Jenny Yi-Chun Kuo (Auth.) - Mandarin Development of Indonesian Immigrants' Children - A Longitudinal Study in Taiwan-Springer Singapore (2016)

Jenny Yi-Chun Kuo
Mandarin
Development
of Indonesian
Immigrants’
Children
A Longitudinal Study in Taiwan
Mandarin Development of Indonesian
Immigrants’ Children
Jenny Yi-Chun Kuo
Mandarin Development
of Indonesian Immigrants’
Children
A Longitudinal Study in Taiwan
123
Jenny Yi-Chun Kuo
National Chiayi University
Chiayi, Taiwan
Taiwan
ISBN 978-981-10-1033-0 ISBN 978-981-10-1035-4 (eBook)

DOI 10.1007/978-981-10-1035-4
Library of Congress Control Number: 2016938656
© Springer Science+Business Media Singapore 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer Science+Business Media Singapore Pte Ltd.
Preface
This book summarizes the results of a 3-year longitudinal project on Mandarin

development by children of Indonesian mothers, the second largest
non-Mandarin-speaking immigrant group in Taiwan. These children were acquiring
their first language while interacting primarily with a second language learner of the
language. The book covers phonological, lexical, and syntactic development. It
provides a better understanding of the language development of children of
immigrants and has valuable implications for language education policy and lan-
guage acquisition theories.
v
Acknowledgements
I would like to thank the Ministry of Science and Technology for the grants to
conduct the 3-year study and write up the book. I am also grateful for my research
assistants, parents and participants of the study, and valuable comments from Prof.
James Tai and anonymous reviewers from Springer.
vii
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Previous Research on Children of Immigrants in Taiwan . . . . . . . 1
1.3 Mandarin Versus Vietnamese Versus Indonesian . . . . . . . . . . . . . 3
1.4 Research on Immigrants in Other Countries . . . . . . . . . . . . . . . . 4
1.5 Learning Chinese as a Second Language . . . . . . . . . . . . . . . . . . 5
1.6 Language Acquisition Theories . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.7 The Present Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.7.1 The Two-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 6
1.7.2 The Three-Year-Old Indonesian Boy . . . . . . . . . . . . . . . 7
1.7.3 The Three-Year-Old Indonesian Girl . . . . . . . . . . . . . . . 7
1.7.4 The Four-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . 7
1.7.5 The Four-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 8
1.7.6 The Five-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . 8
1.7.7 The Five-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 8
1.7.8 The Six-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . . 9
1.7.9 The Six-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . . 9
1.7.10 The Two-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 9
1.7.11 The Two-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 10
1.7.12 The Three-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . 10
1.7.13 The Three-Year-Old Taiwanese Girl. . . . . . . . . . . . . . . . 10
1.7.14 The Four-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 10
1.7.15 The Four-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 10
1.7.16 The Five-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 11
1.7.17 The Five-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 11
1.7.18 The Six-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . . 11
1.7.19 The Six-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . . 11
1.8 Organization of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
ix
x Contents
2 Phonological Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 General Literature on Phonological Development. . . . . . . . . . . . . 17
2.1.1 Theories of the Acquisition of Phonology . . . . . . . . . . . . 17
2.1.2 Phenomena in Phonological Development. . . . . . . . . . . . 23
2.1.3 Adult role in Phonological Development . . . . . . . . . . . . 24
2.2 Phonology of Mandarin, Taiwan Southern Min,
Vietnamese, and Indonesian . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2.1 Phonology of Mandarin . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.2 Phonology of Taiwan Southern Min. . . . . . . . . . . . . . . . 26
2.2.3 Phonology of Vietnamese . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.4 Phonology of Indonesian . . . . . . . . . . . . . . . . . . . . . . . 30
2.3 Mandarin Phonological Development by Native Children . . . . . . . 31
2.3.1 Development of Mandarin consonants . . . . . . . . . . . . . . 31
2.3.2 Development and Acquisition of Vowels
and Final Nasals . . . . . . . . . . . . . . . . . . . . . . . . . . ... 32
2.3.3 Development and Acquisition of Tones and Syllables ... 33
2.4 Mandarin Phonological Development by Children
of Vietnamese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 34
2.4.1 Mandarin Phonological Errors of Vietnamese
Mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 34
2.4.2 Mandarin Phonological Development by Children
of Vietnamese Mothers. . . . . . . . . . . . . . . . . . . . . . ... 34
of Indonesian Mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.5.1 The Two-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 36
2.5.2 The Three-Year-Old Indonesian Boy . . . . . . . . . . . . . . . 41
2.5.3 The Three-Year-Old Indonesian Girl . . . . . . . . . . . . . . . 42
2.5.4 The Four-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . 51
2.5.5 The Four-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 52
2.5.6 The Five-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . 61
2.5.7 The Five-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . 65
2.5.8 The Six-Year-Old Indonesian Boy . . . . . . . . . . . . . . . . . 69
2.5.9 The Six-Year-Old Indonesian Girl . . . . . . . . . . . . . . . . . 73
2.5.10 The Two-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 77
2.5.11 The Two-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 82
2.5.12 The Three-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . 84
2.5.13 The Three-Year-Old Taiwanese Girl. . . . . . . . . . . . . . . . 86
2.5.14 The Four-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 87
2.5.15 The Four-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 92
2.5.16 The Five-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . 93
2.5.17 The Five-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . 99
2.5.18 The Six-Year-Old Taiwanese Boy . . . . . . . . . . . . . . . . . 101
2.5.19 The Six-Year-Old Taiwanese Girl . . . . . . . . . . . . . . . . . 103
2.6 Discussion and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Contents xi
3 Lexical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

3.1 General Literature on Lexical Development . . . . . . . . . . . . . . . . 111
3.1.1 Theories of Lexical Development . . . . . . . . . . . . . . . . . 111
3.1.2 Phenomena of Lexical Development . . . . . . . . . . . . . . . 112
3.1.3 Adult Role in Lexical Development . . . . . . . . . . . . . . . . 112
3.1.4 Lexicon of Mandarin, Vietnamese and Indonesian . . . . . . 113
3.2 Mandarin Lexical Development by Native Children . . . . . . . . . . . 114
3.2.1 Vocabulary Composition . . . . . . . . . . . . . . . . . . . . . . . 114
3.2.2 Classifier Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . 114
3.3 Previous Studies of Mandarin Lexical Development
by Children of Immigrants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
3.4 Mandarin Lexical Development by Children of Indonesian
Mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
3.4.1 Noun Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3.4.2 Verb Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.4.3 Adjective Acquisition. . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.4.4 Classifier Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . 124
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4 Syntactic Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.1 General Literature on Syntactic Development . . . . . . . . . . . . . . . 131
4.1.1 Measuring Syntactic Development by Length—MLU. . . . 131
4.1.2 Acquisition of Sentences Types . . . . . . . . . . . . . . . . . . . 132
4.1.3 Measuring Syntactic Development by Sentence Types . . . 135
4.2 Mandarin Syntactic Development by Native Children. . . . . . . . . . 136
4.3 Mandarin Syntactic Development by Children of Vietnamese
Mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.4 Mandarin Syntactic Development by Children of Indonesian
Mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.4.1 Spontaneous Speech. . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.4.2 Question Elicitation Task . . . . . . . . . . . . . . . . . . . . . . . 142
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.1 Phonological Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.2 Lexical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.3 Syntactic Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.4 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
5.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
List of Figures
Figure 2.1 The mechanism of OT (Kager 1999: 22) . . . . . . . . . . . . . . 19

Figure 3.1 Vocabulary acquired by children of Indonesian
and Taiwanese mothers. . . . . . . . . . . . . . . . . . . . . . . . . . . 117
xiii
List of Tables
Table 1.1 A comparison among Mandarin, Vietnamese,

and Indonesian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Table 1.2 Participant background information . . . . . . . . . . . . . . . . . . . 12
Table 2.1 Mandarin consonants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Table 2.2 Mandarin vowel phonemes. . . . . . . . . . . . . . . . . . . . . . . . . 25
Table 2.3 Mandarin tones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Table 2.4 Tones of Taiwan Southern Min (Cheng 1997) . . . . . . . . . . . 26
Table 2.5 Examples of Taiwan Southern Min tone sandhi
(Cheng 1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Table 2.6 Consonants of Taiwan Southern Min (Cheng 1997) . . . . . . . 27
Table 2.7 Vowels of Taiwan Southern Min . . . . . . . . . . . . . . . . . . . . 28
Table 2.8 Vietnamese consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 2.9 Vietnamese vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 2.10 Vietnamese Tones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 2.11 Indonesian consonants (Macdonald and Darjowidjojo
2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 30
Table 2.12 Indonesian vowels and semivowels (Macdonald and
Darjowidjojo 2007: 6) . . . . . . . . . . . . . . . . . . . . . . ...... 31
Table 2.13 Acquisition and substitutions of the sounds from the
Indonesian girl: 2, 2 + 1, and 2 + 2 years of age . . . ...... 38
Indonesian boy: 3, 3 + 1, and 3 + 2 years of age . . . ...... 43
xv
xvi List of Tables

Indonesian girl: 5, 5 + 1, and 5 + 2 years of age . . . . ..... 66
Indonesian boy: 6, 6 + 1, and 6 + 2 years of age . . . . ..... 70
Indonesian girl: 6, 6 + 1, and 6 + 2 years of age . . . . ..... 74
Taiwanese boy: 2, 2 + 1, and 2 + 2 years of age . . . . ..... 78
Taiwanese girl: 2, 2 + 1, and 2 + 2 years of age. . . . . ..... 84
Table 2.32 Comparison of phonetic variation between children
of Indonesian and Taiwanese mothers . . . . . . . . . . . . ..... 107
Table 3.1 Semantic categories of nouns produced in child
spontaneous speech . . . . . . . . . . . . . . . . . . . . . . . . . ..... 118
Table 3.2 Semantic categories of verbs produced in child
Table 3.3 Semantic categories of adjectives produced in child
Table 3.4 Classifiers produced in child spontaneous speech . . . . ..... 125
Table 3.5 Means of children of Indonesian and Taiwanese
mothers in classifier elicitation task . . . . . . . . . . . . . . ..... 126
Table 3.6 Results of classifier comprehension task . . . . . . . . . . ..... 127
Table 4.1 MLU of children of Vietnamese and Taiwanese
by age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 138
Table 4.2 Index of Productive Syntax by children of Vietnamese
and Taiwanese . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 138
List of Tables xvii
Table 4.3 MLU for children of Indonesian and Taiwanese

mothers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Table 4.4 Mandarin Index of Productive Syntax by children
of Indonesian and Taiwanese . . . . . . . . . . . . . . . . . . . . . . . 140
Table 4.5 Accuracy rate of children’s questions . . . . . . . . . . . . . . . . . 144
List of Tableaux
Tableau 2.1 OT tableau (Dekkers et al. 2000: 3) . . . . . . . . . . . . ..... 20

Tableau 2.2 OT tableau for the Indonesian girl: 2, 2 + 1,
and 2 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 41
Tableau 2.3 OT tableau for the Indonesian boy: 3 years of age . . ..... 47
Tableau 2.4 OT tableau for the Indonesian boy: 3 + 1
and 3 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 47
Tableau 2.5 OT tableau for the Indonesian boy: 4 years of age . . ..... 57
Tableau 2.6 OT tableau for the Indonesian boy: 4 + 1
and 4 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 57
Tableau 2.7 OT tableau for the Indonesian girl: 5, 5 + 1,
and 5 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 69
Tableau 2.8 OT tableau for the Indonesian children: 6, 6 + 1,
and 6 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 77
Tableau 2.9 OT tableau for the Taiwanese male: 2, 2 + 1 years
of age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 84
Tableau 2.10 OT tableau for the Taiwanese male: 2 + 2 years
of age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 84
Tableau 2.11 OT tableau for the Taiwanese children: 3, 3 + 1,
and 3 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 91
Tableau 2.12 OT tableau for the Taiwanese children: 4, 4 + 1,
and 4 + 2 years of age . . . . . . . . . . . . . . . . . . . . . ..... 97
Tableau 2.13 OT tableau for the Taiwanese children:
5, 5 + 1, 5 + 2, and 5 + 3 years of age . . . . . . . . . ..... 101
Tableau 2.14 OT tableau for the Taiwanese children: 6 + 1 years of
age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 106
Tableau 2.15 OT tableau for the Taiwanese children: 6 and 6 + 2
years of age . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 106
xix
Chapter 1
Introduction
1.1 Background
Immigrants or children of immigrants constitute most of the population in Taiwan.

The first immigration to Taiwan was in 1621 with mostly soldiers from Mainland
China. The second immigrant flush was in 1945 when the communist took over
Mainland China. Most of them were government officers and soldiers (Su 2006).
Since 1989, the number of new immigrants by marriage increases tremendously in
Taiwan. According to Directorate-general of Budget, Accounting, and Statistics,
there are 510,930 foreign brides in 2016. Beside the 344,775 (67.48 %) from
China, Hong Kong, and Macao, who speak Chinese, there are still 166,155. Most of
them were from Southeast Asia, especially Vietnam (18.32 %), Indonesia (5.62 %),
and Thailand (1.67 %). On average, one out of eight newborns has a foreign mother
whose native language is not Mandarin. These children thus receive limited input in
Mandarin. It calls for concern whether this will affect their acquisition of Mandarin
or not. Mandarin is the common language for communication and the medium
of instruction in Taiwan. Limited proficiency in Mandarin will hinder not only
personal relationship but also cognitive development.
1.2 Previous Research on Children of Immigrants

in Taiwan
Research has found that many children of foreign mothers delayed in their devel-
opment, and language impairment seems to be the major cause (Wu 2004; Lin
2003; Ke 2004; Mo and Lai 2004; Chen and Zhang 2003; Chen and Chen 2003;
Wang 2004).
© Springer Science+Business Media Singapore 2016 1

J.Y.-C. Kuo, Mandarin Development of Indonesian Immigrants’ Children,
DOI 10.1007/978-981-10-1035-4_1
2 1 Introduction
However, previous studies on the language development of children of immi-

grants have mixed results. Some have shown that children of immigrants delayed in
their development (e.g., Chen and Lin 2006), while others show no such differences
(e.g., Huang 2006). Children of immigrants from Southeast Asia were found to use
fewer words and less complex sentences and initiated conversations less frequently
than their peers (Wang and Yang 2002). Wu (2004) indicated that the learning of
children of immigrants is limited by language communication barrier. Mo and Lai
(2004) described that children’s accented Mandarin from their immigrant mothers
has made them difficult to understand and excluded by their peers. Chen and Chen
(2003) investigated a child of a Vietnamese mother in the first grade of an ele-
mentary school. Her Mandarin Chinese was unintelligible to the teachers and was
then diagnosed as delay in language development. Wang (2004) interviewed five
teachers of kindergartens and found children of foreign mothers had poor academic
performance and the major cause was language barrier. There are also successful
cases. For example, Li and Hung (2004) conducted a case study of a three-year-old
girl with a Vietnamese mother and a Taiwanese father. The child’s oral expression
was clear and accurate. The success was attributed to parenting attitude through
observations of the child and interviews with her parents and teacher. The father
demanded on the mother and the child’s Mandarin pronunciation. Generally,
children of immigrants fall behind in the preschool age and catch up with their peers
as they enter school. Therefore, my research focuses on preschool immigrant
children.
Kuo (2008) collected spontaneous speech samples from 2- to 6-year-old children
of Vietnamese and Taiwanese mothers for three years to investigate their phono-
logical, lexical, and syntactic development. Overall, the phonological development
of children of Vietnamese mothers did not delay as claimed in the previous liter-
ature. Although there were some errors in their pronunciation, most of them were
also observed in children of Taiwanese mothers. All the children have difficulty
with retroflex, nasals, and affricates. All the children replaced retroflex obstruents
with alveolar ones just like adult Mandarin speakers in Taiwan. While younger
children tended to drop nasals in coda, older children tended to change the place of
articulation. Children of Vietnamese mothers simplified affricates to fricatives,
while children of Taiwanese mothers changed them to stops. Children in both
groups were accurate in tones regardless of Vietnamese mothers’ tone errors. The
lexical development of children of Vietnamese did not differ from the native
children in terms of productive vocabulary as claimed in the literature (Kuo 2008).
The syntactic development was measured by mean length of utterance and an
adapted version of Index of Productive Syntax (Scarborough’s 1989) adding some
peculiar structures in Mandarin. There were few differences between children of
1.2 Previous Research on Children of Immigrants in Taiwan 3
Vietnamese and Taiwanese mothers in the MLU for age 2–5. The difference
showed up at the age of 6. The MLU for the six-year-old of a Taiwanese mother
reached 6.37, while that of the Vietnamese age-mate remained 3.62. The sentence
patterns used by children of Vietnamese mothers were not fewer than those used by
children of Taiwanese mothers (Kuo 2008).
Children of immigrants from different countries can have different difficulties
when acquiring Mandarin. Although a study of language development of children
of Vietnamese mothers has been conducted (Kuo 2008), it cannot represent the
language development of all immigrant children. Immigrants’ mother tongues can
influence their acquisition of Mandarin and in turn influence their children’s lan-
guage development. Children with Vietnamese mothers were found to have unique
phonological problems and problems common in children with Taiwanese mothers
(Kuo 2008). The present study investigated Mandarin development by children of
27,036 Indonesian mothers, the second largest non-Mandarin-speaking immigrant
group in Taiwan.
1.3 Mandarin Versus Vietnamese Versus Indonesian
Vietnamese is more similar to Mandarin than Indonesian. Both Mandarin and

