Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
144
Acquisition of complex coda and sonority among selected
bilingual Nigerian children
Grace Boluwatife Olarewaju1
University of Ibadan
Adesina B. Sunday2
University of Ibadan
Abstract
This study examined the role of sonority in the acquisition of complex coda by
Yoruba-English bilingual children with a view to determining the way children
rank constraints to arrive at their outputs. The study adopted Optimality Theory
as the theoretical framework. A wordlist of about 100 words, complemented by
relevant pictures, was used to collect the data for analysis. Spontaneous
speeches were also collected. Ten Yoruba-English bilingual children made up the
population. The data got were phonologically and acoustically analyzed. The
children’s grammar showed a preference for less sonorous consonants over
highly sonorous consonants in coda clusters through deletion and substitution
of segments. They also showed a preference for single consonants that are not
highly sonorous at the coda. They violated *COMPLEX-CODA and sometimes
NOCODA. The grammar of the Yoruba-English bilingual child does not allow for
the formation of coda clusters. This may be as a result of the influence of the
indigenous language.
Keywords: Sonorous consonants, Coda clusters, Yoruba-English bilingual child, Optimality
Theory, Children’s grammar
1. Introduction
Effective communication, which is a necessity in any human society, is
made possible through the use of language. Language cannot be divorced
from human interaction and society. This necessitates the need to acquire
language. Human beings have the innate ability to acquire language as
proposed by Noam Chomsky (Chomsky, 1975; Sunday, 2020). This inborn
ability has made the process of language acquisition easy for normally
developing children. Human beings acquire language at various stages of life
but this study focused on bilingual first language acquisition.
Language acquisition deals with the processes involved in language
development. It refers to a natural process whereby children develop
language ability in their first language effortlessly, as they do not need to
1 Bio: Grace Boluwatife Olarewaju is a PhD student in the Department of English, University
of Ibadan, Ibadan, Nigeria. Her research interests cover Child Language Acquisition and
English Phonology and Phonetics.
2 Bio: Adesina Bukunmi Sunday is a Senior Lecturer in the Department of English at
University of Ibadan, Ibadan, Nigeria. He specializes in Neurolinguistics, English
Phonology and Phonetics, and Pragmatics. He has published in reputable journals, like
English Today, Journal of Pragmatics, HUMOUR: International Journal of Humor Research,
and SKASE Journal of Theoretical Linguistics. Corr. author: sinadaybuk@gmail.com
Received : 03.12.2020
Accepted : 25.12.2020
Published : 30.12.2020
mailto:sinadaybuk@gmail.com
Acquisition of complex coda and sonority Olarewaju, Sunday
145
deliberately learn the grammar of the language (Adegbite, 2009; Meisel,
2011). In the study of language acquisition, children have been discovered to
be able to learn as many languages as they are exposed to. As a result of
this, children become bilinguals/multilinguals at a very young age. This
produces what is called Bilingual First Language Acquisition (BFLA). This is
the subconscious simultaneous acquisition of two languages by children
who have been consistently exposed to both languages (Itani-Adams, 2007;
Ramirez and Kuhl, 2016). The phonological repertoire of bilingual children is
determined by the languages that they are exposed to (Gildersleeve-
Neumann et al., 2008). There is no limit to this repertoire.
In discussing the phonology of any language, the syllable is germane,
particularly when speech is the focus. This is because it gives insights into
the phonotactics of the language. Syllables are made up of speech sounds or
phonemes; the arrangement of the phonemes is determined by the language.
Languages vary as to which type of speech sounds are allowed to occur in
which position of the syllable. The syllable is made up of onset,
nucleus/peak and coda. Only the nucleus/peak is obligatory. At times, the
nucleus/peak and coda are collapsed to form rhyme, in which case, the
syllable now has two components, namely onset and rhyme (Gussenhoven
and Jacobs, 2005; Clark, Yallop, and Fletcher, 2007; Gut, 2009).
To determine how bilingual children acquire a language that permits the
production of complex coda, this study investigated the production of
complex coda by Yoruba-English bilingual children. The Yoruba syllable does
not permit consonant cluster. Conversely, English permits optional
consonants of up to four in the coda position. The way a Yoruba child
acquires English coda will reveal some insights on cluster acquisition and
the manipulation of two different syllable systems by a child. The purpose of
this study is to unravel how children rank and re-rank constraints in their
production of complex coda and their judgement of sounds based on the
level of sonority.
For ease of understanding, there is a need to briefly examine the English and
Yoruba phonological systems.
1.1. Phonology of Yoruba
Yoruba phonology has been greatly studied by language scholars over time
(Oyebade, 2010). Yoruba, like every other language, has its sound system.
