Fut‟ „take‟ „They will take it‟
/wɔ/ + /bɛ h [ h
„3PL‟ „Fut‟ „ low‟ „They will low it ‟
/jɛ/ + f [jɛfwʋ]
„1PL‟ „clim ‟ „We clim it ‟
/bɛ/ + w [ w
„3PL‟ „die‟ „They die‟

46
Figure (15) shows the autosegmental representation of vowel harmony across morpheme
boundaries in the Fiase dialect:
Figure 14: Vowel harmony across morpheme boundaries
Underlying Form Pronoun Vowel Deletion
Syllable Tier [- ATR] [+ ATR] [- ATR] [+ ATR]
Skeletal Tier x x x x x x x x x x
Segmental Tier m w m w
[+ ATR] Spreading and [- ATR] Delinking
Syllable Tier [- ATR] [+ ATR] [+ ATR]
Skeletal Tier x x x x x x x x
Segmental Tier m w m w u
As can be seen from (14) above, the underlying form of the vowel in the first
person singular is deleted. After this deletion the [+ATR] value of the vowel in the verb /u/
spread leftwards to delink the perfect affix /a/ to /æ/ causing it to value the quality of its
vowel in accord with the ATR value of the vowel in the stem. It is important to note that
47
affixes in both dialect have two different pronunciations depending on whether they occur
with a verb stem that has a [+ATR] or [-ATR].
4.2.1.3 Vowels sequence between words
Sequences of two stem morphemes forming a compound also follow the ATR rule
in the Amenfi and Fiase dialects. ATR harmony is not limited to only vowels in the stem
but operates in words and across word boundary (Abakah, 2005). Vowel(s) in the second
word transmits its ATR feature to those in the first word. This situation describe so far is
found in example (10) below:
10. s ɛjɛ [ z jɛ t
„hand‟ „stomach‟ „palm‟
s k [z ŋg Am
„hand‟ „left‟ „left hand‟
The representation in figure (15) below describes the phonological process above.
Figure 15: Vowel harmony across word boundaries
Underlying Form:
ATR Tier [-ATR] [+ATR]
Skeletal Tier x x x x x x x x x
Segmental Tier z b e k
48
[+ATR] Spreading:
Segmental Tier z b ŋ g
[-ATR] Feature Delinking:
Segmental Tier z b ŋ g
Output Form:
ATR Tier [+ATR]
Segmental Tier z æ b ŋ g
49
4.2.1.4 The low vowel and the raising rule in Amenfi dialect
The low vowel /a/ is a central vowel (Dolphyne, 1988; Marfo 2013). In the Wassa
Amenfi dialect however, /a/ may be raised to the next higher mid-back vowels [o] or [ɔ] in
some compound names. The alternation of this low vowel is accounted for by the feature
specification of the vowel in the second word. That is, if the second word has a [+ATR]
word. /a/ will be raised to [o], if it is [-ATR], it moves to the next higher [ɔ]. Examples in
(11) illustrate this point.
Amenfi
11. j : f [j : f
j : m r [jɔ:mƖ
/bʋ nù [ :n
k sṹ [ k sṹ
f nm [ f m
m t ùt [ m:t
j : ʋ [jɔ:bʋ
j kumi [j : kù
Figure 16 the autosegmental representation of the above data
Figure 16: Vowel harmony and raising in Amenfi
A. ATR Tier [-ATR] [+ATR] [-ATR] [+ATR] [+ATR]
Skeletal Tier x x x x x x x x x x x x x x x
Segmental Tie j a: k u m j a: k u m j æ: k u m
50
B. æ o
From the above, the ATR value of the vowels in the second word spread leftwards
and raises the vowel /a/ to [o]. Boadi (2009), however identifies the raising of /a/ to the
next higher [+ATR] vowel [e] in some Fante dialects of the Volta-Comoe in certain
environment of a verb containing a [+ATR].
4.2.1.5 Some infractions of the vowel harmony rule
The low vowel [a] in both dialects violates the ATR harmony rule. This is similar to
the claim made by (Dolphyne, 1988; Bot, 2002; Owusu, 2002; Okeefe, 3003; Afful, 2006;
Boadi 2009; Ofori, 2014). Examples in (12) are:
[d f „talisma‟ [s k „money‟
[ it n „deep hatred‟
Boadi (2009) argues with evidence from some of the Fante dialects that the two low
vowels were distinct in most of the dialects of the Volta-Comoe. He claims that in the
course of historical development, the opposition between them has disappeared from most
varieties in favor of [a].
Another violation could also be found in the direction of harmony flow of those
vowels. In nearly all cases, the vowel in a morpheme or a word on the immediate right of
the boundary transmits its [+ATR] feature specification to the vowel of the morpheme or
word on the immediate left of the boundary as seen in some of the example above.
However, if a vowel occurs in suffix and is not protected by a consonant (Boadi, 2009) it
assimilates to the root vowel. The direction of assimilation could be rightward
(progression) in some limited cases. These are usually seen in the past affirmative suffix
and perfect negative suffix in both dialects. The examples in (13) illustrate this point.
51
13. di [d
„eat‟ „past‟ „ate‟
w [wù
„die‟ „past‟ „died‟
di [n
„NEG‟ „eat‟ „past‟ „have not eaten‟
w [ŋwùù
„NEG‟ „die‟ „past‟ „have not died‟
4.2.2 Consonant nasalization
Nasalization is the perceived nasal resonance heard on speech sounds that are not
originally nasal. Consonant nasal assimilation is the process whereby an oral consonant
acquires nasality from a neighboring nasal consonant (Annan, 2009). Since a lowered
velum determines the production of a nasal segment, an adjacent oral sound found in the
environment of the nasal may adopt the feature of nasality. This is a common feature in the
Wassa language.
Phonetically, all the Wassa dialects have [m], [ɱ], [n], [ɲ and [ŋ . However oadi
(2009) recognizes two underlying nasal consonants: /m n/ where /m/ is [-Cor] and /n/
[+Cor] in which both are redundantly [-High, -Back]. This means that the labiodental [ɱ],
palatal [ɲ and velar [ŋ are conditioned variants: They adopt the place of articulation of
the initial consonant.
52
4.2.2.1 Nasalization of voiced plosives and affricates
In both dialects, when a voiced plosive or an affricate is preceded by a nasal that
has the same place of articulation as the voiced plosive or affricate, the oral consonant
becomes nasal. The following examples in (14) below attest to this:
14. /m-/ + /bɔ [mmɔ]
„NEG‟ „kick‟ „don‟t kick‟
/m-/ + /ɔ [mm
„PL‟ „female‟ „females‟
n- d [n
„NEG‟ „eat‟ „don‟t eat it‟
s ɛm s [ sɛmm s
„matter‟ „ask‟ „question‟
Figure 17 illustrates this phonological process in both dialects.
Figure 17: Nasalization of voiced plosives
Underlying Form: [+Nasal] Feature Spreading Rightwards:
M. A. Tier [+Nasal] [-Nasal] [+Nasal] [-Nasal]
Skeletal Tier x x x x x x
Segmental Tier m b ɔ m b ɔ
53
[+Nasal] Feature Delinking: Output Form:
M. A. Tier [+Nasal] [-Nasal] [+Nasal]
Segmental Tier m b ɔ m m ɔ
Since the negative nasal /m/ has the same place of articulation as the plosive /b/, the nasal
quality spreads to the following consonant for it to be nasalized.
4.2.2.2 Bilabial lenition
Bilabial /b/ lenities to [m] in the Amenfi dialect. Examples are illustrated in data
(15) below:
15. Fiase Amenfi
[ ɲdʑ [m ɲdʑ „cassava‟
[m [mm „shoe‟
[ t [m t „a piece of‟
[bʋtɕ [mʋtɕ „three stoned stove cooker‟
[bƖ ʋɔ] [mƖ ʋɔ „liver‟
[bɔɔdʋw ] [mɔɔdʋw „plantain tu e r‟
54
Below is the autosegmental representation:
Figure 18: Bialabial b/m lenition
P. A. Tier [+Bil] [+Alv] [+Vl] [+Bil] [+Alv] [+V1]
Segmental Tier ɲ ʥ m ɲ ʥ
4.2.3 Consonant Mutation
Boadi (2009) defined consonant mutation under the general umbrella of consonant
alternation where morph-initial obstruents become voiced after a nasal or vowel which is
part of a specifiable grammatical affix. In wassa, obstruents become voiced if preceded by
a nasal in every environment. It applies more consistently and in large portions of stems
morphemes and across word boundaries. This rule is similar to Bono and Nzema but does
not operate in many of the Akan languages. This can be illustrated by the following
examples:
16. s [ z „hands‟
k mf ʋɔ [ŋg ɱvʋɔ „a kind of food‟
kùr ùm [ŋglùm „okro‟
t [d „ e cause‟
mf r m [ɱvl m „wind‟
n- kr [ŋgl
„PL‟ „send‟ „spoken message‟
55
m- fidi [ɱv d
„PL‟ „machine‟ „machines‟
/n-/ + /kɔ [ŋgɔ]
„NEG‟ „go‟ „don‟t go‟
/n-/ + /kɛn/ [ɲʥɥɛn Fs
„NEG‟ „wait‟ „don‟t wait‟
n- k [ŋg
„NEG‟ „say‟ „don‟t say it‟
ɛ k n [ ɛŋgw n Fs
„palm‟ „soup‟ „palm-nut soup‟
The phonological process above can be represented in autosegmental representation as in
Figure 19 below.
Figure 19: Cosonant mutation
Underlying Form: [Dorsal] Feature Leftward Spreading:
P. A. Tier [Coronal] [Dorsal] [Coronal] [Dorsal]
Segmental Tier k k
[Coronal] Feature Delinking: Assimilation at Point of Articulation:

P. A. Tier [Coronal] [Dorsal] [Coronal] [Dorsal]
Segmental Tier k ŋ k
56
[+Voiced] Feature Rightward Spreading: [-Voiced] Feature Delinking:
M. A. Tier [+Voiced] [-Voiced] [+Voiced] [-Voices]
Segmental Tier ŋ k ŋ k
Output Form:
M. A. Tier [+Voiced]
Skeletal Tier x x x
Segmental Tier ŋ g
From the representation above, two processes occur at the same time. Both
progressive and regressive assimilation occur under this phenomenon in Wassa.In the
context where the underlying nasal has lost its vowel, the nasal /n/ assimilates to the
following velar consonant in the point of articulation. This follows a consonant alternation
rule by which velar stop /k/ changes to voiced [g].
4.2.4 Labialization of consonants
Labialization refers to the phonological case where a sound which is not normally
rounded is articulated with a degree of lip rounding. It is phonetically marked with a
superscripted „w‟ on the speech sound to e la ia lized. In Wassa a consonant is rounded
when it is followed by a rounded vowel. Such consonants are said to be labialized. The
following in (17) are some examples:
57
17. /dʋ/ [dwʋ „to ripe‟
/kɔ p ɔ/ [kwɔ w
ɔ „goiter‟
pù ɛ [ pwù ɛ „east‟
/kʋsɛ/ [kwʋsɛ „sorry‟
/bɔkɔɔ/ [bwɔkwɔɔ „calm‟
kùr [kwĺw „city center‟
d ʋm [ d wʋm „cloth‟
d ʋndʋ [ d wʋndwʋ „mosquito‟
k [ ŋ gw „left‟ Am.
Figure 20 shows an illustration of the consonant labialization assimilation process in the
auto segmental representation.
Figure 20: Consonant labialization
Underlying Form: [+Round] Leftward Spreading
M. A. Tier [-Round] [+Round] [-Round] [+Round]
Skeletal Tier x x x x
Segmental Tier d ʋ d ʋ
58
[-Round] Feature Delinking: Output Form:
M. A. Tier [-Round] [+Round] [+Round]
Segmental Tier d ʋ dw ʋ
The alveolar /d/ is unrounded in the underlying form but it is however rounded in
the environment of a following back vowel /ʋ/ as can be seen in the representation above.
However, there is evidence that shows that the back stop consonant is labialized
before non-rounded vowels in Wassa. This is in line with oadi 2009 who claims that we
interpret every la ializ ed consonant as a sequence of a simple consonant and a ack vowel
in the underlying phonology. For instance in slow speech the word [ŋgw is pronounced
[ŋgʋ with some prominence on the back vowel. Examples are in (18) below:
18. ŋkʋ [ŋgw „life‟
/kɔ s [ kw s „foolish‟
/kʋ [kw „polish‟
kʋ t / [ŋgw d „junction‟
/ɛkʋ n [ɛkw n „road‟ Fs
59
Figure 21: Autosegmental representation of consonant labialize before non-
rounded vowels
Underlying Form: Deletion of [ʋ]:
M. A. Tier [-Round] [+Round] [-Round] [-Round] [+Round] [-Round]
Segmental Tier k ʋ k ʋ
[+Round] Leftward Spreading [-Round] Feature Delinking
M. A. Tier [-Round] [+Round] [-Round] [-Round] [+Round] [-Round]
Segmental Tier k k
60
Output Form:
M. A. Tier [+Round] [-Round]
Skeletal Tier x x
Segmental Tier kw
Indeed in the examples above, the high back vowel /ʋ/ becomes weakened in the
development of such morpheme until it finally loses its resonance, leaving a reflex on the
adjacent consonant /k/ as labialization.
4.2.5 Palatalization
Palatalization is another regressive assimilatory process in Wassa. Katamba (1989)
describe it as a process where a velar consonant which is followed by front vowel, attains
some slight anticipatory fronting of the part of the tongue that makes contact with the roof
of the mouth. It is the simultaneous raising of the front part of the tongue towards the hard
palate during the production of a non-palatal sound.
Even though distinction is made between back and non-back consonants is applied
in both dialects. The back consonants are /k g h w/ and in the environment of the following
underlying non-back non-low back consonants are fronted. That is, they are pushed forward
to the palatal region. It should be noted that the stops among these consonants could be
further affricated. Examples are:
61
19. /kɛ/ [tɕ „present as gift‟
k [tɕ „catch‟
k n [ɲʥ n „salt‟
/kɛns [tɕɛnz „silver for cooking‟
/hɛ/ [ɕɛ] „wear‟
Below shows the autosegmental representation:
Figure 22: Palatalization of consonants
Underlying Form: [+Front] Feature Leftward Spreading:
M. A. Tier [-Front] [+Front] [-Front] [+Front]
Segmental Tier k k
[-Front] Feature Delinking: Affrication Accompanying Palatalization:
P. A. Tier [-Front] [+Front] [+Round]
Segmental Tier t t
M. A. Tier [+Affr]
62
Output Form:
P. A. Tier [-Front]
Skeletal Tier x x
Segmental Tier tɕ ɪ
M. A. Tier [+Affr]
4.3 Syllable Structure Process
Syllable structure processes are those processes that lead to the lost or addition of a
sound in the word for morphophonological reasons. For instance, morphological process of
compounding, and also some times in fast speech, sometimes results in segments lost.
Similarly, the adaption of loan words into a language also results in segments insertations.
These processes affect the basic syllable structure of words in a language by altering
syllable shape of words. The syllable structure process that will be discussed here will
include elision, insertion, and compensatory lengthening.
4.3.1 Elision
Elision is a phonological process through which a sound is lost. Abakah (2004)
defines it a sed on how it operates in Fante as “a phonological process y which a vowel a
consonant and sometimes a syllable, which is an intrinsic property of morpheme in the
63
isolative style is dropped in the com ination style.” In Wassa, the vowel, consonant or the
whole syllable may be elided for economic use of speech.
4.3.1.1 Vowel elision
4.3.1.1.1 Deletion of high vowel preceded by a lateral
High vowels are deleted if they are preceded by [1] in the surface realization in both
dialects. In fast speech one will conclude that, this language involves a consonant cluster
following Dolphyne (1988), who made an attempt to discuss the elision of high vowels
before [r] with the claim that Akan does not have a basic CCV structure, but only at the
surface realization of it. Marfoh (2013) also argues that it should be analysis into CVCV
because of application of economy-motivated process of vowel elision that results in
syllable reduction. In a bid to ascertain that the elision of high vowel before [1] in Wassa
ensues from a basic CVCV structure, there is the need to explore data from both dialects to
give some piece of evidence. The data in (20) below illustrates this in both dialects.
20. r [ĺ „come‟
kr [kĺ „ask permission to leave‟
kùr ùm [ŋglùm „okro‟
/bʋrɔ [lɔ „ eatings‟
/pɪr [pl „sweep‟
64
The representation in 23 represents this phenomenon:
Figure 23: Deletion of high vowels
Underlying Form: Deletion of Low Vowel
Syllable Tier
R R R R
Skeletal Tier O N C O N C O N C O N C
Segmental Tier p I Ɩ a p Ɩ a
It is evidence in the above representation that the deletion of the High vowel does
not affect the tone. The tone spreads to re-syllabify the [r] which also lenite to [l].Marfo
(2013), identifies [l], [w], [y] in Anyi and Nzema which occur in the same environment but
the vowel before it is never elided for the succeeding syllable has to be only a labio-dental
approximant, [r] for the realization of CCV. According to him, [l] is not a traditional speech
sound in Akan but [l] and [r] are free variants due to the strong influence of English on
Akan. We take the position here that all high vowels are elided between two consonants
where the sound consonant is [l] in the surface form.
Dolphyne (1988) also identified that [r] carries the tone of the deletion vowel preceding it
in Akuapem and Asante while Fante it carries the tone of the deleted vowel following it.
4.3.1.1.2 Negative vowel deletion
The negative affix is a nasal segment homorganic with the following initial
consonant. It leaves its voicing reflex on the o st ruent that follows it in all environments in
almost all Wassa speaking towns. According to oadi 2009 the nasal segment is a reduce
65
form of its full NV form as appears in Sehwi and Aowin. The negative is [n or [nɛ] in the
context of command and perfect. For example: [nɛ! „don‟t come‟ or [n di „don‟t eat
it‟. The alternation of the vowel is accounted for y the ATR Vowel Harmony Rule. The
autosegmental representation illustrates this point:
Figure 24: Negative vowel deletion
Underlying Form: Syllabic Vowel Deletion: Negative Consonant Syllabicity: Output Form:
Syllable Tier:
Skeletal Tier x x x x x x x
Segmental Tier n ɛ n ɛ n
As can be seen from the representation above, the vowel [ɛ] is deleted through the
historical development of a language (Boadi, 2009). The nasal therefore absorbs the
syllabicity by an assimilatory process. We should note that in the present of a vowel, this