Vietnamese are tone languages and analytical languages and have similar word
order (SVO) (Alves 1999), whereas Indonesian is a stress language and an
agglutinative language, and OVS word order often occurs (Sneddon 2004).
Therefore, tones, which do cause problems for children of Vietnamese mothers
(Kuo 2008), could be difficult for children of Indonesian mothers as they are second
language learners with a stress language L1. A comparison among Mandarin,
Vietnamese, and Indonesian is summarized in Table 1.1.
Table 1.1 A comparison among Mandarin, Vietnamese, and Indonesian

Mandarin Vietnamese Indonesian
Tone language Tone language Stress language
Analytical language Analytical language Agglutinative language
SVO SVO OVS
Modifier head Modifier head Head modifier
4 1 Introduction
1.4 Research on Immigrants in Other Countries
Other developed countries have also faced problems of immigrant language learning.
For example, the USA is a country consisting of immigrants of various language and
culture background. The research on immigrant language has focused on the diversity
that the immigrants have brought to the society, the maintenance of immigrants’
mother tongues, and learning the language of the immigrant country (Mckay and
Wong 2000). According to the research in Canada, the use of mother tongue did not
help immigrants learn the language of the immigrant country. The proficiency of the
target language was predicted by the length of language programs and the time of
staying in the immigrant country (Fazio and Stens 1994). The role of parents plays an
important role in the language development of children of immigrants, for example,
parents’ attitude to the language, parents’ interest in the mother tongue and the local
language, and parent–child activities and conversations (Mushi 2002).
Although some of the immigrants in the USA are highly educated, different from
the foreign brides in Taiwan, the Latino workers in the USA have similar back-
ground with workers or foreign brides from Southeast Asia in Taiwan. The situation
of children of foreign brides in Taiwan is similar to children of immigrants in the
USA. Children of immigrants usually communicated in their parents’ mother ton-
gue, which is not English at home. It was hard for them to communicate with people
when they entered nursery school. They were often mistaken as cognitive deficit for
the failure of following the teacher’s instruction. However, they were fluent in the
language of the immigrant country after a few months (Kowalsky et al. 2003).
Why can’t children of immigrants in Taiwan catch up with their peer in the ele-
mentary school? The children of immigrants in the USA entered nursery school and
interacted with English-speaking children at the age of two. Children of immigrants in
Taiwan usually go to kindergarten or elementary school at five or six. Although it has
been claimed that the critical period for language acquisition is from birth to puberty,
the earlier, the better (Elman 1991). Besides, children in the nursery schools in the
USA have a lot of free time to interact with each other, through which language is
developed. The education in Taiwan focuses on collectivism. Students usually listen
to their teacher’s instruction and have little time to interact individually. Due to the
large class size, the teacher has limited time to interact with individual student. In
addition, the children of immigrants in the USA receive the standard forms of parents’
mother tongue. One the other hand, the foreign mothers are usually not allowed to
communicate with their children in their mother tongues. What their children receive
is their limited Mandarin input influenced by their native language, similar to pidgin.
The characteristic of a pidgin is simplified syntax and limited lexicon. But, when it is
acquired as the first language, the lexicon and syntax increase in amount and com-
plexity and become a full-fledged creole. Is the language spoken by children offoreign
mothers a pidgin of their parents’ language? Has it been creolized? Is this the reason
why their language is difficult to understand? This study investigated this special kind
of language. It will not only improve our understanding in creole but also have
implications for language acquisition.
1.5 Learning Chinese as a Second Language 5
1.5 Learning Chinese as a Second Language
Many minority children in Mainland China learning Chinese as a second language are
in a similar situation as children of immigrants in Taiwan. The major difference is that
they use their minority language at home. On the other hand, Indonesian mothers in
Taiwan do not usually speech Indonesian with their children. Children of immigrants in
Taiwan acquire Chinese as their first language. Gulxa et al. (2014) investigated 10
Kazak Chinese bilinguals in their first three years and found their development followed
a regular development process of the two languages: from concrete to abstract and from
simple to complex. There is an L1 transfer of Kazak Syntax to Chinese, while there is no
influence of Chinese on Kazak. They attributed this phenomenon to the predominant
use of Kazak in the families and the drastic linguistic differences between the two
languages. Sun (2010) reported the effect of school environment on minority
preschoolers’ bilingual development. There were bilingual class with all Uygur chil-
dren instructed by a Uygur Chinese bilingual teacher and Chinese Uygur mixed class,
where Chinese and Uygur children were in the same class. The instruction was given in
Chinese, and the environment was decorated in Chinese. The mixed class suppressed
the bilingual class in Chinese pronunciation, vocabulary, and syntax due to better
language input in quality and quantity. The bilingual teachers were less accurate in
Chinese pronunciation and the mixed class used Chinese more with their peer, while the
bilingual class communicate with their peer in Uygur. Input is very important for
language acquisition. The foreign mothers provided even worse Chinese input than
bilingual teachers for their children in Taiwan.
1.6 Language Acquisition Theories
There are three major language acquisition theories: behaviorism, innatism, and
interactionism. Behaviorism treated language learning as habit formation by imitation
and reinforcement (Skinner 1957). Immigrants with limited language proficiency may
not provide a good model for children to imitate and may not be capable of providing
reinforcement. Therefore, behaviorism will predict that children of immigrants may
delay in their development. With poverty of stimulus and lack of negative evidence,
Chomsky (1959) argued that people are equipped with Universal Grammar to acquire
the language in their environment. This is the called innatism. According to innatism,
children of immigrants should acquire Mandarin normally, because they are
immersed in Mandarin environment. Vygotsky (1978) proposed that language is
learned by social interaction, which is supported by Sachs and Johnson’s (1981) study
that hearing children of deaf parents did not acquire English by watching TV. The
importance of child-directed speech is emphasized. Preschool children interact with
their primary caretakers, who were Indonesian mothers in our case. With limited
Mandarin proficiency, the child-directed speech Indonesian mothers provided is
different from that of native mothers. It may limit children’s development due to
6 1 Introduction
inadequacy or helpful, because it is closer to learner’s level. According to Krashen’s

(1985) input hypothesis, the best input for learners is a little beyond learners’ current
level of proficiency. Recent research shows that children have a capacity to analyze
language input from birth, whose relation to input is mediated by their grammatical
knowledge (Lust 2006). Liu (2014) questioned UG theory by demonstrating that adult
input affects child language acquisition. There is positive correlation between fre-
quency of input and child language acquisition based on a corpus of adult and child
utterance (10–30 months). This study confirms the importance of input in language
acquisition. The present study of language development of this special population has
invaluable implication for language acquisition theories.
1.7 The Present Study
Following the study on the Mandarin acquisition of children of the largest

non-Mandarin-speaking immigrant group—Vietnamese (Kuo 2008), the present
study investigated Mandarin development by children of Indonesian mothers, the
second largest non-Mandarin-speaking immigrant group in Taiwan.
Because of great individual differences, we conducted case studies of children aged
from 2 to 6. It is a primarily cross-sectional and secondarily longitudinal study so that
we can see the developmental trend within reasonable time frame without missing the
insight of individual differences. We visited children at their homes, interviewed
parents, recorded their spontaneous conversation with children, and designed tasks to
elicit data lacking in natural discourse. All the data were transcribed in CHAT format
with Chinese characters, international phonetic alphabet, English gloss, and transla-
tion. Three research assistants transcribed the data and the transcripts were then
checked by the researcher. Then, we analyzed their phonological, lexical, and syntactic
development and compared with children of Mandarin-speaking parents. Pinyin, a
romanization system of Chinese, was used when pronunciation was not relevant.
The participants were nine children of Indonesian mothers and ten children of
Taiwanese mothers. Their age ranged from 2 to 6. There was a boy and a girl in each
age-group with a mother of each nationality except the two-year-old boy of an
Indonesian mother was not available. All the fathers were Taiwanese. For conve-
nience, they were all from Southern Taiwan. Both Mandarin and Taiwanese Southern
Min were used in the family. We focused on their acquisition of Mandarin since it is
the language to be used at school and it is also the language the mothers usually speak
to their children. The family background of each child is described below.
1.7.1 The Two-Year-Old Indonesian Girl
The two-year-old girl of an Indonesian mother is neither very extroverted nor intro-
verted. She is the third child in the family and has an eight-year-old sister and a
1.7 The Present Study 7
six-year-old brother. Her father was graduated from a senior high school in Taiwan
and her mother graduated from a junior high school in Indonesia. Her father works in a
factory as a purchase agent. Her mother is an interpreter in the Foreign Spouse Family
Service Center in Chiayi County. In the daytime, this child goes to kindergarten, so her
parents only spend three or four hours with her in the evening one day. When they have
conversation with her, both of them mostly speak Taiwanese to the child. For most of
the time, this child talked in Taiwanese. During the recording, the investigator and the
child played with toys and practiced writing numbers.
1.7.2 The Three-Year-Old Indonesian Boy
The three-year-old boy of an Indonesian mother is extroverted. His father has a

junior high school diploma and is a farmer. His mother was graduated from ele-
mentary school and takes care of the child at home. The mother has come to Taiwan
for five years but has never received instruction in Chinese. Instead, she learned
Mandarin in daily life after coming to Taiwan. The mother adapted to the life in
Taiwan very well and has no problem communicating with her child. Both parents
speak Mandarin and Taiwanese to the child. During the recording, the Indonesian
mother used blocks and toy cars to play with her child.
1.7.3 The Three-Year-Old Indonesian Girl
The three-year-old girl of an Indonesian mother is very extroverted. She is the third
child in the family and has two older sisters. Her parents were both graduated from
senior high school. Her father works in the hardware store, and her mother is a
housewife and a volunteer at the kindergarten. Her mother has studied Chinese
characters for six months. In the daytime, this child usually stays at home with her
mother or goes to her sister’s kindergarten while her mother works as a volunteer,
so her mother spends all day with her. However, her father spends only five to six
hours with her one day. This child mostly speaks Mandarin with her parents. In the
one-hour recording, the mother and the child played with some toys and were
involved in daily talking.
1.7.4 The Four-Year-Old Indonesian Boy
The four-year-old boy of an Indonesian mother is extroverted. His father has a

junior high school diploma and has a job related to recycling. His mother was
graduated from elementary school and has no job but takes care of the child at
home. The mother has come to Taiwan for 8 years, and she has received instruction
in Chinese for about six months. The Indonesian mother adapted to the life in
8 1 Introduction
Taiwan very well and has no problem in communicating with her children. Both the
parents speak Mandarin and Taiwanese to the child. During the recording, the
Indonesian mother and her child played with blocks.
1.7.5 The Four-Year-Old Indonesian Girl
The four-year-old girl and the five-year-old boy of an Indonesian mother were sister
and brother. The girl was outgoing and the boy was introverted. Their mother
adapted to Taiwan very well, and she has stayed in Taiwan for nine years. Before
she came to Taiwan, she did not attend any Mandarin courses. She has attended
Mandarin courses for five years after coming to Taiwan. Therefore, she was not
worried about her interaction with her children. Additionally, both parents were
graduated from junior high school. Her father worked for transportation, and her
mother worked as a servant part-time. The girl spent five hours with her father and
three hours with her mother a day. She usually talked to her parents in Mandarin or
Taiwanese. The child was usually looked after by her aunt, who also talked to her in
Mandarin or Taiwanese. During the recording, her mother and her aunt used blocks
and drew a picture to interact with her.
1.7.6 The Five-Year-Old Indonesian Boy
The five-year-old boy of an Indonesia mother was introverted. His mother has
stayed in Taiwan for nine years, and she adapted very well. She had not taken any
Mandarin courses before coming to Taiwan. She has taken Mandarin courses for
five years after she came to Taiwan. Besides, she was not worried about her
interaction with the children. Both parents were graduated from junior high school.
The father’s job was fruitage and the mother’s was a part-time servant. Moreover,
the child spent five hours getting along with his father and three hours getting along
with his mother a day. Additionally, he usually talked to his parents in Mandarin
and in Taiwanese. Besides, his aunt usually looked after him, and he usually talked
to her in both Mandarin and Taiwanese. His mother and aunt played with blocks
and drew a picture with him during the recording.
1.7.7 The Five-Year-Old Indonesian Girl
The five-year-old girl of an Indonesian mother is not obviously extroverted or

introverted. Her father has a junior high school diploma and now is a vendor. Her
mother was graduated from an elementary school and worked in a factory. Before
coming to Taiwan, the Indonesian mother had never received instruction in
Chinese. The mother has come to Taiwan for 5 years and learned Mandarin in daily
life. Now, she adapted to the life in Taiwan very well and has no problem in
communicating with her child. Both the parents speak Mandarin and Taiwanese to
the child. During the recording, the investigator used blocks to play with the child
and also told stories to interact with the child.
1.7.8 The Six-Year-Old Indonesian Boy
The six-year-old boy of an Indonesian mother is a very extroverted boy. He is the

second child in the family and has an eight-year-old sister and a two-year-old sister.
His father was graduated from senior high school in Taiwan and his mother graduated
from junior high school in Indonesia. His father works in the factory as a purchasing
agent. His mother is an interpreter in the Foreign Spouse Family Service Center in
Chiayi County. In the daytime, this child went to kindergarten, so his parents only
spend three or four hours with him in the evening each day. When they have con-
versation with him, both of them mostly speak Taiwanese to the child. For the most of
the time, this child talked in Taiwanese. During the recording, the Indonesian mother
and the child read storybooks, drew pictures with crayons, and did the homework.
1.7.9 The Six-Year-Old Indonesian Girl
The six-year-old girl of an Indonesian mother is an introverted girl. She is the

second child in the family and has an eight-year-old sister and a three-year-old
sister. Both of her parents were graduated from senior high school. Her father works
in the hardware store. Her mother is a housewife and sometimes is a volunteer at the
kindergarten. In the daytime, this child goes to kindergarten, so her parents only
spend five to six hours with her each day. When talking to the child, both of her
father and mother mostly speak Mandarin. For the most of the time, this child talks
in Mandarin. In the one-hour recording, the Indonesian mother and the child read
storybooks, colored pictures with crayons, and played with puzzles.
1.7.10 The Two-Year-Old Taiwanese Boy
The two-year-old boy of Taiwanese parents is outgoing. His parents were graduated
from university, and they were both teachers. His father spent three hours, while his
mother spent seven hours getting along with him a day. He usually talked to his dad
in Mandarin and usually talked to his mom in Taiwanese. He was usually looked
after by her grandmother, who talked to him in Mandarin or Taiwanese. His mother
used the blocks to interact with him during the recording.
10 1 Introduction
1.7.11 The Two-Year-Old Taiwanese Girl
The two-year-old girl of Taiwanese parents is not apparently introverted or extro-

verted. She has a younger brother, who was born only a few months ago. Both of
her parents are elementary school teachers. The child goes to a nursery school in the
daytime and is taken care of by the parents after work. The father communicates
with the child in both Mandarin and Taiwanese, while the mother communicates
with the child in Mandarin only. The child speaks Mandarin mostly.
1.7.12 The Three-Year-Old Taiwanese Boy
The three-year-old boy with Taiwanese parents is extroverted. He has a

one-year-old sister. His father studied in the graduate school, and his mother is a
nurse. The child was taken care of by his grand aunt. The parents communicate with
the child in Mandarin and Taiwanese.
1.7.13 The Three-Year-Old Taiwanese Girl
The three-year-old girl of Taiwanese parents is not very extroverted or introverted. She is
the second child in the family and has a sister, who is older than her by three years. Both of
her parents were graduated from vocational high schools. Her father works in a food
factory, and her mother is a waitress in a restaurant. In the daytime, the child is mostly taken
care of by her mother. Her father usually spends seven hours a day with her when her
mother goes to work. When they have conversation with her, both of them mostly speak
MandarinChinesetothechild.Formostofthetime,thischildtalksinMandarinChinese.In
the one-hour recording, the mother played with her with house pokers and a toy crocodile
1.7.14 The Four-Year-Old Taiwanese Boy
The four-year-old child of Taiwanese parents is an extroverted boy. He has a sister.