The Yoruba sound system contains eighteen consonants, seven oral vowels,
and five nasal vowels. The consonant sounds in Yoruba are /b, t, d, k, ɡ, kp,
ɡb, f, s, s̩, h, m, n, r, l, w, j/. The Yoruba vowel system consists of seven oral
vowels and five nasal vowels. The oral vowels are /i, e, e̩, a, o̩, o, u/. The
nasal vowels are /i͂, ɛ͂, a͂, õ, u͂/ (Oyebade, 2010; Arokoyo, 2012; Ogundepo,
2015). Yoruba does not distinguish between long and short vowels. The
Yoruba sound system does not permit clusters (Ogundepo, 2015).
According to Wachuku (2008:131), “the syllable structure refers to the way
vowels and consonants are arranged to form a syllable.” That is the specific
arrangement of consonants and vowels to form a syllable with adherence to
what is permissible in the language. The grammar of Yoruba permits only
open syllables. This means that it does not permit syllables to have codas.
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
146
By implication, Yoruba disallows consonant clusters. The Yoruba syllable
structure permits V, VCV, CV, and CVN structures. The syllable bears the
tone in the language. The language has monosyllabic, disyllabic and
polysyllabic words (Abiodun, 2010; Orie, 2012; Arokoyo, 2012).
1.2. Phonology of English
The English language has twenty-four consonants and vowels. The vowels
comprise twelve monophthongs, eight diphthongs and five triphthongs
(Ladefoged, 2001; Osisanwo, 2012). The consonant sounds in English are
/p, t, k, b, d, ɡ, f, v, ɵ, ð, s, z, ʃ, Ʒ, h, m, n, ŋ, r, l, w, ʧ, ʤ, j/. The English
vowel system consists of twelve monophthongs /i, i:, e, u, u:, a, æ, ͻ, ͻ:, ˄, ɜ:,
ə/. These monophthongs are further classified into long and short vowels. It
also contains eight diphthongs /ei, iə, eə, ai, au, əu, uə, ͻi/ and five
triphthongs /eiə, aiə, ͻiə, əuə, auə/ (Roach, 1997; McCully, 2009).
1.3. Syllable and Sonority
As noted by Gut (2009), the phonotactic structure of syllables is usually
described with reference to the sonority of the phonemes involved. The
sonority of a sound refers to the relative loudness of one sound in
comparison to other sounds and speech sounds are often described in terms
of their degree of sonority (Giegerich, 1992; Ladefoged, 1993). The sonority of
English speech sounds is represented on a sonority scale (Gut, 2009). In the
formation of syllables, speech sounds are organized based on two criteria,
which are Sonority Sequencing Principle and Sonority Distance. The
arrangement of segments on the sonority scale ranges from the sounds with
a high degree of sonority to sounds that are less sonorous.
Oral stops Fricatives Nasals Liquids Semi-Vowels Vowels
Voiceless Voiced Voiceless Voiced High Low
p b f v m
t d t ð n j ¡ a
k ɡ s z ŋ l r w u ɑ
s o n o r i t y
Figure1. Sonority scale
Source: Adapted from Giegerich (1992:133)
The sonority scale presented above reveals the role of natural classes in the
description of sonority. These natural classes have distinctive features that
make it easy to describe them in any phonological analysis. According to
Chomsky and Halle (1968), distinctive features are the minimal elements of
which phonetic, lexical, and phonological transcription are composed, by
combination and concatenation (Sunday, 2014). Some of the distinctive
features relevant to this study are examined here. The classification is based
Acquisition of complex coda and sonority Olarewaju, Sunday
147
on the works of scholars such as Chomsky and Halle (1968), and Schane
(1973) and Giegerich (1992):
i. SONORANT: Vowels, nasals, liquids and semi-vowel are [+sonorant] while
stops, fricatives, affricates and laryngeal glides are [-sonorant].
ii. CONTINUANT: [+continuant] are approximants and fricatives while stops
and affricates (oral and nasal) are [-continuant].
iii. CONSONANTAL: The sounds that are [+consonantal] are stops, fricatives,
affricates, nasals and liquids while laryngeal glides, vowels, and semi-vowels
are [-consonantal].
iv. SYLLABIC: This feature is needed to differentiate syllabic nasals and
liquids from their non-syllabic counterparts. Vowels, syllabic nasals, and
syllabic liquids are [+ syllabic] while obstruents, non-syllabic nasals, glides,
and non-syllabic liquids are [-syllabic].
v. VOCALIC: [+vocalic] are vowels and liquids while [-vocalic] are glides,
nasal, obstruents, devoiced vowels and devoiced liquids.
vi. NASAL: Nasal consonants are [+nasal] while oral consonants are [-nasal].
1.4. Theoretical Framework: Optimality Theory
Optimality Theory (OT) is a theory that accounts for the workings of every
language. It assumes that constraints, which are violable, are essentially
universal; as such, are present in every language. In other words,
constraints are universal. However, the point of divergence in languages is
the order in which the constraints are ranked in each language. This theory,
as proposed by Alan Prince and Paul Smolensky (1993), reflects the
resolution of conflicts among competing constraints (Kager, 1999; Oyebade,
2008; Sunday, 2013a; 2013b; 2013c; Sunday and Oyatokun, 2016; Sunday
and Olarewaju, 2020).