negative nasal segment is non-syllabic because there is already a [+syllabic] segment in the
syllable.
4.3.1.1.3 Deletion of high vowels after nasals in word final position
High vowels after nasals in word final positions in Wassa also get deleted. When
this happen, the finial nasal become syllabic and, therefore, carries the tone of the deleted
high vowel. The following examples illustrate this:
21. /tɪa m [tɪ
„shout‟ „inside‟ „shout‟
66
/ɔm n n t m [m d
„country‟ „PL‟ „twin‟ „inside‟ „region or state‟
4.3.1.1.4 Deletion of final mid vowels in fiase
Final mid vowels are elided in words in the Fiase dialect. Below we attempt to
account for this deletion process in the Fiase dialects.
22. Fiase Amenfi
[ŋg t [ŋg t ɛ „groundnut‟
[ tṹ [ɛtʋɔ „ utt ocks‟
[ d z [ d z ɛ „witness‟
The representation below proves this point:
Figure 25: Deletion of final mid vowels
Final mid vowel deletion:
Syllable Tier
Segmental Tier ŋ g t ɪ ɛ
67
Output Form:
Syllable Tier
Skeletal Tier x x x x x
Segmental Tier ŋ g t ɪ
4.3.1.1.5 Deletion of intervocalic nasal in Amenfi
In the Amenfi dialect if a nasal realized as /n/ occur intervocalically as in CVNV,
the nasal spread to the following vowel and is deleted. This nasalized vowel also
assimilates to the preceding vowel of the deleted nasal. The examples below illustrate this
point:
22. Fiase Amenfi
[ɔm n [ɔmaɪ „country‟
[k n [kaɪ „read‟
[tɔn [tɔʋ] „sell‟
[ dini [ d „name‟
[m ŋg ni [m ŋg „cocoyam‟
[k n [kaɪ „count‟
[dini [d „difficult‟
[nd n [d „veins‟
[ɛkw n [ɛkw ɪ „road‟
68
The autosegmental representation illustrates this rule
Figure 26: Deletion of intervocalic nasal
[+Nasal] Feature Rightward Spreading: Deletion of Intervocalic Nasal:
Syllable Tier
Segmental Tier d n d n
M. A. Tier +Nas -Nas
[+Nasal] Feature Leftward Spreading: Output Form:
Syllable Tier
Segmental Tier d d
M. A. Tier +Nas -Nas
69
In the above, the consonant /t/ alternates to /d/ due to present of the voiced alveolar
nasal. The intervocalic nasal /n/ also spreads to the final high vowel and deletes. This
nasalized vowel also assimilates to the vowel preceded by the deleted nasal.
4.3.1.2 Consonant elision
4.3.1.2.1 Future consonant affix deletion
The future affix in the environment of the first person pronoun is deleted in both
dialects.
Examples of this phenomenon are represented in (23) below:
23. m ɛ/ + /kɔ/ [mɛkɔ]
„lSG‟ „Fut‟ „go‟ „I will go‟
m + /bɛ/ + /fɪrɛ/ [mɛflɛ]
„lSG‟ „Fut‟ „call‟ „I will call‟
The representation below represents this point.
Figure 27: Future consonant affix deletion
a. Deletion
Syllable Tier
R R R R R R
Skeletal Tier O N C O N C O N C O N C O N C O N C
Segmental Tier m ɪ b ɛ k ɔ m _ _ ɛ k ɔ
b. Re-syllabification
70
Syllable Tier
R R R R R
Skeletal Tier O N C O N CON C O N C O N C
Segmental Tier m ɛ k ɔ m ɛ k ɔ
From the above representation, there are two process involves. The pronounce
subject vowel [I] and the future consonant affix [b] is first deleted. The second process is
re-syllabification where the pronoun subject [m] draws the vowel of the future affix [ɛ] to
itself to produce [mɛkɔ].
4.3.1.2.2 The future and progressive affix deletion
The future and progressive affix deletes in the environment of a negative. They are
identical in both dialects. Examples are seen below:
24. m ɛ/ + /n/ + /di/ [minni]
„lSG‟ „Fut‟ „NEG‟ „eat‟ „I will not eat it‟
m r n di [m nni
„lSG‟ „Prog‟ „NEG‟ „eat‟ „I will not eat it‟
/m-/ + /bɛ [mm
„NEG‟ „Fut‟ „come‟ „don‟t come‟
4.3.3 Insertion
Languages tend to simplify their syllable structure for ease of discourse or speech.
Sapaty (2005), claims that insertion occurs when an extra consonant or a vowel find its way
71
into the stem of a word when it originally have not been there. In Wassa when words are
loaned from other languages, there is a tendency to resyllabify these words to suit it
structure. The result is insertion and this marks such words which allow it to conform to the
syllable structure of wassa. According to Owusu (2002), Akan does not have a constant
cluster or word-final non-sonorant consonant, so consonant cluster in borrowed words are
broken by insertion by usually, a high vowel. This phonological phenomenon is applicable
in both dialects of Wassa. The data below support the claim:
25. /bʋk [ ùkù „ ook ‟
/nɛɛs/ [nɛɛs „nurse‟
sku:l [sùk ù „school‟
/mɪlk [mlɪkɪ „milk‟
Figure 28: Insertion of High Vowels
Underlying Form: Output Form:
Syllable Tier
Segmental Tier m ɪ l k m ɪ l ɪ k ɪ
From the above, the consonant cluster is broken by inserting /ɪ/ between /l/ and /k/.
The final vowel /ɪ/ is also inserted since /k/ does not end in a word in this language. It
should be noted that most vowels involved in insertion are [+High] vowels and they follow
the ATR rule.
72
4.3.4 Compensatory lengthening
Compensatory lengthening is an attested process in many languages including the
Wassa languages. According to Hayes (1989), it is the lengthening of a segment triggered
by the deletion or shortening of a nearby segment. When an underlying segment within a
syllable is deleted, an adjacent syllable is lengthened to fill the gap created as a result of the
deletion. Compensatory lengthening in Wassa is found with the realization of the past,
progressive and perfect affixes.
4.3.4.1 The progressive affix lengthening
The progressive affix in it underlying form has a CV. According to Boadi (2009),
the affix is represented in the central dialect of the Volta-Comoe language as [rɪ] derived
from /dɪ/ by consonant lenition.
In both dialects, the derivation of the phonetic forms of the progressive underlying
/rɪ/ is completely deleted to prepare the way for the realization of the various allomorphs.
The progressive affix consonant is first deleted follow by the vowel. This follows the
incorporation of another segment to fill the gap of the lost progressive affix. The choice of
a particular allomorph of the progressive depends on the phonological segment
immediately preceding it and the ATR value of the vowel(s) occurring in the verb on the
right of the affix. In addition to the loss of the progressive affix and segment incorporation,
the Fiase dialect incorporate segment pulls the high tone of the deleted progressive vowel
onto itself.
73
The data below provides evidence in both dialects:
26. Amenfi
/ɔ/ + /rɪ d [ d
„3SG‟ „Prog‟ „eat‟ „he is eating‟
/mɪ/ + /rɪ/ + /kɔ/ [mɪɪkɔ]
„1SG‟ „Prog‟ „go‟ „I am going‟
/ɔ/ + /rɪ [ɔɔ
„3SG‟ „Prog‟ „come‟ „he is coming‟
Fiase
/ɔ/ + /rɪ d [ d
„3SG‟ „Prog‟ „eat‟ „he is eating‟
/mɪ/ + /rɪ/ + /kɔ/ [mɪɪ!kɔ]
„1SG‟ „Prog‟ „go‟ „I am going‟
/ɔ/ + /rɪ [ɔɔ
„3SG‟ „Prog‟ „come‟ „he is coming‟
74
The autosegmental phonology below accounts for the Wassa data
Figure 29: Progressive affix lengthening
Underlying Form: Progressive Affix Consonant Deletion:
Syllable Tier
Segmental Tier ɔ r ɔ r
Progressive Affix Vowel Deletion: Segmental Incorporation
Syllable Tier
Segmental Tier ɔ ɔ b
75
High Tone Dumping: Output Form:
Syllable Tier
Skeletal Tier x x x x x x x
Segmental Tier ɔ: ɔɔ: b
Tone Tier [-High] [+High]
From the representation above, the progressive affix consonant /r/ is first deleted
followed by the vowel /ɪ/. The third person plural vowel /ɔ/ incorporates another segment to
fill the gap of the loss progressive affix.
4.3.4.2 Past affix lengthening
The past suffix in Wassa is realized as [ɪ]. In both dialects a segment becomes long if
it is verb-stem final and is followed by the past suffix and another word. The past suffix
vowel is elided and the elision process is compensated for by a lengthening of the stem-
final vowel or consonant. Examples below illustrate the point:
26. /ɔ h ɪ/ # /nʋ [ hùù
„3SG‟ „see‟ „Past‟ „3SG‟ „he saw him‟
/ɔ k ɪ/ # /nʋ [ kùmm nʋ]
„3SG‟ „kill‟ „past‟ „3SG‟ „he killed him‟
76
Vowel Lengthening: Output:
Syllable Tier
Segmental Tier h h :
4.3.4.3 Perfect affix lengthening in fiase
Another difference between Fiase, on the one hand, and Amenfi on the other hand,
is that whereas in the former the perfect in lengthened after the pronoun or noun vowel is
deleted, there is no such lengthening in the latter. Examples below illustrate this point:
Amenfi
27. m [ m
„lSG‟ „Perf‟ „come‟ „I have come‟
/wɔ t [w t
„3PL‟ „Perf‟ „uproot‟ „They have uprooted‟
Fiase
/mɪ [m
„lSG‟ „Perf‟ „come‟ „I have come‟
/wɔ t [w t
„3PL‟ „Perf‟ „uproot‟ „They have uprooted‟
77
The autosegmental phonology below represents this phenomenon.
Figure 30: Perfect affix lengthening0.
Underlying Form: Pronoun Vowel Deletion:
Syllable Tier
Segmental Tier b ɛ t b ɛ t
Perfect Vowel Lengthening: [+ATR] Spreading:

Syllable Tier
Segmental Tier b : t : t
Tongue Root Tier [-ATR] [+ATR] [-ATR] [+ATR]
78
[-ATR] Delinking: [+High] Tone Dumping:
Syllable Tier
Segmental Tier b t b æ: t
Tongue Root Tier [-High] [+High]
Output Form:
Syllable Tier
Skeletal Tier x x x x x
Segmental Tier b æ: t
In both dialects, the vowel deletion rule applies to delete the morph-final vowel
preceding the perfect prefix vowel. In the Fiase dialect, the prefix becomes long by
incorporating another vowel unto itself. Tone formation rule copies the high tone of the
verb on the right of the low of the perfect prefix to produce a compound gliding tone on the
79
prefix. Finally, the automatic downstep rule applies to drop the tone on the verb from high
to downstep.
80
CHAPTER FIVE
SUMMARY OF FINDINGS, CONCLUSIONS AND RECOMMENDATONS
5.1 Overview
The chapter summarizes the finding of the study; outline the conclusion drawn from
the findings; states their contribution to knowledge; makes recommendations and suggests
area for further study.
5.2 Summery of Findings
This thesis describes the phonological processes in the Amenfi and Fiase dialects of
Wassa, The phonological processes discussed in this work include vowel harmony,
nasalization, consonant mutation, labialization, elision, insertion, and compensatory
lengthening.
1. The tongue root harmony is very consistent in both dialects. There is also evidence
of raising of the low vowel /a/ to mid back vowel in the same compound name in
Amenfi. We should remember that this raising rule depends on the ATR feature
specification of the vowel in the second name.
2. The proces where plosives and affricates change to a nasal is one of the processes
observed in both Amenfi and Fiase dialects of Wassa. Also, the phenomenon of
consonant mutation where obstruents becomes voiced in the environment of nasals
is very productive in Wassa. However, bilabial lenition where bilabial /b/ lenites to
[m] is another difference within the Amenfi dialect and Fiase dialect.
3. A phonological process where a consonant is labialized is also similar to both
dialects.
81
4. Palatalization where consonants are pushed forward to the palatal region is another
phonological process shared by both dialects.
5. Elision of high vowels before a lateral, future consonant affix deletion in the
environment of first person pronoun, deletion of future and progressive affix in the
environment of negative and deletion of a high vowel after a nasal are seen as
similar processes in both dialects. The deletion of a final mid vowel in Fiase and
deletion of intervocalic nasal in Amenfi are the areas of differences.
6. Also, compensatory lengthening in the past and progressive is also similar in both
dialects. However, lengthening of the perfect vowel in Fiase is different from
Amenfi.
7. The insertion of high vowel in loan words to conform to the syllable structure of
Wassa is similar in both dialects.
5.3 Conclusions Drawn from the Study
Based on the findings of the research we can conclude that;
8. Consonant mutation is a common phonological process to both dialects.
5.4 Contribution to Knowledge
The study contributes to knowledge in the following ways;
9. It is known in Akan linguistics that the low vowel is raised to the next front vowel
(Dolpyne, 1988; Boadi, 2009) in Fante. But data in Amenfi dialect shows that the
low vowel /a/ could be raised to the mid back vowel.
10. Bilabial lenition where /b/ lenites to [m] in Amenfi has not received attention in
Akan.
82
11. Also, lenition of /r/ to [I] is also very productive in both dialects of Wassa.
5.5 Suggestions for Further Studies
One study cannot exhaust everything that is worthy of being known. Based on this,
the following areas for further studies are suggested. These areas include;
12. Further studies into the tonal structure of both dialects. The current study captured
phonological processes at the segmental level.
13. A study into the morphology of Wassa. Such a study can take into consideration the
claims made concerning the relation of phonology and morphology.
83
REFERENCES
Abakah E. N. (2004). Elision in Fante. Department of oriental and African Languages,