Both of his parents are elementary school teachers. They use Mandarin for com-
munication at home. The boy is talkative. He often recites texts and poems with his
sister and tells stories to the elders in the family.
1.7.15 The Four-Year-Old Taiwanese Girl
The four-year-old girl of Taiwanese parents was outgoing. Her parents were
graduated from university, and are both teachers. She spent three hours with her
father and seven hours with her mother a day. Additionally, she usually talked to
her dad in Mandarin and talked to her mom in Taiwanese. However, her grand-
mother usually looked after her, and she talked to her grandmother in Mandarin or
Taiwanese. During the recording, her mother used blocks to interact with her.
1.7.16 The Five-Year-Old Taiwanese Boy
The five-year-old boy of Taiwanese parents is extroverted. Both his father and
mother work as a staffs in public organizations. His father has a master’s degree and
his mother graduated from university. The child is usually taken care of by his
grandmother and his mother. The parents speak both Mandarin and Taiwanese to
the child, while the grandmother communicates with the child mostly in Taiwanese.
1.7.17 The Five-Year-Old Taiwanese Girl
The five-year-old girl with a Taiwanese mother is extroverted. She has a brother.
Her father has a master’s degree and works in industry. Her mother was graduated
from junior college and is a business woman. The child went to the kindergarten
and was usually taken care of by her mother at home. The parents communicated
with the child in Mandarin and Taiwanese.
1.7.18 The Six-Year-Old Taiwanese Boy
The six-year-old boy of Taiwanese parents is active and outgoing. He often fights
with other children at school. He has a younger brother. They like to play together.
Both of the parents are elementary school teachers. They lay heavy emphasis on
children’s education. Since the father studied in a graduate program besides work,
the children are mostly taken care of by the mother. The mother teaches first and
second grades in school. She focuses on children’s Mandarin acquisition.
1.7.19 The Six-Year-Old Taiwanese Girl
The six-year-old girl of Taiwanese parents is an extroverted girl. Both her father
and mother have senior high school diplomas. Her father runs an iron factory, and
her mother works as a staff in a public organization. The child is usually taken care
of by her grandmother. The parents and grandmother speak both Mandarin and
Taiwanese to the child. During the recording, the child is doing homework and the
mother asked the child some questions.
12 1 Introduction
Table 1.2 Participant background information

Nationality Age Gender Sibling Parent education Family Length
language
Indonesia 2 Female Older brother F: high school Taiwanese 9
older sister M: elementary
school
Indonesia 3 Male No F: high school Mandarin 5
M: elementary Taiwanese
school
Indonesia 3 Female Older sisters Mandarin 8
Indonesia 4 Male No F: high school Mandarin 8
school
Indonesia 4 Female Old brother Both: high school Mandarin 9
Taiwanese
Indonesia 5 Male Younger sister Both: high school Mandarin 9
Taiwanese
Indonesia 5 Female No F: high school Mandarin 5
school
Indonesia 6 Male Older sister Both: university Taiwanese 12
younger sister
Indonesia 6 Female Older sister Both: high school Mandarin 8
younger sister
Taiwan 2 Male No Both: university Mandarin Native
Taiwanese
Taiwan 2 Female Younger Both: university Mandarin Native
brother Taiwanese
nationality Age Gender Sibling Parent education Family Native
language
Taiwan 3 Male Younger sister Both: university Mandarin Native
Taiwanese
Taiwan 3 Female Older sister Both: high school Mandarin Native
Taiwan 4 Male Younger sister Both: university Mandarin Native
Taiwan 4 Female No Both: university Mandarin Native
Taiwanese
Taiwan 5 Male No Both: university Mandarin Native
Taiwan 5 Female Brother F: MA Mandarin Native
M: junior college Taiwanese
Taiwan 6 Male Younger Both: university Mandarin Native
brother
Taiwan 6 Female No Both: high school Mandarin Native
Taiwanese
Table 1.2 summarizes the background information of the participants. The first
column indicates the mother’s nationality, the second column indicates the chil-
dren’s age in the first year, and the third column shows the gender. The fourth
column presents parents’ education. F stands for father and M stands for mother.
Their languages used at home were given in the sixth column. The last column lists
the mothers’ length of stay in Taiwan.
1.8 Organization of the Book
In this chapter, the background of the present study has been provided. The
remaining of the book discusses phonological development in Chap. 2, lexical
development in Chap. 3, and syntactic development in Chap. 4. Finally, Chap. 5
concludes the study.
References
Alves, M. (1999). What’s so Chinese about Vietnamese? The Ninth Annual Meeting of the
Southeast Asian Linguistics Society. Berkeley: University of California.
Chomsky, N. (1959). Review of verbal behavior by B. F. Skinner. Language, 35(1), 26–58.
Elman, J. L. (1991). The importance of starting small. CRL Technical Report 9101, Center for
Research in Language, University of California, San Diego, pp. 3–9.
Fazio, L. L., & Florence, S. (1994). Using multiple regression to predict minority children’s
second language performance. Applied Linguistics, 15(4), 421–441.
Kowalsky, M., Gheiler, D., Galle, L., Goodrich, W., Kuo, J. Y., & Kwon, K. (2003). Learning a
second language in the preschool years. Tampa, USA: The Biennial Meeting of Society of
Research on Child Development.
Kuo, J. Y. (2008). Mandarin acquisition by children of Vietnamese mothers in Taiwan. Taipei:
Crane.
Krashen, S. (1985). The input hypothesis: Issues and implications. London: Longman.
Lust, B. (2006). Child language: Acquisition and growth. New York: Cambridge University Press.
Mckay, S., & Wong, L. (2000). New immigrants in the United States: Reading for second
language educators. Cambridge: Cambridge University Press.
Mushi, S. L. P. (2002). Acquisition of multiple languages among children of immigrant families:
Parent’s role in the home-school language pendulum. Early Child Development and Care, 172
(5), 517–530.
Sachs, J., Bard, B., & Johnson, M. (1981). Language learning with restricted input: Case studies of
two hearing children of deaf parents. Applied Psycholinguistics, 2(1), 33–54.
Scarborough, H. (1989). Index of productive syntax. Applied Psycholinguistics, 11, 1–22.
Skinner, B. F. (1957). Verbal behavior. New York: Appleton-Century-Crofts.
Sneddon, J. N. (2004). The Indonesian language: Its history and role in modern society.
Kensington: UNSW Press.
Sun, M-X., 孫明霞. (2010). 語言環境與少數民族學前兒童雙語習得探析 “Yuyan huanjing yu
shaoshu minzu xueqian ertong shuangyu xide tanxi” [The study of how minority children learn
second languages in preschool education.]. Language and Translation, 2, 64–68.
Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press.
14 1 Introduction
Chinese
Chen, X.-L., & Lin, Y.-X. 陳璽琳, 林月仙. (2006). 關於外籍配偶子女語言能力研究的一些看

法 “Guanyu waiji peiou zinu yuyan nengli yanjiu de yixie kanfa” [Some opinions about the
studies on the language proficiency of children of foreign spouses]. Journal of Speech and
Hearing, 14, 1–3.
Chen, P.-Z., & Chen, X.-Y. 陳佩足, 陳曉雲. (2003). 外籍新娘子女的語言發展問題 “Waiji
xinniang zinu de yuyan fazhan wenti” [The issue of language development of foreign brides’
children]. Special Education for Elementary School, 35, 68–75.
Chen, X-Q., & Zhang, L.-J. 陳湘淇, 張麗君. (2003). 國小一年級外籍配偶子女在智力, 語文能
力及學業成就表現之研究 “Guoxiao yinianji waiji peiou zinu zai zhili, yuwen nengli ji xueye
chengjiu biaoxian zhi yanjiu” [A comparison of first graders’ intelligence, language ability and
school achievements between children of immigrant mothers and children of native taiwanese
mothers]. M.A Thesis. Tainan Teacher’s College.
Gulxa, A., & Zhang, J.-Y. 古麗夏. 阿克巴爾, 張京魚. (2014). 新疆哈薩克族兒童哈漢雙語的語
言發展特點 “Xinjiang hasake zu ertong ha han shuangyu de yuyan fazhan tedian” [The
developmental features of the Kazak and Chinese Languages in Kazak children in Xinjiang].
Journal of Xinjiang University, 42 (5), 143–146.
Huang, K.-Y. 黃國佑. (2006). 條件或偏見?關於外籍配偶孩童口語表達發展的一些感想.
“Tiaojian huo pianjian? Guanyu waiji peiou haiton kouyu biaoda fazhan de yixie ganxiang”
[Condition or bias? On the oral expression development of children of foreign spouses].
Journal of Speech and Hearing, 14, 3–5.
Ke, S.-H. 柯淑慧. (2004). 外籍母親與本籍母親之子女學業成就之比較研究—以基隆市國小
一年級為例 “Waiji muqin yu benji muqin zhi zinu xueye chengjiu zhi bijiao yanjiu—yi jilong
shi guoxiao yinianji wei li” [A Comparative Study on Academic Achievements for children
with foreign and native mothers at an Elementary School in Keelung]. M.A. Thesis. National
Taipei Teachers College.
Li, X.-L., & Hung, R.-F. 李湘凌, 洪瑞楓. (2004). 外籍配偶家庭之教養觀對其子女口語表達影
響之研究 “Waiji peiou jiating zhi jiaoyangguan dui qi zinu kouyu biaoda yingxiang zhi
yanjiu” [The research on the impact of how foreign brides’ nurture affects their children’s oral
expression]. M.A. Thesis National Hualien Teachers College.
Lin, J.-P. 林璣萍. (2003). 台灣新興的弱勢學生–外籍新娘子女學校適應現況之 “Taiwan
xinxing de ruoshi xuesheng-waiji xinniang zinu xuexiao shiying xiankuang zhi yanjiu”
[Newly Risen Minority Student in Taiwan-Study of the Adaptation Condition of the Children
of foreign Brides at School]. M.A. thesis. National Taitung University.
Liu, Y. 劉穎. (2014). 環境語言輸入與兒童語言習得相關性計量研究 “Huanjing yuyan shuru
yu ertong yuyan xide xiangguanxing jiliang yanjiu” [Qualitative research on the correlation of
language input and children’s language acquisition]. Language teaching and linguistic studies,
2014(5), 43–51.
Mo, L.-L., & Lai, P.-L. 莫藜藜, 賴佩玲. (2004). 台灣社會「少子化」與外籍配偶子女問題初
探 “Taiwan shehui「shaozihua」yu waiji peiou zinu wenti chutan” [Taiwan’s declining
birthrate phenomenon and the problem of foreign brides’ children]. Journal of social
development, 105, 55–65.
Su, J.-F. (2006). 台灣本籍與外籍新娘生育率消長原因之探討 “Taiwan benji yu waiji xinniang
shengyulu xiaozhang yuanyin zhi tantao” [The research on the growth and decline of fertility
between Taiwanese and foreign brides]. Educators and Professional Development, 23(5), 41–51.
Chinese 15
Wang, R.-X. 王瑞勳 (2004). 大陸和外籍新娘婚生子女適應與學習能力之探究 “Dalu he waiji

xinniang hunsheng zinu shiying yu xuexi nengli zhi tanjiu” [The study of adaptability and
learning capacity within Chinese and foreign brides’ children]. Taiwan Education Review, 626,
25–31.
Wang, X.-H., & Yang, Y.-M. 王秀紅, 楊詠梅 (2002). 東南亞跨國婚姻婦女的健康 “Dongnanya
kuaguo hunyin funu de jiankang” [The Health of Southeast Asian Women in Transnational
Marriages in Taiwan]. The Journal of Nursing, 49, 35–41.
Wu, Q.-S. 吳清山. (2004). 外籍新娘子女教育問題及其因應策略 “Waiji xinniang zinu jiaoyu
wenti ji qi yinying celue” [The education issue of foreign brides’ children and the adaptive
strategy]. Teacher’s Friends, 441, 7–12.
Chapter 2
Phonological Development
2.1 General Literature on Phonological Development
2.1.1 Theories of the Acquisition of Phonology
Before understanding children’s phonological development, the theories of the

acquisition of phonology will be introduced. On the topic of phonological devel-
opment, there are two theories which present opposite positions and are respectively
advocated by different scholars. One is nativist theory, and the other is behaviorist
theory. Nativist theories of development hold the belief that children acquire lan-
guage according to their inborn programs, i.e., children assimilate and construct the
external information by themselves. That is, a nativist theory focuses on the sim-
ilarities among children. As for a behaviorist theory, it claims that children acquire
language from external reward and thus improve their pronunciation. That is, adults
may demand children to say the words which could be the objects he wanted, and
thus, children would learn to say the words.
Nevertheless, ‘regression’ in the acquisition of phonology is as key evidence
against both nativist and behaviorist theories. Menn’s (1971) study showed that
children’s pronunciation may become worse as the appearance of regression phe-
nomenon in nasal assimilation. One of his subject produced beans as [minz]
(‘means’), and then, he would apply this rule to a word which he had pronounced
correctly, i.e., down [dæwn] would become [næwn] (‘noun’). Consequently, this
kind of phenomenon provides a counter-example for the behaviorist theories, which
assume external reward for improved pronunciation. And, another regression
phenomenon of Menn’s participant is as well a disproof of nativist theories. The
participant was unable to utter a sound in new words but could maintain the correct
pronunciation in familiar words. That is, he had initial [h] in greeting words such as
hi and hello, but was unable to pronounce words beginning with /h/ such as horse
and hat. Therefore, nativist theories seem not to be flexible enough to explain such
kind of variation between words.