There are core mechanisms in OT. These are Generator (GEN) and Evaluator
(EVAL). GEN generates output forms based on the linguistic input. EVAL
evaluates the candidates generated by GEN to choose the most harmonic
candidate. It selects the optimal candidate by considering a set of ranked,
violable constraints. The role of EVAL is to assess the harmony of outputs
with respect to a given ranking of constraints. Some constraints are
considered to be more important than others as a result of domination
(dominance relation). These mechanisms are interrelated (McCarthy and
Prince, 1995; Kager, 1999).
Constraints (CON) are used to test the well-formedness of the set of
candidates in relation to what is acceptable in the grammar of each
language. Constraints are hierarchically ordered from the highest to the
lowest and they can be violated. There are two major types of constraints:
faithfulness and markedness. Faithfulness constraints require the mapping
of the input form to the output form, while markedness constraints forbid
grammatically ill-formed structures (McCarthy, 2002).
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
148
2. Methodology
The methodology of the research should be detailed very clearly referring to
relevant theories.
2.1. Participants
The data used for analysis were collected from ten Yoruba-English bilingual
children between the ages of one year, seven months (1;7) and five years
(5;0) at the start of the data collection. Yoruba is predominantly spoken in
the south-western part of Nigeria. The participants were selected from
different schools across different local government areas in Ibadan, the
capital of Oyo State, south-western, Nigeria.
2.2. Socio-demographic Analysis of the Participants
The participants were three males and seven females. They were in the age
range of 1 year, 7 months and five years. All of them were Yoruba-English
bilinguals. The participants were put in five groups, made up of two
participants each (1; 7, 2, 3, 4, and 5). The first group was for children that
were one year, seven months old. The second group was for the respondents
who were two years old. The third group catered for the respondents who
were three years old. Those children who were four years old formed the
fourth group. The last group consisted of the children who were five years
old. Each of the groups comprised two children.
Table 1
Demography of the Participants
Subject Age (at the outset) Languages Group
S1 1 year, 7 months English and Yoruba 1
S2 1 year, 7 months English and Yoruba
S3 2 years, 3 months English and Yoruba 2
S4 2 years, 1 month English and Yoruba
S5 3 years, 1 month English and Yoruba 3
S6 3 years English and Yoruba
S7 4 years English and Yoruba 4
S8 4 years English and Yoruba
S9 5 years English and Yoruba 5
S10 5 years English and Yoruba
2.3. Data collection and processing
The respondents were visited three times a week for data collection. The
entire study lasted eight weeks. Age and simultaneous bilingualism were the
criteria employed in grouping the participants. For accuracy and reliability of
data, strictly simultaneous bilingual children were sampled; they were
bilingual in Yoruba and English before the age of three years. The study
employed both longitudinal and cross-sectional designs. A wordlist of about
100 words, complemented by relevant pictures, was used to elicit the data
for analysis from the participants. Additional data were collected from
spontaneous speech of the respondents. The words were tested across the
different age groups to know the structures that each group had acquired at
Acquisition of complex coda and sonority Olarewaju, Sunday
149
their different ages. For each word, two tokens were got from two subjects
across all age groups. The forms presented were chosen based on the age of
the subject who produced it and the similarity between the two tokens got
from the subjects who represented a particular group. The subjects’
renditions are generally presented under each tableau to show the
progressive nature of the grammar of the children.
2.4. Data analysis
Perceptual and acoustic analyses were carried out, with Optimality Theory
used for explanation of the observable patterns. The syllable patterns made
by the children were grouped into stages in line with to the available age
groups. Instances of occurrences of each syllable were then identified. The
items that were repeated by the participants were not considered for
analysis, except if there was variation in their renditions. Phonetic
transcription was done for the sake of accurate description.
2.5. Realization of Coda Clusters
This section captures the analysis and interpretation of the data collected.
Both regular patterns and variations in the renditions of the participants
were identified. Nineteen English words were used to test the acquisition of
coda clusters by the children: drink, translation, instruct, aspect, point, hand,
thank, jump, translation, strange, want, socks, round, front, branch, flask,
plant, task, and strength. Only one category of coda clusters was tested: two-
consonant clusters. Different structures were considered in order to test if
onset clusters affect the production of coda clusters in words that have coda
clusters.