1 (4), 181-213.
Abakah E. N. (2005). Phonological analysis of word-final consonant in Akan.

Department of oriental and African Languages, 1 (5), 47-65.
Afful, M. A. P. (2006). A closer look at the phonology of Wassa: is it more related to

the Asante or the Bono dialects of Akan? Bachelor of Arts long essay submitted
to the Department of Linguistics, University of Ghana.
Agyekum, K. (2012). Documentation and Preservation of the Akan Language. Basic

Research Journal of Education Research and Review, 1(2), 23-37.
Annan, N. (2009). Aspects of phonological processes in Ga. MPhil Thesis submitted to

the Department of Applied Linguistics, University of Education.
Bassaw, A. (1994). Traditional organization of Wassa Mpohor people. MPhil

Dissertation submitted to the Institute of African studies, University of Ghana.
Boadi, L. A. (2009). A comparative phonological study of some verbal affixes in seven

Volta-Comeo languages of Ghana. Cantoment-Accra: Black Mas Limited.
Bakovi , E. (2007). Local Assimilation and Constraint Interaction. In Paul de Lacy (ed.)
The Cambridge Handbook in Phonology. Cambridge:C.U.P.
Bota, G. (2002). Some aspect of Bono phonology. MPhil Dissertation submitted to the
Department of Linguistics, University of Ghana.
Chomsky, N. and Halle, M. (1968). Sound pattern of English. New York: Harper &
Row.
Dolphyne, F. A. (1988). The Akan (Twi-Fante) language: Its sound system and tonal
structure. Accra: Ghana Universities Press.
Durand, J. (1990). Generative and non-linear phonology. New York: Longman.
Essien, P. (2015). Some word formation process in Nzema. MPhil Thesis submitted to
the Department of Applied Linguistics, University of Education, Winneba.
Firth, J. R. (1948).Sound and prosodies. Cambridge University Press.
Ghana Statistical Service (2014). 2010 population and housing census: district analytic
report, Amenfi District.
84
Goldsmith, J. A. (1976). Autosegmental phonology. Ph.D. Dissertation, MIT.
Goldsmith, J. A. (1990). Autosegmental & metrical phonology. Basil Blackwell.
Halle, M. & Clements, G. N. (1983).Problem Book in Phonology. Cambridge, Mass:

MIT Press.
Hayes, Bruce (1989). Compensatory Lengthening in Moraic Phonology. Linguistic

Inquiry, Vol.20. No.2, pp 253-306. MIT Press.
Katamba, F. (1989). An Introduction to Phonology. Essex: Longman.
Leben, W. R. (1973). Suprasegmental phonology. Ph.D. Dissertation MIT.
Marfo . O. 2013 . Optimizing Structure: The ase of the „ rV Sylla le of Akan‟

Studies in Ghana Journal of Linguistics 2 (1): 63-78.
Odden, D. (2005). Introducing phonology. Cambridge: CUP.
Ofori, K. (2014). The grammer of Anum. PhD Thesis submitted to the Department of
Applied Linguistics, University of Education, Winneba.
O‟Keefe M. 2003 . Akan vowel harmony. BA Thesis submitted to the Department of

Linguistics, Swarthmore College.
Owu-Ewie, (2012). Introduction to traditional and action research. Winneba.
Owusu, S. (2002). Aspects of phonological processes in Akan: an autosegmental

analysis. MPhil Dissertation submitted to the Department of Linguistics,
University of Ghana.
Oyebade, F. (1998). A course in phonology. Nigeria: Shebiotimo publications.
Thomas, D. R. (2006).A general inductive approach for analyzing qualitative evaluation

data, American Journal of Evaluation, 27 (2), 237-246. American Evaluation
Association.
Sapaty, G. A. (2005). The study of the phonological structure of the Gbi dialect of Ewe.
MPhil Thesis submitted to the Department of Applied Linguistics. University of
Education, Winneba.
Schane, S. A. (1973). Generative phonology.New Jersey: Prentice Hall Inc., Englewood

Cliffs.
Williams, E. (1971). Underlying Tone in Margi and Igbo. In: Linguistic Inquiry 7(3).
463-484 (1976.)
85
APPENDIX
1.S/N LIST ID LIST GLOSS (ENGLISH) AMENFI WORD FIASE WORD

1 10 eye n n
2 20 ear sũ sũ
3 30 nose ὲɕɥ n ὲɕɥ n
4 40 mouth ànũ ànũ
5 43 tooth ὲs ὲs
6 45 tongue tὲtɕὶlὲma tὲtɕὶlὲma
7 48 lips nũfafa nũfafa
8 60 chin àbɔʥɥ ɛ àbɔʥɥ
9 70 jaw pa da pa da n
10 80 beard àbɔʥɥ sɛ àbɔʥɥ sɛ
11 90 cheek àfʋnʋ àfʋnʋ
12 100 face n n
13 110 forehead mʋma mʋma
14 120 head tl tl
15 150 hair (head) t l ɲɥ t l ɲɥ
16 160 hair (body) hʋna ɲɥ hʋna ɲɥ
17 170 fur ɲɥ ɲɥ
18 180 feather tkl tkl
19 190 horn ὲmɛ ὲmɛn
20 200 neck ɛkɔʋ ɛkɔʋ
21 210 throat ɛm n ɛm n
22 220 shoulder tl tl
23 230 armpit m ɔtʋ m ɔtʋ
24 236 the bush ŋwùl ŋwùl
25 250 breast nṹfṹ nṹfṹ
26 252 udder nṹfṹ nṹfṹ
27 262 milk animal‟s nṹfṹ z o nṹfṹ z o
28 265 butter sl dɛ sl d
29 280 chest kʋkʋɔ kʋkʋɔ
30 310 belly fṹ fṹ
31 320 navel fùnùm fùnùm
32 340 side of 89 body ʥ ʥ
33 342 rib ɱv ɱv
34 360 back ætɕ ætɕ
35 370 buttocks ɛtʋɔ ɛtʋɔ
36 380 tail dɥ dɥ
37 400 leg naɪdɥ naɪdɥ
38 420 thigh ɛs lɛ ɛs lɛ
39 440 knee ŋg t dw ŋg t dw
40 450 foot ɛnaɪ ɛnaɪ
41 500 arm s s
86
42 505 wing t aɪ t a nɪ
43 550 hand z z
44 560 palm z jɛ z jɛ
45 580 finger z t a z t a
46 582 thumb k kùl m t k kùl m t
47 585 toe n zʋ n zʋ
48 590 fingernail z mɔwɥ lɛ z mɔwɥ lɛ
49 591 claw mɔwɥilɛ mɔwɥilɛ
50 600 body n m dɥ n m dɥ
51 610 skin (of man) hʋn hʋn
52 611 hide (animal skin) ŋwʋma ŋwʋma
53 620 wound kl kl
54 621 scar t ɕɥ t ɕɥ
55 630 bone dm dm
56 640 meat ɛn ɛn
57 650 fat sl dɛ sl dɛ
58 651 oil ŋw ŋw
59 660 vein d d n
60 680 egg ksɥ ksɥ
61 700 blood m ʥ m ʥ
62 710 saliva d s d s
63 720 tears ni z ni z
64 740 urine ʥɥʋ z ɔ ʥɥʋ z ɔ
65 750 sweat ɱv fil ɱv fil
66 760 feces / excrement
67 800 heart k ʋma k ʋma
68 830 liver m lɛbʋɔ m lɛbʋɔ
69 860 guts / bowels jɛm d ɛ jɛm d ɛ
70 890 brain ʥɥ n ʥɥ n
71 1000 person nm nm
72 1050 children ɔf l ɔf l
73 1080 elder p ɲini p ɲini
74 1101 husband kn kn
75 1150 son ɛ l m ɛ l m
76 1151 boy l ma l ma
77 1200 woman ɛ ɛ
78 1201 wife ɛj l ɛj l
79 1250 daughter ɛa s ɛa s
80 1251 girl s s
81 1300 father p p p p
82 1310 mother m m m m
83 1320 brother nɥ i l m nɥ i l m
84 1321 elder brother nɥ i l m nɥ i l m
p ɲn p ɲn
85 1322 young brother nɥ i l m k t w nɥ i l m k t w
87
96 1330 sister nɥ b si nɥ b si
87 1331 elder sister nɥ b si p ɲ n nɥ b si p ɲ n
88 1332 younger sister nɥ b si k t w nɥ b si k t w
89 1340 friends ɲwɔŋgʋɔ ɲwɔŋgʋɔ
90 1350 mother‟s r other wɔf wɔf
91 1355 father‟s sister sw sw
92 1360 child (offspring) ɛ ɛ
93 1365 twins d fʋɔ d fʋɔ
94 1380 stranger ɔhɔhʋɔ ɔhɔhʋɔ
95 1390 enemy ɔt ɱv ɔt ɱv
96 1400 chief / king ɛh n ɛh n
97 1410 owner wùl wùl
98 1415 slave kʋ kʋ
99 1430 judge t mm f ʋɔ t mm f ʋɔ
100 1450 god ɲ m ɲ m
101 1460 spirit / ghost s maɪ s maɪ
102 1461 shadow s nz s nz
103 1470 name d d
104 1480 voice ɛn ɛn
105 1485 language k s k s
106 1490 story (tale) n z sɛ n z sɛ
107 1500 animal ʋ ʋ
108 1502 wild animal wùl ʋ wùl ʋ
109 1505 domestic animal fi ʋ fi ʋ
110 1508 sacrifice f ɔl ɔ f ɔl ɔ
111 1610 dog k l maɪ k l maɪ
112 1630 hyena p t k p t k
113 1650 cat kùsù kùsù
114 1660 lion ʥ t ʥ t
115 1670 leopard s ɔ s ɔ
116 1710 rat kùsi kùsi
117 1720 bat paɪ paɪ
118 1750 elephant ɛsʋnʋ ɛsʋnʋ
119 1790 baboon k dùl ɱvi k dùl ɱvi
120 1810 goat p ɔ ʥɪ p ɔ ʥɪ
121 1820 sheep ʥɥaɪ ʥɥaɪ
122 1850 pig pl k pl k
123 1890 horse pɔŋgɔ pɔŋgɔ
124 1900 cow (cattle) n ʥɥ n ʥɥ
125 1910 cow (female) n ʥɥ ɪd ɛ n ʥɥ ɪd ɛ
126 1920 bull n ʥɥ ɲ n n ʥɥ ɲ n
127 2000 bird n ʋmaa n ʋmaa
128 2010 chicken k ʋkɔ k ʋkɔ
129 2012 hen k ʋkɔbɪd ɛ k ʋkɔbɪd ɛ
130 2015 cock k ʋkɔɲ n k ʋkɔɲ n
88
131 2020 eagle kɔd ɛ kɔd ɛ