DOI 10.1007/978-981-10-1035-4_2
18 2 Phonological Development
Since both nativist and behaviorist theories cannot deal with regression in sound
production, an appropriate way is to adopt a cognitive problem-solving approach to
the acquisition of phonology. The cognitive problem-solving approach could offer a
better explanation for the regression phenomenon mentioned above. First, nasal-
ization regression could be the children’s overgeneralization in his trial-and-error
articulation attempts. For the second problem, the problem-solving approach tends
to claim that most of the reward is internal instead of the traditional view of
behaviorism (Gleason 2012).
2.1.1.1 Introduction to Optimality Theory (OT)
Optimality theory (OT), proposed by Prince and Smolensky (2004, 1997), is a

constraint-based theory of phonology. OT represents a language or grammar by
ranking a set of universal constraints. The constraints are universal, while the
ranking is not. Thus, languages differ in the ranking of constraints. The universal
constraints are violable, but they must be minimal. A surface form is ‘optimal’ in
that it violates the least number of constraints and the constraints are low-ranked in
a language-specific hierarchy (Kager 1999).
OT is a development of Generative Grammar, a derivational and rule-based
theory focusing on formal description and questing for universal principles from
language typology and language acquisition. Both theories aim to build Universal
Grammar (UG), but they are fundamentally different. First, Generative Grammar
defines UG as a set of inviolable principles and rule schemata (or ‘parameters’) but
from the viewpoint of OT, UG is defined as a set of violable universal constraints.
Accordingly, to explain cross-linguistic variation, earlier generative model claimed
that the cross-linguistic variation is owing to ‘parametric’ variation of inviolate
principles; however, OT assumes languages differ in rankings of universal con-
straints. Second, Generative Grammar is a derivational theory, which is motivated
by rewriting rules. Hence, the surface forms are derived from the underlying forms
through derivational rules. It is a serial order of rule application. In contrast, OT is
comparative. In the mechanism of OT, given an input, there will be infinite can-
didate output for the input. Then, those candidates will be evaluated by a
language-specific constraint hierarchy. However, there is no parsing order of con-
straints and of candidate output. Therefore, constraints are parsed all at once, and
candidate outputs are evaluated all at once (Kager 1999; Liu 1999; Prince and
Smolensky 2004, 1997; Tesar and Smolensky 2000). In the next section, the
mechanism of OT will be introduced.
2.1.1.2 The Architecture of an OT Grammar
The OT grammar is an input–output mechanism which assumes that each input has
precisely one output. The process of selecting the optimal candidate as the output is
accomplished by the following components:
2.1 General Literature on Phonological Development 19
Eval
Gen C1 >> C2 >> Cn
Candidate a
Candidate b
Input Candidate c
Candidate d Output
Candidate….
Fig. 2.1 The mechanism of OT (Kager 1999: 22)
Con The set of constraints out of which grammars are constructed (McCarthy and
Prince 1994).
Gen A function defining, for each possible i, the range of candidate linguistic
analyses available to i (McCarthy and Prince 1994).
Eval The function defined by composing all of the constraints in the order in
which they are ranked (McCarthy 2008).
The function Gen (Generator) produces a set of candidates, all of which are
logically possible analyses of the given input. Eval (Evaluator) deals with a system
of ranked constraints which is a formal construction on Con that yields the grammar
of an individual language. Therefore, the mechanism of OT is that Gen with the
given input i generates an infinite set of candidate outputs based on UG. Then, Eval
will assess the candidate outputs in terms of their relative harmony with respect to
the language’s ranking of the constraints (McCarthy and Prince 1994). The
mechanism of OT is schematically presented in Fig. 2.1, which supposes that Con
consists of C1, C2, … to the infinite Cn and that language is defined by the ranking
C1 ≫ C2 ≫ ≫ Cn.
2.1.1.3 Principles of Optimality Theory
McCarthy and Prince (1994) proposed five basic principles of OT:

a. Universality. UG provides a set of Con of constraints that are universal and
universality present in all grammars.
b. Violability. Constraints are violable; but violation is minimal.
c. Ranking. The constraints of Con are ranked on a language-particular basis; the
notion of minimal violation is defined in terms of this ranking. A grammar is a
ranking of the constraint set.
d. Inclusiveness. The constraint hierarchy evaluates a set of candidate analyses that
are admitted by very general considerations of structural well-formedness.
e. Parallelism. Best satisfaction of the constraint hierarchy is computed over the
constraint based on whole hierarchy and the whole candidate set. There is no
serial derivation.
OT is aimed to building the universality present in every grammar, so it defines

UG as a set of universal constraints.
According to the principles of violability and ranking, these universal constraints
are violable and ranked in language-specific way. That is, every grammar has its
own specific ranking of constraints. Since these constraints are violable and strictly
ranked, the output structures that minimally violate rankings are optimal and
grammatical.
On the basis of principles of inclusiveness, Gen is free to produce all possible
candidate analyses of the input that are ‘structural well-formedness,’ e.g., segmental
structure, prosodic structure, morphology, and syntax. Then, Eval evaluates can-
didate sets with respect to particular rankings of Con. Moreover, the way of
evaluation of candidate sets is parallel but not serial where one rule’s output is the
next rule’s input (Dekkers et al. 2000; Kager 1999).
2.1.1.4 OT Tableau
The construction of a grammar in OT is to determine the proper ranking of Con, and

the constraint tableau is a useful calculation device. A typical OT tableau is shown
in Tableau 2.1.
In an OT tableau, the top row presents the constraint ranking from left to right
and the first column in the left shows the possible output candidates of the given
input. In the following rows, constraint violations are given for each output
structure. Violation marks are indicated by asterisk ‘*’, and one ‘*’ means one
violation of constraint. The optimal candidate is pointed out by ‘☞’, and fatal
constraint violations are signaled by ‘!’.
As shown in Tableau 2.1, constraint ranking shows that CON1 dominates CON2
and CON3 and there are candidate A to candidate D as possible output candidates.
To select the optimal candidate output, we can examine the OT tableau from left to
right and determine which candidate(s) incur(s) the lowest number of violations for
each column. Candidate B is the only one that violates the highest ranked constraint
CON1, although B has a smaller number of violations than other candidates, this
violation is fatal. Therefore, candidate B is excluded, and the three remaining
candidates will be evaluated by CON2. Candidates A and D violate this constraint
only once, while C does so twice. It means that candidate C is excluded because its
second violation of CON2 is fatal. Hence, only candidates A and D continue to be
evaluated by CON3. Candidate D incurs two violations of CON3, but A violates it
Tableau 2.1 OT tableau CON1 CON2 CON3

(Dekkers et al. 2000: 3)
☞A * *
B *!
C **! *
D * **!
only once. Consequently, A is the optimal candidate and therefore grammatical.

Note that the total number of violations is irrelevant to evaluation. That A incurs
two violations and B incurs only one has no influence on evaluation (Dekkers et al.
2000).
2.1.1.5 Constraints
There are two primary types of constraints. One is the markedness constraints,
which militate against marked structures. The other type consists of the faithfulness
constraints which demand identity between input and output (Gnanadesikan 1996).
In the following sections, the faithfulness constraints and markedness constraints
adopted in the present study will be discussed.
2.1.1.6 Faithfulness Constraints
The faithfulness constraints adopted in the present study are shown in the following
and they are elaborated with examples:
a. Ident (aretroflex): The output must be in accord with its input in retroflex.
(Li 2006). For example, given the input /tʂ/, the candidate output [ts] violates
this constraint, while the candidate [ʂ] does not.
b. Ident (as.g): The output has to correspond to its input in aspiration (Li 2006).
For example, given the input /tɕh/, the candidate output [tɕ] violates this con-
straint, while the candidate [tsh] does not.
c. Ident (acont): The output must be equal to the input in continuance (Li 2006).
For example, given the input /ɕ/, the candidate output [tɕ] violates this con-
straint, while the candidate [s] does not.
d. Ident (aanterior): The segments between input and output share the same value
in anterior (Li 2006). For example, given the input /f/, the candidate output [x]
violates this constraint, while the candidate [p] does not.
e. Ident (astrident): The segments between input and output share the same value
in strident. For example, given the input /ʂ/, the candidate output [t] violates this
constraint, while the candidate [s] does not.
2.1.1.7 Markedness Constraints
The markedness constraints adopted in the present study are as follows:

f. *Retroflex: The retroflex segments are forbidden in the output (Li 2006).
e.g., /tʂ̄ táu/ ! [ts̩̄ táu] Zhi dao ‘know’
As shown in the example, /tʂ/ becomes [ts] in the output, which is not a
retroflex.
g. *(+s.g): This demands the output without aspiration (Li 2006).

e.g., /tɕhiòu/ ! [tɕiòu] Qiu ‘ball’
As shown in the example, aspirated palatal affricate /tɕh/ is replaced by [tɕ],
which is an unaspirated one.
h. *(+cont): This demands the feature value [-continuant] in the output (Li 2006).
e.g., /fēi tɕī / ! [pēi tɕī ] Fei ji ‘plane’
As shown in the example, /f/ in /fēi/ becomes [p] in the output, which is a
non-continuant sound.
i. *(+anterior): This demands the feature value [-anterior] in the output.
e.g., /thoù fǎ / ! [khoù fǎ ] Tou fa ‘hair’
As shown in the example, /th/ in /thoù/ becomes [kh] in the output, which is a
segment with feature [-anterior].
j. *(+strident): This demands the feature value [-strident] in the output.
e.g., /lău ʂŭ ! [lău tŭ] Lao shu ‘rat’
As shown in the example, /ʂ/ in /ʂŭ/ is substituted by [t] in the output, which is
not a strident segment.
k. *(+high): This demands the feature value [-high] in the output.
e.g., /ɕiău ʂ̀ xóu/ ! [tsiău s̩ ̀ xóu] Xiao shi hou ‘childhood’
As shown in the example, /ɕ/ in /ɕiău/ becomes [ts] in the output, which is a
non-high segment.
l. *Affricate: The affricate segment is forbidden in the output (Li 2006).
e.g., /tɕh iŋ̄ thiŋ/̀ ![ɕiŋ̄ thiŋ]̀ Qing ting ‘dragonfly’
As shown in the example, the aspirated palatal affricate /tɕh/ in /tɕhiŋ̄ / becomes
[ɕ], which is a palatal fricative but not an affricate.
m. *Stop: The stop segment is forbidden in the output.
e.g., /thú/ ! [tshú] Tu ‘vomit’
As shown in the example, the aspirated stop /th/ in /thú/ becomes [tsh] in the
output, which is not a stop segment.
Faithfulness constraints demand the input and output to share the same values in
some features, represented by ‘Ident(a__).’ Markedness constraints forbid output
from bearing some features, so it has asterisk ‘*’ as forbidness for the output.
Markedness and faithfulness constraints are conflicting with each other inherently.
As the constraints shown above, faithfulness constraints such as items a, b, c, d, e
conflict with markedness constraints such as items f, g, h, i, j, l, m in some aspects.
2.1.1.8 Constraint Demotion
OT (Prince and Smolensky 2004) made a claim that learning is a process of

reranking a set of universal constraints. The mechanism of reranking the constraints
is by demoting the constraints that would make the winner to lose in the hierarchy.
Besides, the constraints can only be demoted in the hierarchy. Therefore, those
constraints violated by the winner would be demoted in the hierarchy so that they
would be dominated by the constraints violated by the loser to ensure that each
winner is more harmonic than its competing losers (Tesar and Smolensky 2000).
However, Boersma and Hayes (1999) developed a different theory of learning, in
which the constraints can be demoted and promoted. Different from classic OT, the
mechanism of ranking constraints proposed by Boersma and Hayes (1999) is
bidirectional. For example, if there are two incorrectly ranked constraints in a
learner’s ranking, the higher ranked one is demoted and the lower ranked one is
promoted slightly in the full hierarchy of constraints. After learning, the ranking is
reranked and the two constraints are ranked closer, until the ranking is reversed
finally.
In the present study, the classic OT model is adopted to analyze the data since
the one directional mechanism of constraint ranking is simple and clear to observe
which constraint is demoted in the full hierarchy. Besides, by using OT, we can
explain the phonological development of children of Vietnamese and Indonesian
mothers by the constraint rankings and investigate whether there is any similarity or
difference with native children.
2.1.2 Phenomena in Phonological Development
In children’s early speech, there is still some babble contained in their speech.
Moreover, during the transition from prespeech to speech, children may produce
words invented by them, defined as ‘protowords.’ Protowords may with or without
adult models, but the appearance of protowords shows that children have begun to
voluntarily control over their vocalizations.
Then, when children start out to learn to pronounce, several phenomena may
appear in the process such as feature changing, cluster reduction, and assimilation.
For children, they learn adult words through trial and error. And, by descriptive
features, it would be more clearly for us to interpret children’s changing of sounds
during the process. So, the often seen patterns in early child phonology include all
initial stops being voiced, using alveolar place of articulation for both alveolar and
velar consonants, and nasal assimilation or assimilating the place of articulation.
Besides the patterns mentioned above, the accuracy of children’s perception of
sounds is also a key point for them to learn speech. For example, some children
learning English may have problem distinguishing [f] from [h] the two sounds are
similar for audition, because which leads to late acquisition of [h] (Velleman 1988).
In addition, in the early period of phonological development, word pronunciation is
affected by the length of the word and stress patterns. For example, children tend to
omit the initial syllable of a multisyllabic word as the syllable is unstressed. So, we
may hear ‘posed’ for ‘supposed.’ This phenomenon may be due to the fact that the
weakly stressed syllables are harder to perceive so that children miss hearing the
whole words.
During the process of learning speech sounds, we can see the differences in
learning strategies among children. Some children would not use a word if they
cannot produce the sounds accurately. But, other children may feel free to rearrange
adult words to fit them into their speech. So, some children may attempt one word
at a time until the word is clear to them. On the other hand, some children may try a
whole phrase at once. Nevertheless, as children begin to put words together and
become precise in their articulation, the differences in learning strategies will dis-
appear finally.
By three years of age, the basic features of the adult phonological system are
present in children’s pronunciation and most children can produce all the vowel
sounds and nearly all the consonant sounds. However, the liquids /r/ and /l/ are still
easy to be mistaken even until the age of four or five. Besides, in longer words such
as vacuum cleaner, children may still mispronounce the /v/ sound although utter
accurately in shorter word such as vase. Around the age of seven, children could
pronounce all sounds accurately (Gleason 2012).
2.1.3 Adult role in Phonological Development
If parents correct children’s pronunciation adequately, it would be helpful for

children’s learning speech. Studies of Malsheen (1980) and Bernstein Rater (1984)
have shown that parents indeed clarify their articulation as speaking to their chil-
dren at the one-word stage. Moreover, their findings indicated that parents’ clari-
fication of vowel production can be seen as a good model for children’s learning to
use (Gleason 2012). Consequently, parents or adults actually have effect on chil-
dren’s phonological development.
2.2 Phonology of Mandarin, Taiwan Southern Min,

Vietnamese, and Indonesian
Before discussing Mandarin phonological development, Mandarin phonology will

be introduced first. In addition, Taiwan Southern Min, widely spoken language in
Taiwan, a possible source of influence of the Mandarin in Taiwan, is also intro-
duced. Moreover, since Mandarin is not a native language for those immigrant
mothers, they may tend to be influenced by their first languages while learning
Mandarin, which in turn affect their children’s Mandarin phonological develop-
ment. Immigrant mothers’ native languages, Vietnamese and Indonesian, will be
discussed. Thus, this section focuses on the comparison of phonological systems
among the four different languages: Mandarin, Taiwan Southern Min, Vietnamese,
and Indonesian.
2.2 Phonology of Mandarin, Taiwan Southern Min, Vietnamese, and Indonesian 25
2.2.1 Phonology of Mandarin
Mandarin has twenty-two consonants including three palatals, which were analyzed
as consonant–glide combinations by some scholars (i.e., Cheng 1973; Duanmu
2000). I will follow the tradition in the acquisition literature treating them as single
sound for the ease of transcription and discussion. Table 2.1 shows the twenty-two
phonemes in phonetic symbols.
Mandarin has five vowel phonemes as shown in Table 2.2 (Duanmu 2000).
Mandarin has three glides [j, w, ɥ], as in [ja] ‘tooth,’ [wa] ‘frog,’ and [ɥ] ‘moon,’
which correspond with vowels [i, u, y]. Since the two sets do not contrast with each
other, the vowel symbols were used for transcription. A vowel in Mandarin can
carry four tones with contrastive meanings. Some examples are given in Table 2.3
(Ladefoged and Johnson 2015). The pitch levels are represented by five numerals
with one representing the lowest and five representing the highest.
A Mandarin syllable is made up of an optional onset limited to one consonant
and an optional coda restricted to nasals. Since our focus is acquisition, we will not
get into sophisticated phonological analysis.
Table 2.1 Mandarin consonants

Manner Place
Bilabial Labiodental Dental Retroflex Palatal Velar
Stop (vl.) p, ph t, th k, kh
Fricative (vl.) f s ʂ ɕ x
(vd.) ʐ
Affricate (vl.) ts, tsh tʂ, tʂh ʨ, ʨh
Nasal (vd.) m n ŋ
Liquid (vd.) l
Table 2.2 Mandarin vowel Front Central Back

phonemes
High i, y u
Mid ə
Low a
Table 2.3 Mandarin tones Name Description Tone value Example

Tone 1 High level 55 ma ‘mother’
Tone 2 High rising 35 mà ‘hemp’
Tone 3 Low falling rising 214 mă ‘horse’
Tone 4 High falling 53 ma ‘scold’
2.2.2 Phonology of Taiwan Southern Min
Taiwan Southern Min is a Min dialect of Chinese spoken in Taiwan. In the liter-
ature of linguistics, there is a variety of names of Taiwan Southern Min: South Min,
Southern Min, Taiwanese, and Taiwanese Hokkien. Taiwan Southern Min is a tone
language like Mandarin, but it has seven tones to distinguish lexical meanings. In
the following, tones, consonants, and vowels of Taiwan Southern Min will be
introduced.
2.2.2.1 Tones of Taiwan Southern Min
There are seven tones in Taiwan Southern Min including five long tones: Yinping,
Yinshang, Yinqu, Yangping, and Yangqu and two short tones (entering tones):
Yinru and Yangru. Every tone has two forms: juncture tone and sandhi tone.
Because of the tone sandhi phenomenon in Taiwan Southern Min, every juncture
tone has a matched sandhi tone which appears in context position. According to
Cheng (1997), seven tones of Taiwan Southern Min are shown in Table 2.4.
Entering tones, Yinru and Yangru, only exist in checked syllables, i.e., a syllable
ending with unaspirated stops: /p, t, k, ʔ/, for example, /hap5/ ‘close,’ /lak5/ ‘six,’
and /phaʔ3/ ‘beat.’ The tone sandhi phenomenon in Taiwan Southern Min results in
two tonal forms. For example, given the word ‘short,’ the juncture form is [te53] and
the sandhi form is [te55 te53]. Basically, tone sandhi is applied regressively (Chung
2002; Liu 2009). As the example shown, when the lexical item ‘short’ is alone, it is
a juncture tone [te53] and tone value is high falling; when a lexical item is followed
by another lexical item within a tone group, the tone sandhi rule is applied.
Therefore, in [te55 te53] ‘short,’ tone value of the first one [te] is realized as
high-level tone. Examples of tone sandhi phenomenon of each tone are given in
Table 2.5.
Table 2.4 Tones of Taiwan Southern Min (Cheng 1997)