For the OT analysis, the following constraints were deployed:
a) ONSET: This constraint was suggested by Prince and Smolensky
(1993) and it stipulates that every syllable must have an onset. A
syllable must begin with a consonant. This constraint captures the
universal preference of language for CV syllables (McCarthy, 2008).
b) NOCODA: This constraint was also proposed by Prince and
Smolensky (1993) and it prohibits closed syllables while maximizing
open syllables. A syllable must not end with a consonant sound. It
stipulates that syllables do not have codas.
c) MAX (or MAX-IO): This requires the input segments to appear in
the output. It forbids deletion of segments (Kager, 1999).
d) *COMPLEX-ONSET: It does not permit clusters at onset (McCarthy,
2008).
e) *COMPLEX-CODA: It forbids complex stringing of sounds at coda
(McCarthy, 2008).
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
150
3. Findings
3.1. CC-Clusters
The section presents the analysis of English words with two final-consonant
clusters. The analyses of four of these words are presented below.
3.2. Analysis of ‘jump’
Jump is phonemically represented as /ʤᴧmp/. It is graphically represented
based on the hierarchy of the sounds on the sonority scale as:
Figure 2. Sonority Scale of jump
The sonority scale shows that /ʤ/ is less sonorous than /ᴧ/, which is more
sonorous than /m/ and /p/.
With respect to the acquisition of the CVCC-cluster, the children in groups
1, 2, and 3 (60% of the participants) could not properly articulate jump as
[ʤᴧmp] but as [ɛʃ] while the other four subjects pronounced it as [ʤͻmp].
Those who pronounced /ʤᴧmp/ as [ɛʃ] reduced the entire syllable into
something different from the input, thereby reducing it from its original
CVCC structure to a CV type. The syllable was replaced with two new
sounds: the open-mid front unrounded vowel /ɛ/ and the voiceless
postalveolar fricative /ʃ/. The latter sounds are sonorous sounds; as such,
this may be responsible for the replacement. It also shows that only the
subjects who were between ages three and five have acquired the CVCC
syllable type. The OT account of jump as pronounced by the participants is
shown below:
Tableau 1
The emergence of /ɛʃ/
Input / ʤᴧmp/ → Output / ɛʃ/
/ ʤᴧmp/ *COMPLEX-CODA *MAX ONSET NOCODA
(i) ʤͻmp *! *!
(ii) ʤᴧmp *! *!
(iii) ɛʃ * *
Constraint ranking: *COMPLEX-CODA>> *MAX >> ONSET >> NOCODA
Optimal candidate: /ɛʃ/
Acquisition of complex coda and sonority Olarewaju, Sunday
151
Tableau 1 above presents the analysis for the production of /ʤᴧmp/ with
respect to the emergence of the optimal candidate for the majority of the
participants. The constraints involved in the analysis are *COMPLEX-CODA,
*MAX, ONSET, and NOCODA. The grammar showed a preference for non-
clusters over clusters. And this is essentially the same in Levelt, Schiller,
and Levelt’s (2000) study. The participants also simplified the word using
both deletion and substitution. There is actually no correlation between the
output form and the input form. The constituents are entirely replaced with
other sounds. Candidate (i) and candidate (ii) are similar except for difference
in the vowel used. They violate the highest ranked constraints, thus
incurring a fatal violation for each of the constraints. However, they both
obey ONSET and NOCODA. Candidate (iii), which is the optimal candidate,
obeys the two highly ranked constraints and violates the two constraints
that are lowly ranked. The voiced palato-alveolar affricate /ʤ/ in the onset
position and the open central neutral vowel /ᴧ/ are replaced with the half-
open front spread vowel /ɛ/ with the features [+ vocalic, + sonorant, -
consonantal], while the bilabial nasal /m/, with features [-continuant, +
sonorant], and the voiceless bilabial plosive /p/, with features [+obstruent, -
continuant, +consonantal, -voice], that made up the /-mp/ cluster were
replaced with the voiceless palatal-alveolar fricative /ʃ/ with features [+
continuant, -sonorant]. The production shows the constraint ranking as:
*COMPLEX-CODA >> *MAX >> {ONSET >> NOCODA}.
Below is the Praat image of /ɛʃ/:
Figure 3. Praat image of jump
This Praat image shows the frequency and intensity values for jump as
pronounced by a year, seven months old subject which is a representation of
the production of the subjects in groups 1, 2, and partly 3. This
monosyllabic word has an intensity value of 67.1 dB and a frequency of
190.8 Hz. The values are given to guide in the interpretation of the
spectrogram.
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
152
3.3. Analysis of ‘translation’
• • •
• •
• •
• •
•
/t r æ n z. l ei. ʃ ə n/
Figure 4. Sonority scale of translation
The scale is a graphical representation of the level of sonority of each of the
speech sounds that make up the syllables in translation. The sonority scale
reveals that /t/ is less sonorous than /r/, which is less sonorous than /æ/,
which is more sonorous than /n/ and /z/. In the second syllable, /l/ is less
sonorous than /ei/. In the third syllable, /ʃ/ is less sonorous than /ə/,
which is more sonorous than /n/. It can be deduced that plosives and
fricatives are lowly ranked on the sonority hierarchy, nasals are ranked
higher than plosives and fricatives, while glides and vowels are highly
ranked.