132 2030 vulture pɛtɛ pɛtɛ
133 2040 guinea fowl akɔɱvɛ akɔɱvɛ
134 2050 dove l l
135 2060 pigeon ʋlɔnʋma ʋlɔnʋma
136 2110 tortoise t ɕ tɕ d ɛ t ɕ tɕ d ɛ
137 2120 lizard p p
138 2130 crocodile dɛ ʥɛ dɛ ʥɛ
139 2150 snake ɛwɔ ɛwɔ
140 2190 worm (earth) ŋg l m ŋg l m
141 2200 fish nz n nz n
142 2250 crab kɔtɔ kɔtɔ
143 2310 louse eʥɥ eʥɥ
144 2320 fly (house) ŋw z n ŋw z n
145 2330 honey ɛwʋɔ ɛwʋɔ
146 2332 honeybee mɔbʋlɔ mɔbʋlɔ
147 2340 grasshopper ɛbɛ ɛbɛ
148 2350 ant (soldier) d ɛt d ɛt
149 2360 termite ɱvɔt ɛ ɱvɔt ɛ
150 2370 spider n z n z
151 2380 scorpion ɲ ɲ ʥ lɛ ɲ ɲ ʥ lɛ
152 2500 tree dɥ dɥ
153 2510 bark (tree) d ɥ t ɕi d ɥ t ɕi
154 2520 leaf ŋg mɔ ŋg mɔ
155 2530 branch ŋgʋl t ŋgʋl t
156 2535 stick
157 2540 root nd nd
158 2550 thorn k sɛ k sɛ
159 2580 fruit dɥ dɥ
160 2590 seed
161 2620 millet t ʋkʋɔ t ʋkʋɔ
162 2640 maize ùl ùl
163 2650 rice s k s k
164 2660 banana kw d kw d
165 2665 plantain bɔɔd ɛ bɔɔd ɛ
166 2670 palm tree ɛ ɛ
167 2680 yam j lɛ j lɛ
168 2685 cassava m ʥ m ʥ
169 2690 groundnut ŋg t ɛ ŋg t ɛ
170 2710 tobacco d d
171 2720 grass nz nzaɪ nz nzaɪ
172 2750 forest kw ɛ kw ɛ
173 2760 “the ush” ɱvʋfʋ ɱvʋfʋ
174 2790 field (s) ŋwùl ŋwùl
175 2810 country ɛmaɪ ɛmaɪ
89
176 2820 tribe / ethnic group ùs ɥ k ùs ɥ k

177 2830 village kùl s kùl s
178 2840 home f f
179 2850 house ɛd ɛd
180 2860 roof nʥɛnz d i nʥɛnz d i
181 2880 wall ɛaɪ ɛaɪ
182 2890 door ɛpʋnʋ ɛpʋnʋ
183 2900 gate ɛpʋnʋkɛs ɛ ɛpʋnʋkɛs ɛ
184 2910 fence ɛaɪ ɛaɪ
185 2930 path n mɔʋ kwaɪ n mɔʋ kwaɪ
186 2940 road ɛkwaɪ ɛkwaɪ
187 2970 well (water) ɛ l ɛ l
188 3000 thing (object) dɛ dɛ
189 3100 clothing t lɛ t dɛ t lɛ t dɛ
190 3120 shirt t lɛ t dɛ t lɛ t dɛ
191 3130 trousers tʋlɔz tʋlɔz
192 3150 shoe (s) mm ʋ mm ʋ
193 3180 ring (finger) k w k w
194 3200 rope h ʋma h ʋma
195 3210 string / thread h ʋma h ʋma
196 3220 mat (sleeping) m m
197 3310 chair kŋw kŋw
198 3320 stool s s ʥɥ s s ʥɥ
199 3350 drum (n) tɕɥ n tɕɥ n
200 3360 boat / canoe ɕɛma ɕɛma
201 3410 calabash / gourd ɛkʋl ɛkʋl
202 3420 basket kɛndɛɪ kɛndɛɪ
203 3430 load d sʋ d sʋ
204 3450 rubbish / garbage bɔl bɔl
205 3460 hole (in ground) m ɪna m ɪna
206 3510 mortar (grinding) w dl w dl
207 3520 pestle ʥɥʋma ʥɥʋma
208 3550 medicine dl dl
209 3555 poison wdl wdl
210 3570 salt ʥn ʥn
211 3600 pot (cooking) kk kk
212 3610 water pot (earthen) h na h na
213 3620 cooking pot (metal) tɕɛ z ɪɪ tɕɛ z ɪɪ
214 3700 iron (metal) d dɛ d dɛ
215 3710 knife s kaɪmm s kaɪmm
216 3720 bush-knife s kaɪ s kaɪ
217 3730 axe kuma kuma
218 3740 hoe sɔ sɔ
219 3750 arrow ʥaɪ ʥaɪ
220 3770 spear p p
90
221 3775 fish-spear d d

222 3780 shield (n) ŋg t sʋɔ ŋg t sʋɔ
223 3799 war kʋ kʋ
224 3810 law ɛm l ɛm l
225 3910 charcoal d d
226 3920 fire ʥ ʥ
227 3940 smoke ŋwùs ŋwùs
228 3950 ash(es) zʋ zʋ
229 4010 night (time) anaʥɥʋ anaʥɥʋ
230 4011 darkness s s
231 4050 moon bʋsʋmɪ bʋsʋmɪ
232 4051 month bʋsʋmɪ bʋsʋmɪ
233 4050 star z ʋlʋma z ʋlʋma
234 4070 sun æwɥ æwɥ
235 4071 heat of day ɛɕ ʋ ɛɕ ʋ
236 4120 morning n ɔap n ɔap
237 4180 evening æɲɥùm l æɲɥùm l
238 4210 sky w ɥi w ɥi
239 4220 cloud (s) mṹnṹŋgṹm mṹnṹŋgṹm
240 4230 fog / mist kùsùk k kùsùk k
241 4250 wind ɱvil ma ɱvil ma
242 4260 storm hṹ hṹ
243 4300 water z z
244 4316 thunder gl d gl d
245 4340 dew s s
246 4460 sea ɛpʋ ɛpʋ
247 4510 mountain p ɔ p pɔ p ɔ p pɔ
248 4512 rock ɔbʋtaɪ ɔbʋtaɪ
249 4520 stone bʋ bʋ
250 4551 ground ɛfɔm ɛfɔm
251 4560 sand ɛɲwaɪ ɛɲwaɪ
252 4570 dust ɱv t l ɱv t l
253 4585 mud atɛtɕɛ atɛtɕɛ
254 4600 year f f
255 4610 rainy season z ùtɔ l ɛ z ùtɔ l ɛ
256 4620 dry season ɛpɛ l ɛ ɛpɛ l ɛ
257 5000 how many ɛjɛ sɛɪ ɛjɛ sɛɪ
258 5010 one kʋ kʋ
259 5020 two mi nù mi nù
260 5030 three m ɛnza m ɛnza
261 5040 four ɛnaɪ ɛnaɪ
262 5050 five n n
263 5060 six nz a nz a
264 5070 seven ɛnzʋʋ ɛnzʋʋ
265 5080 eight ɛnwɔtɕɥ ɛnwɔtɕɥ
91
266 5090 nine ɛŋgʋnʋ ɛŋgʋnʋ