Name Juncture tone Description Sandhi tone Description
Yinping 55 High level 33 Mid-level
Yinshang 53 High falling 55 High level
Yinqu 21 Low falling 53 High falling
Yinru 32 Mid-short 5 High short
Yangping 13 Low rising 33 Mid-level
Yangqu 33 Mid-level 11 Low level
Yangru 54 High short 3 Low short
Table 2.5 Examples of Juncture tone Gloss Sandhi tone Gloss

Taiwan Southern Min tone
sandhi (Cheng 1997) sa55 ‘Cloth’ sa33 kho21 ‘Cloth’
te53 ‘Short’ te55 te53 ‘Short’
kho21 ‘Pants’ kho53 tua21 ‘Waist belt’
khuaʔ32 ‘Broad’ khuaʔ5 khuaʔ32 ‘Broad’
laŋ13 ‘People’ laŋ33 laŋ13 ‘Everyone’
phĩ3 ‘Nose’ phĩ11 bi33 ‘Smell’
tit54 ‘Straight’ tit3 tit54 ‘Straight’
2.2.2.2 Consonants of Taiwan Southern Min
Based on Cheng (1997), there are eighteen consonant phonemes in Taiwan

Southern Min as shown in Table 2.6.
In Taiwan Southern Min, there are eighteen consonant phonemes including
stops, nasals, fricatives, affricates, and glides. Stops are distinguished by voice,
aspiration, and nasality. Oral stops are /p, t, k/, and /ph, th, kh/ are distinguished by
[+/− aspiration]. /b, l, g/ are the voiced counterparts of voiceless stops /p, t, k/.
Nasal stops are /m, n, ŋ/, in which /m/ and /ŋ/ are also syllabic consonants, serving
as the nucleus of a syllable.
2.2.2.3 Vowels of Taiwan Southern Min
There are six oral vowels and four nasal vowels in Taiwan Southern Min as given in
Table 2.7.
Taiwan Southern Min has oral vowels including front vowels /i, e/, central
vowels /ə, a/, and back vowels /u, ɔ/. Besides, different from Mandarin, Vietnamese,
and Indonesian, Taiwan Southern Min has four nasal vowels /ĩ, ẽ, ɔ̃, ã/, which can
stand alone even they do not occur with nasal consonants as examples shown in
Sect. 2.1 (Chung 2002).
Table 2.6 Consonants of Taiwan Southern Min (Cheng 1997)

Manner Place
Bilabial Alveolar Palatal Velar Glottal
Stop vl. unasp. /p/ /t/ /k/ /ʔ/
asp. /ph/ /th/ /kh/
vd. /b/ /l/ /g/
Nasal vd. /m/ /m/ /ŋ/
Fricative vl. /s/ /h/
Affricate vl. unasp. /ts/
asp. /tsh/
Glide vd. /j/
Note ‘vl.’ is an abbreviation for the feature voiceless, ‘vd.’ voiced, ‘unasp.’ unaspirated, and ‘asp.’
aspirated
Table 2.7 Vowels of Taiwan Front Central Back

Southern Min
Oral Nasal Oral Nasal Oral Nasal
High /i/ /ĩ/ /u/
Mid /e/ /ẽ/ /ə/ /ɔ/ /ɔ̃/
Low /a/ /ã/
(2.1) [ĩ33] ‘yard’

[pĩ53] ‘flat’
[ẽ55] ‘child’
[pẽ33] ‘illness’
[tã55] ‘bear’
[sã55] ‘three’
[ɔ̃55] ‘infant sleeps’
2.2.3 Phonology of Vietnamese
Vietnamese has northern and southern dialects. The northern dialect does not have
retroflex consonants, while the southern one does. Affricates are lacking in southern
Vietnamese, while they exist in northern Vietnamese. The consonants of the two
dialects are given in Table 2.8.
Vietnamese has ten vowels. Below is the vowel chart for the Hanoi variety, a
northern dialect, and other dialects may have different vowel inventories
(Vietnamese language 2007) (Table 2.9).
All vowels are unrounded except the back vowels. Therefore, the high front
rounded vowel [y] in Mandarin is particular challenging for Vietnamese learners of
Mandarin. On the other hand, like Mandarin, all vowels in Vietnamese carry an
inherent tone. Vietnamese tones are not identical with Mandarin tones though.
There are six tones in Vietnamese: high level, low falling, high rising,
dipping-rising, breaking-rising, and low falling constricted. The tone value is listed
in Table 2.10 (Vietnamese Language 2007; Lin 2005).
Table 2.8 Vietnamese consonants

Northern Vietnamese
Manner Place
Labial Alveolar Palatal Velar Glottal
Stop (vl.) p t, th c k
Stop (vd.) b d
Fricative (vl.) ɸ s x h
(vd.) v z ɣ
Affricate (vl.) tɕ
Nasal (vd.) m n ɲ ŋ
Approximant w l j
Southern dialect
Manner Place
Labial Alveolar Retroflex Palatal Velar Glottal
Stop (vl.) p t ʈ c k
Stop (vd.) b d
Fricative (vl.) ɸ s ʂ x h
(vd.) ʐ ɣ
Nasal (vd.) m n ɲ ŋ
Approximant w l j
Table 2.9 Vietnamese vowels

Front Central Back
High i ɨ u
Upper mid e ə/əː o
Lower mid e ɔ
Low a
Table 2.10 Vietnamese Tones

Name Description Tone Example
value
Ngang ‘level’ High level 33 ma ‘ghost’
Huyền Low falling 21 mà ‘but’
‘hanging’
sắc ‘sharp’ High rising 35 má ‘cheek, mother (southern)’
Hỏi ‘asking’ Dipping-rising 313 mả ‘tomb, grave’
Ngã ‘tumbling’ Breaking-rising 435 mã ‘horse (Sino-Vietnamese),
code’
nặng ‘heavy’ Low falling 3 mạ ‘rice seedling’
constricted
2.2.4 Phonology of Indonesian
Indonesian is not a tone language as Mandarin or Vietnamese; instead, Indonesian

is an intonation language as English. Intonation, which means distinctive pitches in
a phrase, conveys information about the syntactic components of the utterance.
However, tone is distinctive pitches within a word, which may convey both lexical
information about the meaning of the word and the grammatical function of the
word. Speakers of intonation language often find it hard to consider the tone as an
important, meaningful part of a word. Consequently, it is possible that speakers of
Indonesian may make some mistakes in pronouncing Mandarin tones.
2.2.4.1 Indonesian Consonants
Indonesian has twenty-two consonants. Twelve of them can occur in both initial
and final positions, and ten of them can only occur initially in the syllable. The
twenty-two consonants are shown in Table 2.11.
As shown in Table 2.11, Indonesian has stops including voiceless–voiced pairs
/p, b/, /t, d/, /k, g/ and one glottal stop /ʔ/, in which /p, t, k, ʔ/ can occur initially and
finally in the syllable, but /b, d, g/ only occur in the initial position. Indonesian
fricatives include /f, v/ and /s, z/ as two pairs of voiceless–voiced fricatives, one
palatal fricative /ʃ/, one velar fricative /x/, and one glottal fricative /h/. /f, s, h/ occur
both initially and finally in the syllable, but /v, z, ʃ, x/ occur only in the initial
position of a syllable. Affricates include voiceless palatal /tʃ/ and the corresponding
voiced palatal /dʒ/ and both of them only occur initially in the syllable. Besides,
Indonesian has nasals /m, n, ɲ, ŋ/, one alveolar trill /r/ and an alveolar lateral /l/.
Palatal nasal /ɲ/ only occurs initially, but /m, n, ŋ, r, l/ occur both initially and finally
in the syllable. In comparison with Vietnamese, the voiced velar stop /g/ and the
Table 2.11 Indonesian consonants (Macdonald and Darjowidjojo 2007)

Manner Place
Bilabial Labiodental Alveolar Palatal Velar Glottal
Stop vl. /p/ /t/ /k/ /ʔ/
vd. /b/ /d/ /g/
Nasal vd. /m/ /n/ /ɲ/ /ŋ/
Fricative vl. /f/ /s/ /ʃ/ /x/ /h/
vd. /v/ /z/
Affricate vl. /tʃ/
vd. /dʒ/
Trill vd. /r/
Liquid vd. /l/
Note ‘vl.’ is an abbreviation for the feature voiceless, ‘vd.’ as voiced, ‘unasp.’ as unaspirated, and
‘asp.’ as aspirated
Table 2.12 Indonesian Front Central Back

vowels and semivowels
(Macdonald and Darjowidjojo Semivowel /j/ /w/
2007: 6) High /i/
Mid /e/ /^/ /o/
Low /a/
voiceless glottal stop /ʔ/ only exist in Indonesian but not in Vietnamese. However,
Vietnamese has one voiceless aspirated alveolar stop /th/ which also exists in
Mandarin but not in Indonesian as well as a voiceless unaspirated palatal stop /c/. In
terms of fricatives, Indonesian has palatal fricative /ʃ/, but Vietnamese does not; on
the other hand, Vietnamese has velar fricative /ɣ/ which is absent in Indonesian. As
for affricates, Indonesian has palatal affricates /tʃ/ and /dʒ/ distinguished by the
feature [+/− voiced]. Mandarin has three pairs of affricates which are voiceless and
distinguished by [+/− aspirated]: alveolar affricates /ts, tsh/, retroflex affricates /tʂ,
tʂh/, and palatal affricates /tɕ, tɕh/. Vietnamese has no affricates. Both Vietnamese
and Indonesian have four nasals /m, n, ɲ, ŋ/ and an alveolar lateral /l/, but alveolar
trill /r/ exists only in Indonesian. In contrast to Mandarin, both Vietnamese and
Indonesian lack retroflex sounds; moreover, the difference between Mandarin,
Vietnamese, and Indonesian is that most of Mandarin consonants are distinguished
by the feature [+/− aspirated], but both consonants of Vietnamese and Indonesian are
distinguished by the feature [+/− voiced] (Macdonald and Darjowidjojo 2007).
2.2.4.2 Indonesian Vowels
There are two semivowels and six vowel phonemes in Indonesian and they are
presented in Table 2.12.
Indonesian has two semivowel phonemes /j/ and /w/ and both of them can occur
initially and finally in the syllable. In comparison with Vietnamese, both
Vietnamese and Indonesian have front vowels /i/ and /e/ and back rounded vowels /
u/ and /o/. Indonesian has a mid-central unrounded vowel phoneme /ʌ/ which
Vietnamese does not have. However, compared with Mandarin, both Vietnamese
and Indonesian lack high front rounded vowel /y/, which is easily replaced by high
front unrounded vowel /i/ if there is no such a sound in the native language.
2.3 Mandarin Phonological Development

by Native Children
2.3.1 Development of Mandarin consonants
According to Hsu (1996), children’s phonological development in Mandarin can be

divided into five stages. The first stage is from twelve to twenty months (1; 0–1; 8).
In this stage, children do not seem to have a productive sound system because they
are still in the period of exploration of sounds. However, they have acquired [p],
[m], [h], and [t]. It seems that the bilabial sounds are acquired quite early, but [n] is
still unstable and sometimes even deleted from pronunciation. Besides, a variety of
substitution sounds are used by the age of twenty months. For example, [l] is often
replaced by [t] and [n].
The second stage is from twenty to thirty months (1; 8–2; 6). In the stage,
children begin the active process of acquisition of a complex set of linguistic rules,
so the development becomes a major part of acquisition in this stage. In this period,
[f], [tɕh], and [tsh] begin to appear but are still unstable. [f] is often substituted by
[p]. And, at this time, [ph] and [th] are substituted by their unaspirated counterparts
although they have been acquired at the preceding stage. Likely, [t] and [n] are still
the most often substitutions for [l]. Besides, all the affricates except [tɕ], the
fricatives, and the retroflex sounds are difficult for children. So, [ɕ] is replaced by
[t], [k], or [tɕ]. Both [ts] and [s] are substituted by [t] or [tɕ]. As for [ʑ], [l] is the
most often substituted sound for it. [tʂ] and [tʂh] are often replaced by [tɕ] and
sometimes by [t]. [ʂ] has many variants such as [t], [k], and [tɕ] and English
interdental fricative [h].
The third stage is from thirty-one to thirty-eight months (2; 7–3; 2). During this
stage, children progress from putting two words together with the well-formed
simple sentences. Moreover, children still continue to add new sounds to his/her
language. In this period, the mutual substitution between [l] and [n] is observed.
Besides, all affricates and fricatives have emerged but are not very stable. The
retroflex sounds are substituted by different ways including the corresponding
affricates or fricatives and some sound approximations.
The fourth stage is from forty to forty-eight months (3; 4–4; 0). In this period,
aspirated stops such as [th] and [ph] can be pronounced, and the substitution of sounds
such as [l], [tɕh], and [ɕ] is dramatically narrowed down. As to the retroflex sounds
and [ts], [tsh], and [s], they do not make any significant progress in this period.
The final stage is from fifty-two to seventy-two (4; 4–6; 0). In this stage,
phonological development progresses considerably in that children begin to
develop his/her perceptual abilities and acquire a number of phonemic contrasts in
speech production. The number of substitution for the retroflex sounds seems to
decrease gradually. But, sometimes [ts] is still used for [tʂ], and [ts] or [tsh] is used
for [tʂh]. The factors for the delay in acquiring the retroflexes might be the influence
of Taiwanese (Hsu 1996). In fact, deretroflexion is quite common among adult
Mandarin speakers in Taiwan (Rau and Li 1994).
2.3.2 Development and Acquisition of Vowels

and Final Nasals
Since vowel production involves more invisible articulators, children may need to
adopt trial and error in order to find proper positions of articulators. By the age of
twelve months, children produce [a] first and followed by [i]. Then, by the age of
2.3 Mandarin Phonological Development by Native Children 33
thirteen and fourteen months, children are able to pronounce [u] sound. By the age
of eighteen months, children have good control over all the single vowels except
[y]. That is, they can produce vowels such as [a], [i], [u], [ə], and [o]. Moreover,
from thirteen months to eighteen months, children appear to pronounce the diph-
thongs such as [ai], [au], [ua], [ie], [ui], and [ia] although these diphthongs are not
so stable in this period. Besides, they may be used to substitute for other sounds
which are not acquired. By the age of six, children might still have problem with
[ei] and [ou]. For example, [ei] may lose its glide and pronounced as [e]; [ou] may
be pronounced as [o]. [y] may not be accurately pronounced by children till the age
of six; otherwise, it is substituted by [i] or [ə].
Triphthongs are most difficult for children. Some children even cannot produce
them accurately by the age of six. The first emergent triphthong is [iau] at around
2.6 and followed by [iou] at around 3.2; occasionally, [uai] would appear as early as
one year of age. As for [uei], it is observed very late in language development and is
often replaced by [ue].
Observations show that Chinese final nasals are very often used to substitute
diphthongs without nasal ending. Sometimes the final nasals were dropped. The
first final nasal that appeared was [uan] but was not so stable. It sometimes was
replaced by [an] and [ua]. At around 3.2, [uan] was finally stabilized. [uang] had
three alternatives for the period from 1.8 to 3.2. They are [ua], [ang], and [uang].
[yen] is replaced by [ien]. [uen] emerged around the age of 2.6 and was the only one
used accurately (Hsu 1996).
2.3.3 Development and Acquisition of Tones and Syllables
In the early stage of language development, children may rely on the prosodic
features such as stress and intonation. The declarative utterances rise gradually and
then fall in fundamental frequency. The imperative utterances rise sharply and then
fall. The questions have fundamental frequency contours which rise sharply at the
end of the utterance. Usually, the declarative and imperative intonations are
acquired at the beginning of one-word stage. The question intonation appears
around the twentieth month. Besides, stress patterns are also made by children to
show their emphasis on the intonation.
According to the literature, around one year of age, children appear to use the
level tone and falling tone though not stable. But, before the age of one year, all
rising and falling rising tones are substituted by the level or falling tones. The rising
tone begins to appear shortly after 1.0. The falling rising tone does not emerge till
the age of 1.6. By the age of 2.6, children have acquired the four tones, but they are
sometimes confused between the rising and falling rising tones.
The development of syllable structures can be divided into the following four
stages, beginning with simple structure to more complex one. In the first stage, the
syllable structures appear to be CV, V, VV, and CVC, and CV is the dominant one
in this period. In the second stage, the CVC pattern becomes stabilized and CVV
structure is new to appear in the period. Then, the CVV structure is a common one
in this period; there appeared to be nasal syllabic consonants. In the final stage, the
most complex structures appeared are VVV, CVVC, and CVVV.
There is some relationship between syllable structures and tones in children’s
development of speech. In early stage, children deal with simple structures such as
V, CV, and VV and the level, falling, and rising tones are used. The falling rising
tone begins to appear when children construct more complex syllables such as CVV
and VVV. So, it seems that the dipping tone begins to appear after acquiring
complex syllable structures (Hsu 1996).