The first syllable in the word translation has the CCVCC structure. In the
production of translation, five (50%) participants realised it as [tʰӕleɪʃən], one
(10%) participant pronounced it as [tʰӕnleɪʃən] and four (40%) participants
produced it as [tʰrӕnzleɪʃən].
Tableau 2
The emergence of /thæ.lei. ʃən /
Input /trænz.lei. ʃən/ → Output /thæ.lei. ʃən /
/trænz.lei.
ʃən/
*COMPLEX
-
ONS
ET
*COMPLEX-
CODA
NOCOD
A
MAX
☞ (i) thæ.lei.
ʃən
* *
(ii) trænz.lei.
ʃən
*! *! **
Constraint ranking: *COMPLEX-ONSET >> *COMPLEX-CODA >> NOCODA
>> MAX
Optimal candidate: /thæ.lei. ʃən /
Tableau 2 above shows the constraint ranking for the production of
translation by the majority of the participants. Particular focus is on the first
syllable, which has a complex coda. The constraints involved are
*COMPLEX-ONSET, *COMPLEX-CODA, NOCODA, and MAX. They are
hierarchically ranked as: *COMPLEX-ONSET >> *COMPLEX-CODA >>
NOCODA >> MAX.
Acquisition of complex coda and sonority Olarewaju, Sunday
153
These participants prefer unmarked structure in the onset position.
Candidate (i) violates highly ranked constraints: *COMPLEX-ONSET and
*COMPLEX-CODA. The former forbids clusters in the onset, while the former
bans clusters from appearing in the coda position. The first syllable in
translation has an initial CCVCC structure which is simplified in the optimal
output form to a CV structure. In onset position, the voiced alveolar trill /r/
is deleted, while the voiceless alveolar stop, which is the least sonorous of
the pair, is retained. Similarly, the alveolar nasal /n/, with features [-
continuant, +sonorant], and the voiced alveolar fricative /z/, with features [+
continuant, -sonorant, +voice], were deleted in coda position. Candidate (i)
emerged as the optimal candidate despite its violation of NOCODA and MAX.
Candidate (ii) fatally violates highly ranked constraints *COMPLEX-ONSET
and *COMPLEX-CODA. Consequently, it is outright eliminated from further
evaluations.
Figure 5. The spectrogram of translation
This image above is a spectrogram of translation as produced by a
participant in group 1 and it is a representation of most of the children’s
rendition of the word.
3.4. Analysis of ‘front’
Front is phonemically represented as /frᴧnt/ and it is graphically
represented on the sonority scale as:
•
•
•
•
•
/f r ᴧ n t/
Figure 6. Sonority Scale of front
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
154
The sonority scale shows that /f/ is less sonorous than /r/, which is less
sonorous than /ᴧ /. While /n/ is less sonorous than /ᴧ/, it is more
sonorous than /t/. /ᴧ/ has the highest degree of sonority.
Four (40%) out of the children used for this research pronounced front as
[thͻt], while another two (20%) pronounced it as [frͻt] and others (40%), who
were older, with almost adult proficiency, pronounced it as [frͻnt]. Those who
pronounced it as [thͻt] were between ages of 1 years, 7 months and 2 years,
3 months at the time of data. Those who pronounced the word as [thͻt]
reduced the entire syllable into something different from the input, thereby
reducing it from its original CCVCC structure to a CVC type. The syllable
was replaced with new sounds at onset and coda. For the complex coda [-nt],
the voiceless alveolar stop /t/, with features [+obstruent, +consonantal, -
sonorant], which is the less sonorous sound, is retained while the alveolar
nasal /n/, with features [-obstruent, +consonantal, +sonorant, +continuant,
+nasal], is deleted. The OT account of front is shown below:
Tableau 3
The emergence of [thͻt]
Input / frᴧnt / → Output [thͻt]
/ frᴧnt / *COMPLEX-
ONSET
*COMPLEX-
CODA
ONSET NOCODA MAX
☞ (i) thͻt * **
(ii) frᴧnt *! *! *
(iii) frͻnt *! *! * *
Constraint ranking: *COMPLEX-ONSET >> *COMPLEX-CODA >>ONSET>>
NOCODA >> MAX
Optimal candidate: [thͻt]
Tableau 3 above presents the constraint ranking for the emergence of [thͻt]
as the optimal candidate. The constraints involved are *COMPLEX-ONSET,
*COMPLEX-CODA, ONSET, NOCODA, and MAX. The subjects, in
simplifying the coda cluster /-nt/, deleted the sound with high sonority /n/
while retaining the one that is less sonorous /t/. This nasal-stop cluster
obeys the sonority condition which prescribes that the first consonant in a
complex coda must be less sonorous than the second. The grammar of the
children in these age groups (1; 7 - 2; 3) selects the sound that is least
sonorous in a cluster situation at coda position. This phenomenon is
observed in the works of Levelt et al. (2000) and Kappa (2002). The subjects’
preferred means of satisfying *COMPLEX-CODA is through deletion. In the
bid to obey *COMPLEX-CODA, the production is unfaithful to the input form
and, as such, violates MAX. Any attempt by the subjects to preserve both
consonants at the coda position will result in the violation of the highest
ranked constraint *COMPLEX-CODA, which will lead to the disqualification
of the candidate. This is the case with candidates (ii) and (iii). By violating
*COMPLEX-CODA, candidates (ii) and (iii) are knocked out of the
competition. Candidate (i) emerges as the optimal candidate, having obeyed
Acquisition of complex coda and sonority Olarewaju, Sunday
155
the highest ranked constraint, but violates constraints *NOCODA, ONSET,
and MAX.