267 5100 ten d d
268 5120 twelve d m i nu d m i nu
269 5150 fifteen d nu d nu
270 5160 twenty dù nu dù nu
271 5170 thirty dù sa dù sa
272 5180 hundred ɛh ɛh
273 5190 thousand p p
274 5210 hot weather ɛɕ ɛɕ
275 5220 cold (weather) wɔ wɔ
276 5310 long (thing) t dn t dn
277 5311 tall t d ɪɪ t d ɪɪ
278 5312 deep m dɔ m dɔ
279 5320 short (thing) t t t t
280 5330 big kɛs ɛ kɛs ɛ
281 5240 wide ɛtilɛ ɛtilɛ
282 5350 small ktw ktw
283 5352 thin (thing) ɕɥɪaa ɕɥɪaa
284 5380 round (adj.) kùl kùl w kùl kùl w
285 5410 heavy dldl dldl
286 5412 difficult d d ɪɪ d d ɪɪ
287 5420 light in weight h l h l
288 5422 easy m lɛ m lɛ
289 5432 strong dɪɪ dɪɪ
290 5440 soft (surface) m lɛ m lɛ
291 5450 smooth tʋlʋ t ʋlʋ tʋlʋ t ʋlʋ
292 5455 shiny ɕilɛɪ ɕilɛɪ
293 5470 clean / pure ɛhʋ t ɛhʋ t
294 5475 defiled gù hʋ f gù hʋ f
295 5480 good p p p p
296 5490 bad ɛɲɛ ɛɲɛ
297 5510 bitter ɲɥʋnʋ ɲɥʋnʋ
298 5520 sour s s
299 5540 sweet dɛɛdɛɛ dɛɛdɛɛ
300 5565 truth nʋkʋlɛ nʋkʋlɛ
301 5580 false k ɕɥ k ɕɥ
302 5650 right (side) nifa s nifa s
303 5660 left side ŋ g s ŋ g s
304 5710 new fʋfʋlɔ fʋfʋlɔ
305 5720 young kùmaa kùmaa
306 5750 old (worn) p ɲn p ɲn
307 5810 all ɲinaa ɲinaa
308 5820 many / much p l p l
309 5830 crowd (of people) ɛdɔm ɛdɔm
310 5850 few k kl k kl
92
311 5860 some

312 5870 other / different fʋfʋlɔ fʋfʋlɔ
313 5910 red kɔkɔɔ kɔkɔɔ
314 5930 yellow k ʋkɔsɪl d ɪɛ k ʋkɔsɪl d ɪɛ
315 5940 green h a ɪmʋnʋ h a ɪmʋnʋ
316 5950 black tu du tu du
317 5960 blue ùlùù ùlùù
318 5980 white ff ff
319 6000 who? hwaɪ hwaɪ
320 6010 i mɪ mɪ
321 6020 you wʋ wʋ
322 6030 (s)he ɔnʋ ɔnʋ
323 6060 we jɛɪ jɛɪ
324 6070 you (pl.) mʋ mʋ
325 6080 they wɔʋ wɔʋ
326 6100 what? dɛɪ dɛɪ
327 6110 this w w
328 6120 that ɛnʋ ɛnʋ
329 6200 where? ɛhɪ ɛhɪ
330 6210 here ɛh ɛh
331 6220 there ɛhɔ ɛhɔ
332 6240 at ɛwɔ ɛwɔ
333 6250 (going) towards ɛɛkɔ ɛɛkɔ
334 6260 (coming) from ɛɛ ɛɛ
335 6280 near ɛbɛɪ ɛbɛɪ
336 6290 far nʋhʋa nʋhʋa
337 6300 before n n
338 6310 behind / after tɕil tɕil
339 6350 in mu mu
340 6370 above ɛsʋlʋ ɛsʋlʋ
341 6380 below sɛ sɛ
342 6400 how? sɛɪ sɛɪ
343 6420 with ɛn ɛn
344 6430 also ɛnzʋ ɛnzʋ
345 6435 and ɛn ɛn
346 6440 if sɛ sɛ
347 6450 when d ɛɪ d ɛɪ
348 6460 then / that time s l ɛ nʋ s l ɛ nʋ
349 6470 today ɛnɛ ɛnɛ
350 6480 yesterday ɛnʋl ɛnʋl
351 6490 tomorrow ɛtɕɪna ɛtɕɪna
352 6500 why? d ɛɪ d ɛɪ
353 6510 because d d
354 6550 no, not d d
355 6610 alive, to be t sɪ t sɪ
93
356 6615 life ŋgʋa ŋgʋa

357 6620 dirty / to become jɛ f jɛ f
358 6630 become dry wʋ wʋ
359 6640 full / filled ɛma ɛma
360 6650 ripe, be dʋ dʋ
361 6660 botten pʋlɔ pʋlɔ
362 6670 sharp na na
363 6679 blunt, dull ku ku
364 6690 become wet fɔ fɔ
365 6790 sit (down) tn s tn s
366 6770 stand up ʥɪna hɔ ʥɪna hɔ
367 6771 to be standing ʥɪna hɔ ʥɪna hɔ
368 6772 to stop ʥ ʥ
369 6780 lie down d hɔ d hɔ
370 6781 to be lying ɔd hɔ ɔd hɔ
371 6840 sleep (v) d d
372 6850 dream sʋm d sʋm d
373 6860 fear sùl sùl
374 6870 anger; be angry bʋ f bʋ f
375 6880 hunger; be hungry ɛkɔ ɛkɔ
276 6890 thirst; be thirsty z ùkɔ z ùkɔ
377 6900 shame; be ashamed n w n w
378 6910 illness; be ill jl ɛ jl ɛ
379 6920 sorrow; be sad w lɛhʋɔ w lɛhʋɔ
380 6930 joy; be joyful nʥ nʥ
381 7010 to bite k k
382 7020 eat d d
383 7025 food dɥ anɪ dɥ anɪ
384 7030 drink za za
385 7032 to smoke (something) h h
386 7050 vomit f f
387 7110 to cough; a cough bɔ ɛw bɔ ɛw
388 7115 sneeze hwɪtɪ hwɪtɪ
389 7150 suck f f
390 7152 suck (breast) nu nu
391 7170 spit t d s t d s
392 7190 blow h h
393 7192 the wind blows ɱv l ma ɔ ɱv l ma ɔ
394 7195 breathe hʋmɪ hʋmɪ
395 7230 yawn (v) hl m hl m
396 7315 song ɛɲwʋ ɛɲwʋ
397 7320 dance (v) s s
398 7325 dance (n) s s
399 7340 play (v) d g ʋlɔ d g ʋlɔ
400 7350 laugh (v) sl sl
94
401 7360 weep su su

402 7370 to bark p p
403 7372 cry out su su
404 7375 make noise j d d j d d
405 7400 say ka ka
406 7410 to talk / speak k s k s
407 7420 answer ji nʋ ji nʋ
408 7450 ask (question) s s
409 7455 ask for s ma s ma
410 7460 to command ɕɛ sʋ ɕɛ sʋ
411 7465 to rule d sʋ d sʋ
412 7470 obey d sʋ d sʋ
413 7475 refuse dù sʋ dù sʋ
414 7480 to swear an oath ka da ka da
415 7490 to curse (someone) bɔ kʋlɔ bɔ kʋlɔ
416 7495 insult d atɛ d atɛ
417 7510 see hu ad ɛ hu ad ɛ
418 7520 look at ɕɥɛ ɕɥɛ
419 7535 teach tɕ lɛ tɕ lɛ
420 7550 hear t t
421 7560 listen to t t
422 7570 to smell (something) hua hua
423 7575 smell; stink bɔʋ bɔʋ
424 7600 know (something / one) n n
425 7610 learn sua sua
426 7614 remember k ɪ k ɪ
427 7615 forget w lɛ f w lɛ f
428 7620 count (v) kaɪ kaɪ
429 7622 read kaɪ kaɪ
430 7625 book ŋwʋma ŋwʋma
431 7630 write tɕɥ lɛ tɕɥ lɛ
432 7650 think ʥɥ n ʥɥ n
433 7700 to love dɔ dɔ
434 7730 need (v) h a h a
435 7750 to seek ɕɥ ɕɥɛ ɕɥ ɕɥɛ
436 7790 carry sʋ sʋ
437 8000 take ta ta
438 8004 catch tɕ tɕ
439 8005 hold s s
440 8010 to lift / raise p ʥ p ʥ
441 8020 give ma ma
442 8022 gift atɕɛd ɛ atɕɛd ɛ
443 8030 pay tɥ tɥ
444 8035 money sk sk
445 8040 cost ɛbʋɔ ɛbʋɔ
95
446 8050 to get / receive ʥ ʥ