of Vietnamese
Since adult input has effects on children’s phonology acquisition, it is important to

know the characteristics of input they received. Just like most adult second lan-
guage learners, Vietnamese mothers’ Mandarin is influenced by the first language
Vietnamese.
2.4.1 Mandarin Phonological Errors of Vietnamese

Mothers
Vietnamese immigrants in Taiwan have some common Mandarin phonological

errors. First, they are confused about the four tones. The mix of Tone 3 (rising
falling) and Tone 2 (rising) and Tone 1 (high level) and Tone 4 (falling) occurs
most frequently. The most difficult consonants are fricatives and affricates. They
tend to be confused about aspirated affricates and voiceless fricatives. They are also
confused about nasals and laterals and have difficulty distinguishing final alveolar
and velar nasal. The most frequently observed vowel error lies in the high front
rounded vowel /y/ and the dropping of glides in diphthongs (Lin 2005).
2.4.2 Mandarin Phonological Development by Children

of Vietnamese Mothers
Children of Vietnamese mothers did not delay in phonological development as

claimed in the previous literature. Despite pronunciation problems, most of them
were also observed in the children of Taiwanese mothers such as retroflexes, nasals,
and affricates. They both substituted retroflex obstruents with alveolar ones just like
adult Mandarin speakers in Taiwan. While younger children tended to delete nasals
2.4 Mandarin Phonological Development by Children of Vietnamese 35
in coda, older children tended to alter the place of articulation. Children of

Vietnamese mothers simplified affricates to fricatives, while children of Taiwanese
mothers changed them to stops. Children in both groups were accurate in tones,
regardless of the Vietnamese mothers’ tone errors.
The common difficult place of articulation is retroflex. Both children of
Vietnamese mothers and Taiwanese mothers across all ages have problems with
retroflex obstruents. They tend to replace retroflex obstruents with alveolar
obstruents. This phenomenon is also observed in the adult speakers of both
Vietnamese and Taiwanese (Lin 2005). Replacing retroflex obstruents with alveolar
obstruents is actually a characteristic of Mandarin spoken in Taiwan (Rau and Li
1994). It is the target adult form that children are achieving, which may not be
considered as an error.
In addition, children in both groups have difficulty with nasals. The
four-year-old child of a Vietnamese mother replaced an alveolar nasal [n] with an
alveolar lateral liquid [l]. The same error was found in the speech of the
two-year-old child of a Taiwanese mother, but was not observed in older children.
The opposite process—change of [l] to [n]—was found in the Vietnamese
two-year-old and the Taiwanese four-year-old as a result of nasal assimilation as in
/nali/ ‘where’ ! [nani]. Nasals in codas are most difficult for children. The
four-year-old Vietnamese and Taiwanese tend to drop nasals in codas, which are the
only coda allowed in Mandarin. The five- and six-year-old Vietnamese and
six-year-old Taiwanese changed the place of articulation of nasals in codas. The
five-year-old Vietnamese replaced a velar nasal with an alveolar nasal just like the
six-year-old Taiwanese. The six-year-old Vietnamese did the opposite, who
replaced an alveolar nasal with a velar nasal. The errors in nasals were not found in
adults, neither Vietnamese nor Taiwanese. They are due to children’s difficulty in
velum control.
The common difficult manner of articulation is affricate. Affricates are composed
of stops followed by fricatives. Children younger than five-year-olds in both groups
were found to simplify this complex phonological process. While two- and
four-year-old children of Vietnamese mothers changed affricates to fricatives, the
three- and four-year-old children of Taiwanese mothers changed affricates to stops.
Aspiration seems to confuse children, too. The two-year-old child of a Vietnamese
mother tended to pronounce aspirated stops without aspiration. The three-year-old
child of a Taiwanese mother pronounced aspirated stops as affricates. The problems
with affricates were also observed in Vietnamese mothers but not in Taiwanese
adults.
Some errors were unique to children of Vietnamese mothers such as pro-
nouncing high front rounded vowels without lip rounding and replacing a labio-
dental fricative with a velar fricative followed by a labiovelar glide. The errors in lip
rounding were observed in the two-year-olds and six-year-olds of Vietnamese
mothers. The replacement of labiodental fricatives was found in the three- and
four-year-olds of Vietnamese mothers. Such errors were also found in their
mothers. They are also quite common in adult Mandarin speakers in Taiwan due to
Taiwanese interference. However, our children of Taiwanese groups in the study

did not show difficulty in these two sounds.
As for the suprasegmental aspect, children seem to acquire well. Although
Vietnamese mothers showed errors in tones, they did not influence the children’s
speech. No tone errors were found in children’s speech. This confirms the literature
that suprasegmental is the first thing first language children acquire while is the
most difficult part for adult second language learners.
Children of Vietnamese mothers are not necessary slower language learners.
Many of the difficulties they faced are also faced by the children of Taiwanese
mothers. They can be due to innate biological constraints or poverty of input (Kuo
2008).

of Indonesian Mothers
Due to great individual differences, we conducted case studies of nineteen children

aged from 2 to 6, a boy and a girl in each age-group except that the two-year-old
Indonesian boy was missing. It is a primarily cross-sectional and secondarily lon-
gitudinal study so that we can see the developmental trend within reasonable time
frame without missing the insight of individual differences. We followed the
children for three years. We visited children at their homes, interviewed parents,
and recorded their spontaneous conversation with children. All the data were
transcribed in CHAT format with Chinese characters, international phonetic
alphabet, English gloss, and translation. In the following, we described their
phonological development and compared with children of Mandarin-speaking
parents. Finally, some samples were illustrated in OT.
2.5.1 The Two-Year-Old Indonesian Girl
2.5.1.1 Phonological Development in the First Year
The two-year-old Indonesian girl has not acquired most of the phonemes. She has
problems with fricatives, affricates, retroflexes, and the high front rounded vowel /
y/. She sometimes substituted labiodental fricative /f/ with high back vowel /u/. For
example, she pronounced /faǹ tsuà n/ ‘sail’ as [uaǹ tsuà n]. She also sometimes
deleted the velar fricative /x/. For example, she pronounced /xua/ ‘flower’ as [ua].
She changed fricatives to affricates as /sā n/ ‘three’ was pronounced as [tsā n]. The
child had problems with the aspirated alveopalatal affricate /tɕh/and sometimes
changed it to the unaspirated one [tɕ], as /tɕhī n tshaí / ‘vegetable’ was pronounced
as [tɕī n tshaí ] or the alveolar one [ts] as /fā n tɕhiè / ‘tomato’ was pronounced as
2.5 Mandarin Phonological Development by Children of Indonesian Mothers 37
[fā n tshiè ]. The alveopalatal fricative was changed to affricates at the same place of
articulation. For example, /tiá u ɕià tɕhy/́ ‘fall’ was pronounced as [tiá u tɕiá tɕhy].
She replaced retroflexes with alveolars like her mother and many adults in Taiwan.
For example, /tʂǔ / ‘cook’ was pronounced as [tsǔ ], /tʂ̄ h/ ‘eat’ was pronounced as
[ts̄ h], and /ʂ/̀ ‘ten’ was pronounced as [s̩ ].
̀ The high front rounded vowel /y/ was
replaced with the unrounded one [i] half of the time; for example, /tɕhy/́ ‘go’ was
pronounced as [tɕhi]́ .
2.5.1.2 Phonological Development in the Second Year
In the second year, aspiration was acquired. All the aspirated stops were pro-
nounced correctly. The unaspirated bilabial stop /p/ was omitted sometimes, e.g., /
pù/ ‘no’ ! [øù].
The child had acquired alveopalatal affricates. The problems with the labiodental
and retroflex still remained. She continued pronouncing the labiodental fricative /f/
as a voiceless velar fricative [x] and replacing retroflex affricates with alveolars,
e.g., /tʂuō tʂ/ ‘table’ ! [tsuō tʂ]. She was able to produce the voiced retroflex
fricative correctly, but not the voiceless one. The child showed regression in the
aspirated alveolar affricate, which was pronounced correctly in the first year.
2.5.1.3 Phonological Development in the Third Year
In the third year, the child still deleted /x/ and deaspirated the aspirated alveopalatal
affricate /tɕh/ occasionally. All the retroflex consonants were still replaced by
alveolars, and the retroflex schwa was pronounced without retroflex. Final velar
nasals were sometimes changed to alveolar nasals or dropped like the final alveolar
nasal sometimes. /u/ was deleted in /uo/ sequence occasionally.
The phonological development of the two-year-old child of an Indonesian
mother is summarized in Table 2.13. The first column lists the target phonemes.
The second column shows the phonetic representations and frequency of occur-
rences in the first year, two-year-old; the third column indicates those in the second
year, three-year-old; the last column presents those in the third year, four-year-old.
The frequency was omitted when there were no phonetic variations.
Let me give an example in OT framework. The Indonesian two-year-old girl
ranked *Retroflex constraint quite high and so replaced the retroflex fricative with
alveolar fricative for three years as shown in Tableau 2.2.
Table 2.13 Acquisition and substitutions of the sounds from the Indonesian girl: 2, 2 + 1, and 2 + 2 years of age
38
Phoneme Phonetic representations, first year Phonetic representations, second year Phonetic representations, third year
Frequency Frequency Frequency
p p p, ø p: 96.87 %, ø: 3.13 % p
ph ph 100 % ph ph
m m m m
f f, u u: 20 % f f
t t t t
th th th th
n n m n
l l l l, n l: 11/12 = 91.67 %, n: 1/12 = 8.33 %
k k k k
kh kh kh kh
x x, ø, u ø: 4 %, u: 2 % x x, ø x: 8/10 = 80 %, ø 2/10 = 20 %
tɕ tɕ tɕ tɕ
tɕh tɕh, tɕ, tsh tɕ: 2 %, tsh: 2 % tɕh, tɕ, ø tɕh: 81.82 %, tɕ: 9.09 %, ø: tɕh, tɕh: 4/5 = 80 %, tɕ: 1/5 = 20 %
9.09 % tɕ
ɕ ɕ, tɕ, tɕh tɕ: 52 %, tɕh: 9 % ɕ, tɕ ɕ: 40 %, tɕ: 60 % ɕ
tʂ ts 100 % ts 100 % ts 100 %
tʂh tsh 100 % tsh 100 % tsh 100 %
ʂ s, ts s: 96 %, ts: 4 % s 100 % ʂ, s, ʂ:1/17 = 5.89 %, s:15/17 = 88.22 %, ts:
ts 1/17=5.89 %
ʐ ʐ s 100 % ʂ: 1/17 = 5.89 %, s: 15/17 = 88.22 %
ts: 1/17 = 5.89 %
ʐ ʐ ʐ l 1/1 = 100 %
ts ts ts, tsh ts: 95 %, tsh: 5 % ts
(continued)
2 Phonological Development
Table 2.13 (continued)
tsh tsh tsh tsh
s s, ts ts: 33 % s, ts s: 75.86 %, ts: 24.14 % s
ŋ ŋ ŋ ŋ
i i i, ø i: 96.77 %, ø: 3.23 % i
u u u, ø, y u: 81.82 %, ø: 13.64 %, y: u
4.55 %
y y, i i: 50 % y, i y: 90.91 %, i: 9.09 % y
a a a, ai a: 80 %, ai: 20 % a
o o o o
ə ə ə ə
ei ei ei ei
ai ai, a a: 5 % ai ai
e e e e
au au au au
ou ou ou ou
an an, a, aŋ, a: 7 %, aŋ: 7 %, uan: an an
uan 2%
ən ən ən ən
aŋ aŋ aŋ, a, an, aŋ: 80 %, a: 4 %, an: 4 %, en: aŋ, aŋ: 10/12 = 83.33 %, an: 2/12 = 16.67 %
2.5 Mandarin Phonological Development by Children of Indonesian Mothers
en 12 % an
əŋ əŋ əŋ əŋ
ə˞ ə˞ ə 100 % ə 100 %
ia ia ia ia
(continued)
39
40
ua ua ua ua
ͻŋ ͻŋ ͻŋ ͻŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in, i in: 5/6 = 83.33 %, i: 1/6 = 16.67 %
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ ua ua: 1/1 = 100 %
iŋ iŋ iŋ, in iŋ: 85.71 %, in: 14.29 % iŋ
jͻŋ jͻŋ jͻŋ jͻŋ
uo uo uo uo, o uo: 6/7 = 85.71 %, o: 1/7 = 14.29 %
Tableau 2.2 OT tableau for the Indonesian girl: 2, 2 + 1, and 2 + 2 years of age
ʂ Ident (astrident) *Retroflex *Affricate *(+cont) *(+anterior)
☞s * *
ʂ *! *
ts *! *
tʂ *! *
2.5.2 The Three-Year-Old Indonesian Boy
The three-year-old boy of an Indonesian mother has problems with aspiration,