Figure 7. Praat image of front
The waveform illustrates the production of front as enunciated by one of the
subjects and it represents the production of the majority of the participants,
especially those in groups 1, 2, and 3.
3.5. Analysis of ‘branch’
This is the phonemic representation of /brɑːnʧ/ and it is graphically
represented on the sonority scale as:
•
•
•
• •
/b r ɑ: n ʧ/
Figure 8. Sonority Scale of branch
The sonority scale shows that /b/ is less sonorous than /r/, which is less
sonorous than /ɑ:/. While /n/ is less sonorous than /ɑ:/, it is more
sonorous than /ʧ/.
In the rendition of branch, two of the groups (40%) pronounced it as [th:t],
while the production of the word by those in the third group (20%) was not
clear, thus could not be presented. The other two groups (40%) pronounced
it as [brɑ:nʧ] because they can be said to have attained adult-like use of the
language. Those who pronounced it as [thɑ:t] reduced the entire syllable into
something different from the input form, thereby reducing it from its original
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
156
CCVCC structure to a CVC type. They simplified the structure through
deletion and substitution. For the complex coda [-nʧ], the alveolar nasal /n/
with features [+consonantal, +sonorant, +nasal], which is the more sonorous
sound, is deleted, while the post-alveolar affricate /ʧ/, with features
[+obstruent, +consonantal, -continuant, +coronal, -voice], is deleted and
replaced with /t/, with features [+obstruent, +consonantal, -voice]. The
subjects’ grammar shows a preference for less sonorous sounds over sounds
that are highly sonorous. The OT account of branch is shown below:
Tableau 4
The emergence of [thɑ:t]
Input /brɑ:nʧ/ → Output [thɑ:t]
/brɑ:nʧ/ *COMPLEX-
CODA
*COMPLEX-
ONSET
ONSET NOCODA MAX
☞ (i) thɑ:t * **
(ii) brɑ:nʧ *! *! *
Constraint ranking: *COMPLEX- CODA >> *COMPLEX-ONSET >>ONSET>>
NOCODA >> MAX
Optimal candidate: [thɑ:t]
Tableau 4 above shows how [thɑ:t] emerged as the optimal candidate. The
constraints involved are *COMPLEX-CODA, *COMPLEX-ONSET, ONSET,
NOCODA, and MAX. In order to simplify the cluster /-nʧ/, the alveolar
nasal /n/, with features [+consonantal, +sonorant, +continuant, +nasal],
and the voiceless post-alveolar affricate /ʧ/, with features [+obstruent,
+consonantal, -sonorant, -continuant, +coronal, -voice], were deleted and
replaced with a less sonorous sound /t/, with features [+obstruent, -
continuant, -voice]. This is in accordance with the sonority condition, which
states that the first consonant in a complex coda must be less sonorous
than the second. In the bid to obey *COMPLEX-CODA, the subjects’ output
is unfaithful to the input form, thereby violating MAX.
Figure 9. Praat image of branch
Acquisition of complex coda and sonority Olarewaju, Sunday
157
The image in Figure 9 is a waveform of branch as produced by the subjects
in group 1 and group 2, which is a reflection of the children’s speech
production.
3.6. Analysis of ‘socks’
The phonemic representation of socks is /sͻks/; it is graphically illustrated
on the sonority scale as:
Figure 10. Sonority Scale of socks
The sonority scale shows the sonority hierarchy of each sound that make up
the word. Going by the graphical representation above, the voiceless alveolar
fricative /s/ is less sonorous than /ͻ/, which is often described as open
back rounded vowel. The voiceless velar stop /k/ is lowly ranked on the
sonority scale.