447 8060 steal wɥi wɥi
448 8070 to hide (something) f s f s
449 8080 lie (tell lies) k ɕɥ k ɕɥ
450 8090 deceive d d d d
451 8100 buy tɔ tɔ
452 8105 sell tɔʋ tɔʋ
453 8110 to marry (a wife) w l w l
454 8130 bear child wʋ wʋ
455 8131 bear twins wʋ d wʋ d
456 8135 be born wʋ wʋ
457 8210 to die w w
458 8220 kill ku ku
459 8250 to live t s t s
460 8310 to leave, depart ʥ hɔ kɔ ʥ hɔ kɔ
461 8320 to go (somewhere) kɔ kɔ
462 8340 arrive
463 8341 return saɪ saɪ
464 8342 go out p p
465 8334 enter dl dl
466 8344 go up kɔ sʋlʋ kɔ sʋlʋ
467 8346 follow d tɕ d tɕ
468 8347 bring f l f l
469 8350 send (someone) sʋma sʋma
470 8410 fly (v) t t
471 8415 jump hùl hùl
472 8430 swim bʋlʋ bʋlʋ
473 8431 dive tɕɥ tɕɥ
474 8432 float tɛ n tɛ n
475 8450 walk na d na d
476 8460 run tù m l k tù m l k
477 8470 fall tɔ tɔ
478 8480 turn danɪ danɪ
479 8510 scratch t t
480 8512 scratch oneself t t t t
481 8530 rub w s w s
482 8550 wipe pp pp
483 8650 pour ɕɥ ɕɥ
484 8660 wash hʋlʋ hʋlʋ
485 8662 to bathe (oneself) ʥɥ l ʥɥ l
486 8662 wash one‟s hands hʋlʋhʋ wʋ z hʋlʋhʋ wʋ z
487 8665 wash (clothes) hʋlʋ t dɪɛ hʋlʋ t dɪɛ
488 8670 sweep pl pl
489 8675 broom p l jɛ p l jɛ
490 8710 open (v)
96
491 8720 shut (v) tu tu

492 8800 break (tr)
493 8810 split (wood) p p
494 8812 tear (v tr) t t
495 8814 divide tɕɛ tɕɛ
496 8820 cut tɕɥ tɕɥ
497 8840 chop tɕɥ tɕɥ
498 8850 stab (pierce) wɔ wɔ
499 8880 fight kʋ kʋ
500 8900 hit, strike bɔ bɔ
501 8910 beat bʋ bʋ
502 8930 to hurt someone pl pl
503 8940 help (v) bʋ bʋ
504 8950 heal s s
505 8955 healer s fʋɔ s fʋɔ
506 9010 throw (v) tʋ tʋ
507 9015 throw away tʋ tɕɥɪnɪ tʋ tɕɥɪnɪ
508 9020 push (v) p a p a
509 9030 pull (v) tɕɥɪ tɕɥɪ
510 9050 press hʋ hʋ
511 9052 squeeze tɕi tɕi
512 9100 tie tɕ tɕl tɕ tɕl
513 9200 build s s
514 9210 make jɛ jɛ
515 9220 do jɛ jɛ
516 9230 work (v) jɛ ʥɥuma jɛ ʥɥuma
517 9235 work ʥɥuma ʥɥuma
518 9240 to create ɲɥ n ɲɥ n
519 9255 blacksmith ɔtʋɱvʋɔ ɔtʋɱvʋɔ
520 9260 begin ɕɛ as ɛ ɕɛ as ɛ
521 9270 finish w w
522 9280 to sew pa pa
523 9290 weave ɲɥɪnɪ ɲɥɪnɪ
524 9295 weaver ɔɲɥɪnɪfʋɔ ɔɲɥɪnɪfʋɔ
525 9400 dress ɕɛ ɕɛ
526 9450 to braid, to plait ŋwa ŋwa
527 9500 hunt (v) kɔ h kɔ h
528 9505 hunter bɔmmɔfʋɔ bɔmmɔfʋɔ
529 9510 shoot tʋ tʋ
630 9520 gun t t
531 9550 cook (v) nʋa nʋa
532 9700 cultivate dɥ dɥ
533 9710 to plow fu da m fu da m
534 9720 plant (v) dɥ dɥ
535 9730 to dig (a hole) tù m ɪna tù m ɪna
97
536 9750 bury s s

537 9800 burn (something) ɕ ɕ
538 9810 kindle sɔ sɔ
539 9830 extinguish du du
540 9900 shiver pʋ pʋ
541 9950 swell hʋnʋ hʋnʋ
98

A Study of Phonological Processes in Wassa

Uploaded by

Copyright:

Available Formats

A Study of Phonological Processes in Wassa

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Study of Phonological Processes in Wassa

Uploaded by

Copyright:

Available Formats

University of Education,Winneba http://ir.uew.edu.

UNIVERSITY OF EDUCATION, WINNEBA

A STUDY OF PHONOLOGICAL PROCESSES IN WASSA

CLEMENT KWAKU KUMI

UNIVERSITY OF EDUCATION, WINNEBA

A STUDY OF PHONOLOGICAL PROCESSES IN WASSA

CLEMENT KWAKU KUMI

A THESIS IN THE DEPARTMENT OF APPLIED LINGUISTICS,

FACULTY OF FOREIGN LANGUAGES EDUCATION AND

COMMUNICATION, SUBMITTED TO THE SCHOOL OF GRADUATE

STUDIES, UNIVERSITY OF EDUCATION, WINNEBA IN PARTIAL

FULFILLMENT OF THE REQUIREMENTS FOR AWARD OF THE

MASTER OF PHILOSOPHY (APPLIED LINGUISTICS) DEGREE

part or in whole, for another degree elsewhere.

I express my sincere gratitude to my supervisor, Prof. Thomas Essilfie for his

useful information which assisted me greatly in the writing of this work.

thanks go to my colleague graduate students.

I wish to take this opportunity to express my heartfelt thanks to my wife, Tawiah,

my late mother, siblings and to the whole family.

I dedicate this thesis to my family especially, my wife and children.

LIST OF TABLES viii

GLOSSARY AND ABBREVIATIONS xi

CHAPTER ONE: INTRODUCTION 1

1.2 Background of the Study 1

1.3 The Wassa People and their Language 2

1.4 Statement of Problem 9

1.5 Objectives of the Study 9

1.6 Research Questions 10

1.7 Significance of the Study 10

1.9 Delimitation of the Study 11

1.10 Organization of Thesis 11

CHAPTER TWO: LITERATURE REVIEW 12

2.2 Generative Phonology 12

2.3 Autosegmental Phonology 13

2.4 Basic Tenets of Autosegmental Phonology 15

2.4.1 Skeletal Tier 15

2.4.2 Universal Association Convention (UAC) 16

2.4.3 Linkage Condition 18

2.4.4 Obligatory Contour Principle (OCP 19

2.4.5 Well Formedness Condition (WFC) 20

2.5 Motivation for Autosegmental Phonology 21

2.6 Tiers Used in this Thesis 21

2.7 Distinctive Features 23

2.7.1 The Major Class Features 23

2.7.2 Cavity Features 24

2.7.3 Manner Feature 25

2.8 Literature Review 29

CHAPTER THREE: METHODOLOGY 37

3.2 Research Design 37

3.3 Selected Site 38

3.4 Population and Sampling 39

3.5 Data Collection 40

3.6 Data Analysis 40

CHAPTER FOUR: RESULTS AND FINDINGS 41

4.2 Assimilation Process 41

4.2.1 Vowel Harmony 42

4.2.2 Consonant Nasalization 52

4.2.2.1 Nasalization of Voiced Plosives and Affricates 53

4.2.3 Consonant Mutation 55