Socks was pronounced as [thͻk] rather than /sͻks/. The participants who
pronounced it as [thͻk] simplified the word by deleting one of the sounds that
make up the cluster at coda [-ks]. The voiceless velar stop /k/, with features
[+obstruent, +consonantal, -sonorant, -voice], which is the less sonorous
sound of the two, is retained, while the voiceless alveolar fricative /s/, with
features [+consonantal, -sonorant, +continuant, -voice], is deleted and
replaced with /t/, which has features [+obstruent, +consonantal, -sonorant,
-voice]. The subject’s grammar shows a preference for less sonorous sounds
over highly sonorous sounds. The OT account of socks is shown below:
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
158
Tableau 5
The emergence of [thͻk]
Input /sͻks/ → Output [thͻk]
/sͻks/ *COMPLEX-
CODA
*COMPLEX-
ONSET
ONSET NOCODA MAX
(i)
sͻks/
*! *
☞ (ii)
thͻk
* **
Constraint ranking: *COMPLEX- CODA>>*COMPLEX-ONSET>>ONSET>>
NOCODA >> MAX
Optimal candidate: [thͻk]
As seen in Tableau 5, candidate (i) violates the highest ranked constraint,
thereby incurring a fatal violation. Conversely, candidate (ii) violates two
lowly ranked constraints, NOCODA and MAX, but did not violate the highly
ranked constraints. However, in spite of the violation of two constraints,
candidate (ii) emerges as the optimal candidate. In the bid to simplify the
coda cluster /-ks/, the voiceless velar stop /k/, with features [+obstruent,
+consonantal, -sonorant, -voice], was retained, while the voiceless alveolar
fricative /s/, with features [+consonantal, +continuant, -voice], was deleted.
This is in tandem with Salami’s (2004) study that establishes the fact that
children employ different phonological processes, such as deletion, to reduce
existing structures into their preferred structures. In the bid to obey
*COMPLEX-CODA, the subject’s output is unfaithful to the input form,
thereby violating MAX. Below is the spectrograph of socks:
Figure 11. Praat image of socks
Acquisition of complex coda and sonority Olarewaju, Sunday
159
The waveform indicates that each segment is produced with a different
degree of prominence with the nucleus of the syllable having the highest
degree of prominence.
4. Conclusion
This paper examined the production of consonant clusters at the coda
position in relation to sonority by Yoruba-English bilingual Nigerian
children. The analysis revealed that the grammar of the children is quite
different from that of adults, in that it does not permit complex stringing of
consonants at the coda position. The children showed a preference for single
consonants at the coda. This choice is made based on the levels of sonority
of the sounds. The children constantly violated *COMPLEX-CODA, which
forbids the existence of more than one consonant at the coda (Kager, 1999),
and sometimes violated NOCODA. They were able to do this through
constraint re-ranking. The subjects also showed a preference for less
sonorous consonants over highly sonorous ones. The observed pattern
changes as the children mature physiologically. However, some differences
may occur as a result of other factors, like speech deficiency, interference
and lack of exposure to novel words in English. The findings of this study
have implications for early childhood education, as they reveal the kinds of
words that are suitable for teaching bilingual children, the findings could
also assist speech therapists in designing rehabilitation materials suitable
for addressing speech disorders that affect consonant clusters.
References
Abiodun, M. (2010). Basic linguistics for Nigerian languages. Ed. Yusuf, O.
Ijebu-Ode: Shebiotimo Publications. 38-65.
Adegbite, W. (2009). The psycholinguistics of English language in Nigeria.
Ibadan: Kraft Books limited.
Arokoyo, B. E. (2012). A comparative phonology of the Olukumi, Igala, Owe and
Yoruba languages. Towards proto-Niger Congo: comparison and
reconstruction. Paris. 18-21.
Chomsky, N. (1975). Reflections on language and mind. New York: Pantheon.
Chomsky, N. & Halle, M. (1968). The sound pattern of English. New York:
Harper & Row Publishers.
Clark, J., Yallop, C., & Fletcher, J. (2007). An introduction to phonetics and
phonology. Malden, MA: Blackwell Publishing.
Giegerich, H. (1992). English Phonology: an introduction. Cambridge: Cambridge
University Press.
Gildersleeve-Neumann, C., Kester, E., Davis, B., & Pena, E. (2008). English
speech sound development in preschool-aged children from bilingual
English-Spanish environments. American speech-language-hearing
association. American Speech-Language-Hearing Association 39, 314-328.
Gussenhoven, C., & Jacobs, H. (2005). Understanding phonology. 2nd Edition.
London: Hodder Arnold
Journal of Child Language Acquisition and Development – JCLAD
Vol: 8 Issue: 4 144-161, 2020, December
ISSN: 2148-1997
160
Gut, U. (2009). Introduction to phonetics and phonology. In Textbooks in
English Language and Linguistics (TELL). Hube, M. & Mukherjee, J. (eds).
Germany: Peter Lang.
Itani-Adams, Y. (2007). One child, two languages: Bilingual first language
acquisition in Japanese and English. Thesis. University of Western
Syndney. Retrieved April 4, 2018, http://www. jstor.org/stable.
Kager, R. (1999). Optimality Theory. Cambridge: Cambridge University Press.
Kappa, J. (2002). On the acquisition of syllabic structure in Greek. Journal of
Greek Linguistics 3, 1-52.
Ladefoged, P. (1993). A course in phonetics. 3rd edition. USA: Harcourt Brace &
company.
Ladefoged, P. (2001). Vowels and consonants. UK: Blackwell Publishers.
Levelt, C., Schiller, N., & Levelt, W. (2000). The acquisition of syllable types.
Language Acquisition. Vol 8(3), 237-264. Retrieved from
http://www.informaworld.com/smpp/title-content+775653668.
McCarthy, J. J. (2002). A thematic guide to Optimality Theory. Cambridge:
Cambridge University Press.
McCarthy, J. J. (2007). Doing Optimality Theory: applying theory to data.
Malden, MA: Blackwell Publishing.
McCarthy, J., & Prince, A. (1995). Faithfulness and reduplicative identity
University of Massachusetts occasional papers in linguistics 18. J.
Beckman, L.W. Dickey and S. Urbanczyk. Eds. Amherst, MA: GLSA. 249-
384.
McCully, C. (2009). The sound structure of English: an introduction. New York:
Cambridge University Press.
Meisel, J. M. (2011). First and second language acquisition: parallels and
differences. Cambridge: Cambridge University Press.
Ogundepo, A. O. (2015). A contrastive study of English and Yoruba
morphological. European Centre for Research Training and Development.
UK International Journal of English Language and Linguistics Research
3(4), 1-8.
Orie, O. O. (2012). Two cases of adaptation mismatches in Yoruba loan
phonology. Selected proceedings of the 42nd annual conference on African
linguistics. In R.M. Michael et al (Eds). Somerville, MA: Cascadilla
proceedings project. 99-111.
Osisanwo, A. (2012). Fundamentals of English phonetics and phonology. Lagos:
Femolus-Fetop Publishers.
Oyebade, F. (2008). A course in phonology. Ijebu-Ode: Shebiotimo.
Oyebade, F.O. (2010). Yoruba phonology. In Yusuf, O. (Ed.) Basic linguistics for
Nigerian Languages. Ijebu Ode: Shebiotimo. 228-246.
Prince, A., & Smolensky, P. (1993). Optimality Theory: constraints interaction
in Generative Grammar. Report no. Ruccs-TR 2, Now Brunswick, NJ:
Rutgers University Center for Cognitive Science.
Ramirez, N. F., & Kuhl, P. K. (2016). Bilingual language learning in children.
University of Washington. Retrieved March 20, 2018, from
https://ilabs.uw.edu/Bilingual_Language_Learning_in_Children.pdf.
Roach, P. (1997). English phonetics and phonology. Cambridge: Cambridge
University Press.
http://www/
http://www.informaworld.com/smpp/title-content+775653668
https://ilabs.uw.edu/Bilingual_Language_Learning_in_Children.pdf
Acquisition of complex coda and sonority Olarewaju, Sunday
161
Salami, O. L. (2004). A natural analysis of the acquisition of English
consonants by a Yoruba child. In K. Owolabi & A, Dasylva (Eds.) Forms
and functions of English and indigenous languages in Nigeria: a festschrift
in honour of Ayo Banjo. Ibadan: Group Publisher. 593-611.
Schane, S. (1973). Generative phonology. New Jersey: Prentice-Hall Inc.
Sunday, A. B. (2013a). Constraints ranking among bilingual Nigerian adult
aphasics. International Journal of English Linguistics. 3(1), 46-59.
Sunday, A. B. (2013b). Segmental phonology of bilingual Nigerian adult
Wernicke’s aphasics. World Journal of English Language. 3(1), 50-61.
Sunday, A. B. (2013c). The segmentals of bilingual Nigerian adult Broca’s
aphasics. International Journal of Applied Linguistics & English Literature.
2(3), 83-94.
Sunday A. B. (2014). Deletion, aspiration and nasalisation rules in Nigerian
English. Ibadan Journal of English Studies. 10, 21-39.
Sunday A. B. (2020). Issues in neurolinguistics. In A. Osisanwo, W. Bamgbade,
E. Igwebuike and A. Tella (Eds.) Applied linguistics, linguistic variations
and English usage in the Nigerian context: a festschrift for Moses Alo.
Ibadan: University Press Ltd pp. 17-31.
Sunday, A. B., & Oyatokun, O. O. (2016). Optimality theoretical analysis of
word-stress in educated Nigerian English. Skase Journal of Theoretical
Linguistics 13(1), 87-106.
Sunday A. B. & G. B. Olarewaju (2020). Sonority and acquisition of onset
clusters among Yoruba-English Bilingual Children. Journal of English
Scholars Association of Nigeria 22(2), 26-52
Wachuku, U. (2008). An introduction to the phonetics and phonology of English.
2nd edition. Owerri: Alphabet Nigeria publishers.