fricatives, laterals, nasals, retroflexes, and complex vowels. He sometimes aspirated
and unaspirated stops and affricates as /píŋ/ ‘sick’ was pronounced as [phíŋ] and /tɕi
tɕhíʐə̀n/ ‘robot’ was pronounced as [tɕhi tɕhí lə̀n]. He pronounced the labiodental
fricative /f/ as aspirated or unaspirated voiceless bilabial stops as /fē i tɕi/ ‘plane’
was pronounced as [péi tɕi] and /fáŋ/ ‘put’ was pronounced as [pháŋ]. The alveolar
lateral was used interchangeably with the alveolar nasal. For example, /pú nə̀ŋ/
‘can’t’ was pronounced as [pú lə̀ŋ] and /luán tɕiǎ ŋ xúa/ ‘speak nonsense’ was
pronounced as [núan tɕiǎ ŋ uá]. The velar fricative was omitted as in the above
example, and also, /xóu mién/ ‘back’ was pronounced as [óu mén]. The alveopalatal
fricative was replaced by the palatal one, as /tɔ̄ ŋ ɕi/ ‘thing’ was pronounced as [tɔ̄ ŋ
ɕi]. The alveolar affricate was replaced by the alveopalatal one, for example, /tś tɕi/
‘self’ as [tɕí tɕí]. The retroflex obstruents were replaced by their alveolar coun-
terparts, for example, /tʂə́/ ‘this’ as [tsə́], /tʂhə̄ ts/ ‘car’ as [tshə̄ ts], /ʂ́/ ‘is’ as [ś], and /
ʐə̀n/ ‘person’ as [1ə̀n]. The high front rounded vowel was pronounced without lip
rounding, for example, /uàn tɕý/ ‘toy’ as [uàn tɕí]. The child had a constraint against
vowel sequences, so complex vowels were simplified, for example, /thiē tš̩ / ‘sticker’
as [thē tš̩ ], /tuán tiáu/ ‘break’ as [tán tiáu], /tɕhì kuái/ ‘strange’ as [tɕhì kuá], and /
phǎ u tsǒ u/ ‘run away’ as [phǎ tsǒ u]. He seemed to do better in the vowels in syllable
final. However, he tended to drop the final nasal; for example, /ɕ̌ iau ɕīn/ ‘careful’
was pronounced as [ɕiǎ u ɕī] and /uàn/ ‘play’ was pronounced as [uà].
Some of errors were also observed in the mother. The mother also replaced
retroflex obstruents with alveolars just like most Mandarin-speaking adults in
Taiwan. She simplified vowel sequence but not as productive as the child. She only
omitted /i/ in /ie/ and /iou/ sequence, for example, /thiē / ‘paste’ as [thē ] and /tɕióu xú
tʂhə̄ / ‘ambulance’ as [tɕíu xú tʂhə̄ ]. She was also observed to change alveolar
affricate to alveopalatal, for example, /tś tɕǐ / ‘self’ as [tɕí tɕǐ ] like her child. She did
not drop final nasals like her child, but changed the place of articulation, for
example, /pú ɕiŋ̀/ ‘may not’ as [pú ɕiǹ].
In the second year, the voiceless labiodental fricative [f] has been acquired and
pronounced accurately. However, he continued replacing retroflex obstruents with
alveolar obstruents, i.e., /ʂuě i kuǒ / ‘fruit’ ! [suě i kuǒ ]. There were also variants of
stops. For example, the unaspirated voiceless bilabial stop /p/ was sometimes
replaced as the aspirated one [ph] as in /tɕhiaŋ̀ pi/́ ‘wall,’ or the alveolar one [t] as in
/pǎ n tʂ/ ‘board.’ The aspirated voiceless velar stop /kh/ was sometimes replaced
with the alveolar one [th], e.g., /khǎ phień / ‘card’ ! [thǎ phień ]. Some errors in the
vowels occurred in the second year. A coda [n] was added to the vowel /a/, e.g., /tǎ
tɕia/́ ‘fight’ ! [tǎ tɕian]́ or replaced the velar nasal coda, i.e., /tʂhuaŋ̀ /
‘bed’ ! [tʂhuan ̀ ].
In the third year, the child was already six years of age. He had acquired most
phonemes, but still had problems with retroflexes and pronounced them as alveolars
just like most adult Mandarin speakers in Taiwan. For example, retroflex affricate /
tʂ/ was still replaced by alveolar affricate [ts] as in /tʂə́/ ‘this’; retroflex obstruent /ʂ/
was still replaced by alveolar obstruent [s] as in /ʂí/ ‘be.’ The other problem is
replacing the vowel /u/ with [ou], e.g., /phìŋ kuǒ / ‘apple’ ! [phìŋ koǔ ]. The
phonological development of the three-year-old boy of an Indonesian mother is
summarized in Table 2.14.
Let me illustrate the phonological development of the three-year-old Indonesian
boy in the OT framework. The three-year-old Indonesian boy ranked the marked-
ness constraints *(+s.g), *(+cont), and *Stop over the faithfulness constraint, so the
labiodental fricative became the unaspirated voiceless bilabial stop as shown in
Tableau 2.3. In the second and the third year, *(+s.g) and *(+cont) were demoted
and thus ended with the correct target [f] as in Tableau 2.4.
2.5.3 The Three-Year-Old Indonesian Girl
The three-year-old girl of an Indonesian mother has acquired all the stops and had
mistakes only occasionally. She had some typical Taiwan Mandarin pronunciation.
She replaced labiodental fricative /f/ with the velar one [x], for example, /fɔ́ŋ lì/
‘pineapple’ as [xɔ́ŋ lì]. She also replaced retroflexes with alveolar counterparts, for
example, /tʂə́ kə/ ‘this one’ as [tsə́ kə], /tʂ̄ h/ ‘eat’ as [ts̄ h], /ʂ́/ ‘is’ as [ś], and /ʐə̀n/
‘person’ as [lə̀n]. The child has acquired all the vowels except sometimes they
replaced the rounded high front vowel /y/ with the unrounded one [i]. For example,
/ý mǐ / ‘corn’ was pronounced as [í mǐ ].
Table 2.14 Acquisition and substitutions of the sounds from the Indonesian boy: 3, 3 + 1, and 3 + 2 years of age
Phoneme Phonetic representations, first year Phonetic representations, second year Phonetic representations,
third year
p p, ph p (93.59 %), ph (6.41 %) p p
ph ph ph ph, m
m once
m m m m
f f, p, ph, x f (21.74 %), p (69.57 %), ph (4.35 %), f f
x (4.35 %)
t t, k, l t (87.18 %), k (1.71 %), l (11.11 %) t t
th th, x, ø th (70 %), x (1.15 %), ø (1.12 %) th th
n n, l n (94.95 %), l (5.05 %) m n
l l, n, k, ø l (97.21 %), n (0.56 %), k (1.12 %), ø l l
(1.12 %)
k k, kh k (99.04 %), kh (0.96 %) k k
h h
k k kh kh
x x, ø x (96.94 %), ø (3.06 %) x x
tɕ tɕ, tɕh, ts, th tɕ (77.27 %), tɕh (19.09 %), ts (2.73 %), th tɕ, ø, t tɕ (81.82 %), ø (4.54 %), t (13.64 %) tɕ
(0.91 %)
tɕh tɕh tɕh tɕh
h h
ɕ ɕ, ɕ, s, k , tɕ ɕ (88.68 %), ɕ (1.89 %), s (1.89 %), k ɕ, tɕ ɕ (81.82 %), tɕ (18.18 %) ɕ
(1.89 %), tɕ (5.66 %)
tʂ ts, tɕ ts (99.34 %), tɕ (0.66 %) tʂ, ts, tɕ, tʂ (4.08 %), ts (85.72 %), tɕ (4.08 %), th ts 100 %
th , t (4.08 %), t (2.04 %)
tʂh tsh, tɕh tsh (97.18 %) tʂh, tsh tʂh (9.09 %), tsh (90.91 %) tsh 100 %
tɕh (2.82 %)
ʂ s, ɕ s (97.26 %), ɕ (2.74 %) s 100 % s, ɕ
43
(continued)
44
third year
s (97.83 %), ɕ
(2.17 %)
ʐ l 100 % ʐ l 100 %
ts ts, tɕ ts (4.74 %), tɕ (5.26 %) ts, tʂ, k ts (87.5 %), tʂ (8.33 %), k (4.17 %) ts
tsh tsh tsh tsh, tsh (92.59 %), ø
ø (7.41 %)
s s s s
ŋ ŋ ŋ ŋ
i i i i
u u u u
y y, i y (59.1 %), i (40.91 %) y y
a a, ia, au a (96.97 %), ia (1.82 %), au (1.21 %) a a
o o o o
ə ə, i, ei, o ə (99.25 %), i (0.30 %), ei (0.15 %), o ə ə
(0.30 %)
ei ei, ie ei (93.65 %), ie (6.35 %) ei ei
ai ai, a, ia ai (68.18 %), a (29.09 %), ia (2.73 %) ai, a ai (66.67 %), a (33.33 %) ai
e e e e
au au, a, iau au (71.95 %), a (26.83 %), iau (1.22 %) au au
ou ou ou ou
an an, a an (90 %), a (10 %) an an, uan
uan once
(continued)
third year
ən ən ən, in ən (77.78 %), in (22.22 %) ən
aŋ aŋ aŋ aŋ
əŋ əŋ, an əŋ (93.33 %), an (6.67 %) ən 100 % əŋ
ə˞ ə˞ ə 100 % ə 100 %
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie, e, i, ei, iau ie (24 %), e (40 %), i (12 %), ei (20 %), ie, e ie (56.25 %), e (43.75 %) ie
iau (4 %)
ye ye ye ye
uai uai, ua uai (88.64 %), ua (11.36 %) uai uai
uei uei, u uei (97.87 %), u (2.13 %) uei uei
iau iau, ia iau (86.13 %), ia (13.87 %) iau iau
iou iou, ə, ou, iͻ iou (90.82 %), ə (2.04 %), ou (3.06 %), iou iou
iͻ (4.08 %)
ien ien, en, iŋ, ei, ien (65.08 %), en (25.40 %), iŋ (3.17 %), ien, en ien (73.68 %) ien
i ei (4.76 %), i (1.59 %) en (26.32 %)
uan uan, an, ua, uan (73.44 %), an (1.56 %), ua (12.5 %), uan uan
uaŋ, ͻŋ uaŋ (4.69 %), ͻŋ (7.81 %)

yen yen yen yen
in in, i in (92 %), i (8 %) in in
uən uən uən uən
yn yn yn yn
(continued)
45
46
third year
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ, ͻŋ uaŋ (42.86 %), ͻŋ (57.14 %) uaŋ uaŋ, uaŋ (86.67 %), ͻŋ
ͻŋ (13.33 %)
iŋ iŋ iŋ, in iŋ (28.57 %), in (71.43 %) iŋ
jɔŋ jɔŋ jɔŋ jɔŋ
uo uo, o, ou uo (71.31 %), o (26.23 %), ou (2.46 %) uo, o, u uo (80 %), o (14.29 %), u (5.71 %) uo, o uo (98.91 %), o
(1.09 %)
Tableau 2.3 OT tableau for the Indonesian boy: 3 years of age

/f/ Ident (aanterior) *(+s.g) *(+cont) *Stop Ident (astrident)
f *! *
☞p * *
ph *! * *
x *! * *
Tableau 2.4 OT tableau for the Indonesian boy: 3 + 1 and 3 + 2 years of age
/f/ Ident (aanterior) *Stop *(+s.g) *(+cont) Ident (astrident)
☞f * *
p *! *
ph *! * *
x *! * *
In the second year, the phonetic variations decreased in both types and tokens. The
replacement of labiodental fricative /f/ with the velar one [x] has disappeared. The
replacement of retroflexes with alveolars still happens occasionally. The alveopa-
latal fricative was still pronounced as affricates most of the time. The alveolar
fricative was also pronounced as affricates sometimes. The vowel /u/ was dropped
occasionally. She still replaced the rounded high front vowel /y/ with the unrounded
one [i] occasionally. /a/ was replaced with [ai] occasionally. She dropped nasals in
codas or changed the place of articulation occasionally, which was not observed in
the first year.
In the third year, the only problems remained were retroflexes and nasals in coda.
Retroflexes were replaced with alveolars most of the time. The velar nasals in codas
were still replaced with alveolar nasals when following /i/ but not other vowels
sometimes. The frequency is lower than that in the second year. The phonetic
representations of the three-year-old girl are summarized in Table 2.15.
48
Phoneme Phonetic representations, first year Phonetic representations, second Phonetic representations, third year
year
p p p p
ph ph, x, f, th x: 8 %, f: 8 %, th: ph, p ph: 77.78 %, p: ph
8% 22.22 %
m m m, l m: 87.5 %, l: 12.5 % m
f f, x x: 33 % f f
t t, p, k, ts p: 4 %, k: 2 %, ts: t t
2%
th th th th
n n m n
l l l, n l: 98 %, n: 2 % l
k k k k, ø ø: once
once
kh kh kh kh
x x x x
tɕ tɕ, t t: 2 % tɕ tɕ
tɕh tɕh, ɕ ɕ: 6 % tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ tʂ, ts ts: 94 % ts, tɕ ts: 93.33 %, tɕ: 6.67 % ts 53/53 = 100 %
tʂh tʂh, tsh, kh tsh: 81 %, kh: 2 % tsh 100 % tsh 8/8 = 100 %
ʂ ʂ, s, tsh, kh, x, k, s: 66 %, tsh: 2 %, kh: ʂ, s ʂ: 25 %, s: 75 % s 43/43 = 100 %
th 18 %, x: 5 %, k:
5 %,
th: 2 %
ʐ l 100 % ʐ ʐ, l ʐ: 2/4 = 50 %, l: 2/4 = 50 %
(continued)
year
ts ts, k, tʂ k: 10 %, tʂ: 10 % ts, tʂ ts: 90.91 %, tʂ: 9.09 % ts, tʂ, tɕ tʂ once
tɕ once
tsh tsh, kh kh: 6 % tsh tsh
s s, ʂ, kh ʂ: 9 %, kh: 27 % s, ʂ s: 97.92 %, ʂ: 2.08 % s, ʂ ʂ twice
ŋ ŋ ŋ ŋ
i i i i
u u u, ø u: 94.34 %, ø: 5.66 % u
y y, i i: 48 % y y
a a, au au: 3 % a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au, iau iau twice
ou ou ou ou
an an, ən, uan ən: 10 %, uan: 15 % an an
ən ən ien 100 % ən
aŋ aŋ, an an: 9 % aŋ, aŋ: 81.25 %, an: aŋ

an 18.75 %
əŋ əŋ əŋ əŋ
ə˞ ə˞ ə 100 % ə˞
ia ia ia ia
49
(continued)
50
year
ua ua ua ua
ͻŋ ͻŋ ͻŋ ͻŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ iŋ, in iŋ: 69.23 %, in: iŋ, in, iŋ: 5/8 = 62.5 %, in: 2/8 = 25 %, i:
30.77 % in 1/8 = 12.5 %
uo uo uo uo
2.5.4 The Four-Year-Old Indonesian Boy
The Indonesian four-year-old boy had problems with stops, fricatives, laterals,
nasals, alveopalatals, and retroflexes. The child replaced the unaspirated voiceless
bilabial stop /p/ with the voiceless labiodental fricative [f] as /pań fǎ / ‘solution’ was
pronounced as [fan ́ fǎ ]. On the other hand, he also changed fricatives to stops. For
example, /xē i sə/́ ‘black’ was pronounced as [xē i tə]́ and /sɔ̄ ŋ ʂǔ / ‘squirrel’ was
pronounced as [tɔ̄ ŋ tǔ ]. The aspirated voiceless alveolar stop /th/ was replaced by
alveolar affricates [tsh]. For example, /thú/ ‘throw up’ was pronounced as [tshu]́ . The
alveolar lateral /l/ and alveolar nasal /n/ were used interchangeably as /naǹ ʂə̄ ŋ/
‘male’ was pronounced as [laŋ̀ tə̄ ŋ] and /laǹ tɕhioù / ‘basketball’ was pronounced as
h
[naǹ tɕ iou ̀ ]. Alveolar nasals and velar nasals in codas were mixed. For example, /
kuan ́ tɕȳ n/ ‘champion’ was pronounced as [kuaŋ́ tɕȳ n] and /tshəŋ̀ kɔ̄ ŋ/ ‘succeed’
was pronounced as [tʂhəǹ kɔ̄ ŋ]. The alveopalatal fricative /ɕ/ was changed to the
palatal fricative [ҫ] or alveolar fricative [s]. For example, /ɕiǎ ŋ taú / ‘think of’ was
pronounced as [ҫiǎ ŋ taú ] and /tʂhɔŋ̀ ɕī n/ ‘restart’ was pronounced as [tshɔŋ̀ sī n]. The
retroflexes were replaced by their alveolar counterparts. For example, /khǎ tʂu/́
‘stuck’ was pronounced as [khǎ tsu]́ , /tʂhū an/ ‘wear’ was pronounced as [tshū an],
and /ʂā y/̀ ‘shark’ was pronounced as [sā y]̀ . The voiced retroflex fricative was
replaced by the alveolar lateral [l] all the time as /ʐəǹ tɕiā / ‘I’ was pronounced as
̀ tɕiā ]. The child sometimes changed alveolar affricates to retroflex or alveo-
[lən
palatal ones. For example, /thú ts̝/ ‘rabbit’ was pronounced as [thú tʂ] and /ts̩ /́
‘rabbit’ was pronounced as [tɕhi]́ . The child did not really have problems with any
manner or place of articulation, but was confused about when to use which one. All
the vowels were correct. His mother also replaced retroflexes and final velar nasals
with alveolars as she pronounced /tʂǎ u/ ‘find’ as [tsǎ u] and /nań ʂə̄ ŋ/ ‘male’ as [naǹ
sən]. The mother’s errors were more systematic.
In the second year, the child produced the labiodental fricative /f/ correctly, but still
have problems with the voiceless retroflex fricative, which was replaced with an
alveolar one [s], e.g., /ʂū / ‘book’ ! [sū ]. The other problem was the dropping of
velar nasal coda, i.e., /ɔŋ/ ! [ɔ] as in /xɔŋ̀ lɔ̀ puo/ ‘carrot.’ Moreover, the child still
replaced retroflex affricate /tʂ/ with alveolar affricate /ts/ as in /tʂuō tʂ/ ‘table.’
In the third year, the child was already seven years of age. He had acquired all the
phonemes and pronounced them almost correctly. But, sometimes, the child may
replace the unaspirated voiceless alveolar stop /t/ with the aspirated one [th], e.g.,
/tɔŋ́ u/́ ‘animal’ ! [thɔŋ́ u]́ . Besides, the retroflex affricate /tʂ/ was sometimes
replaced with alveolar affricate /ts/, i.e., /tʂə́ kə/ ‘this’ ! [tsə́ kə]. The phonological
development of the four-year-old boy of an Indonesian mother is summarized in
Table 2.16.
Let me illustrate with an example of the four-year-old boy’s phonological
development in the OT framework. *(+cont) was high-ranked in the first year, so
the voiceless bilabial stop was changed to voiceless labiodental fricative as in
Tableau 2.5.
In the second and the third year, *(+cont) was demoted and ended with the target
[p] as in Tableau 2.6.
2.5.5 The Four-Year-Old Indonesian Girl
The Indonesian four-year-old girl has acquired most of the phonemes and just had
problems with retroflexes and high front rounded vowels like many
Mandarin-speaking adults in Taiwan. She replaced retroflexes with alveolars. For
example, /tʂǔ / ‘cook’ was pronounced as [tsǔ ]. /tʂhū / ‘out’ was pronounced as
[tshū ]. /ʂ̄ / ‘teacher’ was pronounced as [s̩̄ ]. /ʐàn/ was pronounced as [làn]. The high
front rounded vowel was pronounced without lip rounding, as /tɕhý/ ‘go’ was
pronounced as [tɕhí] occasionally.
year
p p, f p: 2/6 = 33.3 %, f: 4/6 = 66.6 % p p
ph ph ph ph
m m m, l m: 5/6 = 83.33 %, m
l: 1/6 = 16.67 %
f f f f, p p: once
t t t t
th th, tsh, th: 5/12 = 41.6 %, tsh: 6/12 = 50 %, tʃ: 1/12 = 8.3 % th, th: 6/7 = 85.71 %, th
tʃ tsh tsh: 1/7 = 14.29 %
n n, l n: ½ = 50 % m n
l: 1/2 = 50 %
l n, l l: 11/13 = 84.6 % l l
n: 2/13 = 15.3 %
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ, k k twice
tɕh tɕh tɕh tɕh
ɕ ɕ, s, ts ɕ: 4/8 = 50 %, s: 3/8 = 37.5 %, ts: 1/8 = 12.5 % ɕ ɕ, k k once
tʂ ts, t ts: 14/17 = 82.3 %, t: 3/17 = 17.7 % tʂ, tʂ: 1/4 = 25 %, t: tʂ, tʂ: 4/144 = 2.78 %, ts:
t, ts 2/4 = 50 %, ts: 1/4 = 25 % ts, 139/144 = 96.53 %, tɕ:

tɕ 1/144 = 0.69 %
tʂh tʂh, tsh tʂh: 4/23 = 17.3 %, tsh: 19/23 = 82.6 % tʂh, tʂh: ¾ = 75 %, t+: tʂh, tʂh: 3/30 = 10 %
t, h 1/4 = 25 % tsh tsh: 27/30 = 90 %
(continued)
53
54
year
ʂ s, ts, s: 3/31 = 9.6 %, ts: 1/31 = 3.2 %, tʂh: 23/31 = 74.1 %, ʂ, s ʂ: 3/4 = 75 %, s: ʂ, s s: 7/78 = 8.97 %, s: 71/78 = 91.03 %
tʂh, t, ɕ t: 3/31 = 9.6 %, ɕ: 1/31 = 3.2 % 1/4 = 25 %
ʐ l 1/1 = 100 % ʐ l, n l: 16/17 = 94.12 %, n: 1/17 = 5.88 %
ts ts, tʂ, t ts: 1/3 = 33.3 %, tʂ: 1/3 = 33.3 %, t: 1/3 = 33.3 % tʂ 7/7 = 100 % ts, l: 35/38 = 92.11 %, n: 3/38 = 7.89 %
tʂ
tsh tɕh, tʂh tɕh: ½ = 50 %, tʂh: 1/2 = 50 % tsh, tsh: 4/5 = 80 %, tʂh: tsh, tsh: 8/9 = 88.89 %, tʂh: 1/9 = 11.11 %
tʂh 1/5 = 20 % tʂh
s s, t, ɕ s: 4/19 = 21.05 %, t: 13/19 = 68.42 %, ɕ: s, ʂ s: 7/12 = 58.33 %, ʂ: s, ʂ s: 28/30 = 93.33 %, ʂ: 2/30 = 6.67 %
2/19 = 10.52 % 5/12 = 41.67 %
ŋ ŋ ŋ ŋ
i i i i
u u u u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an, aŋ an: 3/11 = 27.27 %, aŋ: 8/11 = 72.72 % an an
ən ən ən ən
(continued)
year
aŋ aŋ aŋ aŋ
əŋ əŋ, ən əŋ: 3/6 = 50 %, ən: 3/6 = 50 % əŋ əŋ, əŋ: 2/7 = 28.57 %, ən: 5/7 = 71.43 %
ən
ə˞ ə˞ ə˞ ə 19/19 = 100 %
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
(continued)
55
56
year
iŋ in 3/3 = 100 % iŋ iŋ
uo uo uo uo
Tableau 2.5 OT tableau for p Ident *Stop * *

the Indonesian boy: 4 years of (aanterior) (+cont) (+strident)
age
p *!
☞f * *
Tableau 2.6 OT tableau for p Ident * * *Stop

the Indonesian boy: 4 + 1 and (aanterior) (+cont) (+strident)
4 + 2 years of age
☞p *
f *! *
Her mother also replaced retroflexes with alveolars and pronounced the high
front rounded vowel without lip rounding. In addition, she replaced alveolar nasals
with alveolar laterals, changed the place of articulation of final nasals, simplified
complex vowels, and pronounced the first tone as the fourth tone. For example, the
mother pronounced /nień / ‘read’ as [lień ]. She changed the final alveolar nasal to a
velar one, as /uən ́ / ‘ask’ was pronounced as [uəŋ́ ]. She tended to drop the first vowel
in a sequence; for example, /tɕhień mień / ‘front’ was pronounced as [tɕhieǹ meń ]
and /xǔ o tɕien
́ / ‘rocket’ was pronounced as [xǒ tɕień ]. She pronounced the first tone
as the fourth tone as /xiā ŋ kū / ‘mushroom’ was pronounced as [xiaŋ́ kū ]. These
errors were not observed in the child.
In the second year, the girl still replaced retroflexes with alveolars all the time. In
addition, she had problems with place of articulation of nasals in codas. She tended
to replace velar nasals with alveolar nasals next to /i/, which could be caused by
assimilation. Nasals following other vowels were pronounced correctly.
In the third year, retroflexes were still replaced with alveolars over 90 % of the
time. Velar nasals were replaced with alveolar nasals next to /i/ with even higher
frequency than in the second year. The acquisition and substitutions of the sounds
of the four-year-old girl are summarized in Table 2.17.
58
year
p p p p
ph ph ph ph
m m m m
f f f f, xw f: 13/16 = 81.25 %, xw: 3/16 = 18.75 %
t t t t
th th th th
n n, m n: 98.24 %, m: 1.75 % m n
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ ts 100 % ts 100 % tʂ, ts tʂ: 8/134 = 5.97 %, ts: 126/134 = 94.03 %
tʂh tʂh 100 % tsh 100 % tʂh, tʂh: 2/31 = 6.45 %, tsh: 29/31 = 93.55 %
tsh
ʂ s 100 % s 100 % ʂ, s ʂ: 22/341 = 6.45 %, s: 319/341 = 93.55 %
ʐ l 100 % ʐ ʐ, l ʐ: 24/52 = 46.15 %, l: 28/52 = 53.85 %
ts ts ts ts
tsh tsh tsh tsh
s s s s
(continued)
year
ŋ ŋ ŋ ŋ
i i i i
u u u u
y y, i y: 93.75 %, i: 6.25 % y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au, aŋ au: 98.33 %, aŋ: 1.66 % au au
ou ou ou ou
an an, aŋ an: 96.77 %, aŋ: 3.22 % an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ əŋ əŋ
ə˞ ə 100 % ə˞ ə˞, ə ə˞: 3/24 = 12.5 %, ə: 21/24 = 87.5 %
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ, au, ɔŋ: 88.88 %, au: 5.55 %, ͻ: ɔŋ ɔŋ
5.55 %
(continued)
59
60
year
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau, iau: 96.87 %, tɕiau: 3.12 % iau iau
tɕiau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ iŋ, in iŋ: 57.14 %, in: 42.86 % iŋ, in iŋ: 14/50 = 28 %, in: 36/50 = 72 %
uo uo, o uo: 99.31 %, o: 0.68 % uo uo
2.5.6 The Five-Year-Old Indonesian Boy
The Indonesian five-year-old boy had problems with labiodental fricatives, retro-
flexes, final nasals, and complex vowels. He replaced the labiodental fricative /f/
with [xu]; for example, [fē i tɕī ] ‘airplane’ was pronounced as [xuē i tɕī ]. He replaced
all the retroflexes with alveolars as /tʂə́ iaŋ́ / ‘this way’ as [tsə́ iaŋ́ ], /tʂ̄ h ‘eat’ as [ts̄ h],
/ʂ/́ ‘is’ as [s̩ ],
́ and /ʐəǹ / ‘person’ as [ləǹ ]. One of the vowels in a sequence was
omitted sometimes. For example, /kǒ u/ ‘dog’ was pronounced as [kǒ ], /mień /
‘noodles’ as [men ́ ], /tʂ ū aŋ/ ‘bed’ as [ts ā ŋ], and /uǒ / ‘I’ as [ǒ ]. The velar nasal /ŋ
h h
following /i/ was changed to the alveolar nasal [n] as /tɕhī ŋ/ ‘clean’ was pronounced
as [tɕhī n]. Many of these errors were also observed in the mother (Table 2.21).
His mother also replaced retroflexes with alveolars, changed the place of artic-
ulation of final nasals, pronounced the high front rounded vowel without lip
rounding, and simplified complex vowels. In addition, she replaced the alveolar
nasal with the alveolar lateral, the first tone with the fourth tone. For example, the
mother pronounced /nień / ‘read’ as [lień ]. She changed the final alveolar nasal to a
velar one, as /uən ́ / ‘ask’ was pronounced as [uəŋ́ ]. She tended to drop the first vowel
in a sequence; for example, /tɕhieǹ mié n/ ‘front’ was pronounced as [tɕhieǹ meń ]
and /xǔ o tɕien ́ / ‘rocket’ as [xǒ tɕień ]. She pronounced the first tone as the fourth
tone as /xiaŋ́ kū / ‘mushroom’ was pronounced as [xiaŋ́ kū ] like the stress in a stress
language like her L1, Indonesian. The tone errors were not observed in the child.
In the second year, the alveolar variants [s, ts, tsh] of retroflex obstruents /ʂ, t ʂ, tʂh/
coexisted with retroflexes [ʂ, tʂ, tʂh] and alveopalatals [ɕ, ʨ] except the voiceless
aspirated retroflex affricate /tʂh/ and the voiced retroflex fricative /ʐ/ were consis-
tently replaced by the alveolars [tsh] and [z].
In the third year, almost all the phonemes were acquired except that the aspirated
voiceless retroflex affricate /tʂh/ was still replaced by the unaspirated one [tʂ], e.g., /
tʂhī / ‘eat’ ![tʂī ]. The phonological development of the five-year-old boy of an
Indonesian mother is summarized in Table 2.18.
62
Phoneme Phonetic representations, first year Phonetic representations, Phonetic representations, third year
second year
p p p
ph ph ph ph
m m m m
f f, x f: 91.66 % f f
x: 8.33 %
t t t t
th th th th
n n m n
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ, tɕ ɕ: 96.77 %, tɕ: 3.23 %
tʂ tʂ 100 % tʂ, ts tʂ: 25 %, ts: 75 % tʂ, ts tʂ: 26.53 %, tʂ: 73.47 %
tʂh tʂh 100 % tsh 100 % tʂh, tsh tʂh: 25.81 %, tsh: 74.19 %
ʂ s 100 % s 100 % ʂ, s ʂ: 16.38 %, s: 83.62 %
ʐ l 100 % ʐ ʐ, l ʐ: 36.36 %, l: 63.64 %
ts ts ts ts
tsh tsh tsh tsh
s s s s
(continued)
second year
ŋ ŋ ŋ ŋ
i i i i
u u, ə u: 96.96 %, ə: 3.03 % u u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au, a au: 75 %, a: 25 % au
ou ou, o ou: 93.54 %, o: 6.45 % ou ou
an an an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ əŋ əŋ
ə˞ ə 100 % ə˞ ə˞, ə ə˞: 5.26 %, ə: 94.74 %
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
ye ye ye ye
(continued)
63
64
second year
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien, en ien: 97.43 %, en: 2.56 % ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ, aŋ uaŋ: 83.33 %, aŋ: 16.66 % uaŋ uaŋ
iŋ iŋ 100 % iŋ iŋ, in iŋ: 27.5 %, in: 72.5 %
uo uo, o uo: 95.23 %, o: 4.76 % uo uo
2.5.7 The Five-Year-Old Indonesian Girl
The Indonesian five-year-old girl had problems with alveolar laterals, retroflexes,
final nasals, and high front rounded vowels. She changed the alveolar lateral /l/ to
nasal [n]; for example, /lə̌ ŋ/ ‘cold’ was pronounced as [nə̌ ŋ]. All the retroflexes
were replaced by their alveolar counterparts. For example, she pronounced / kɔ̄ ŋ ʂǔ /
’ princess’ was pronounced as [kɔ̄ ŋ tsǔ ], [ʂhū laì ] ‘come out’ as [tshū laì ], and /ʂueí
́ / ‘sleep’ as [sueí tɕiaú ]. There were also aspiration and palatal assimilation. She
tɕiau
pronounced the unaspirated voiceless alveopalatal affricate as the aspirated one as /
́ tɕ y/́ ‘enter’ was pronounced as [tɕ iń tɕ y]́ . The final nasals tended to be
h h h
tɕin
dropped; for example, the child pronounced /ɕiǎ u/ ‘small’ as [ɕiǎ ], /uaŋ́ tɕi/́ ‘forget’
as [uá tɕi]́ , /phaŋ́ / ‘fat’ as [pha]́ , /khaú tɕiń / ‘get close’ as [khaú tɕi]́ , and /pì kɔ̌ ŋ/ ‘nose
hole’ as [pì kɔ̌ ]. The high front rounded vowels were sometimes pronounced
without lip rounding; for example, /tɕhy/́ ‘go’ was pronounced as [tɕi]́ .
In the second year, the alveolar lateral /l/ was pronounced correctly. The retroflexes
were still replaced by alveolars with lower frequency. She had no more problems
with aspiration and palatal assimilation. The velar final nasals were sometimes
changed to alveolar nasals.
In the third year, the girl has acquired all the phonemes except that retroflexes were
consistently replaced by alveolars. The retroflex mid-central lax vowel was also
pronounced without retroflex (Table 2.19).
Let us put /tʂ/ in the OT framework. The five-year-old Indonesian girl
high-ranked *Retroflex and so turned it to an alveolar affricate and remained for
three years as shown in Tableau 2.7.
66
p p p p
ph ph ph ph
m m m m
f f f f
t t t t
th th th th
n n n n
l l, n l: 81.82 %, n: 18.18 % l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ, tɕh tɕ: 77.27 %, tɕh: 22.73 % tɕ tɕ, ts tɕ: 70.83 %, ts: 29.17 %
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ ts 100 % tʂ, ts tʂ: 13.64 %, ts: 86.36 % ts 100 %
tʂh tsh 100 % tsh 100 % tsh 100 %
ʂ s 100 % s 100 % s 100 %
ʐ l 100 % ʐ l 100 %
ts ts ts ts
tsh tsh tsh tsh
s s s s
ŋ ŋ ŋ ŋ
i i i i
(continued)
u u u u
y y, i y: 88.89 %, i: 11.11 % y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an an, a an: 95 %, a: 5 % an
ən ən ən ən ən
aŋ aŋ, an, a aŋ: 36.36 %, an: 45.45 %, a: 18.18 % aŋ aŋ
əŋ əŋ 100 % əŋ əŋ
ə˞ ə˞ ə 100 % ə 100 %
ia ia ia ia
ua ua ua ua
ie ie ie ie
ye ye ye ye
uai uai uai uai

uei uei uei uei
iau iau iau iau
iou iou iou iou, ͻ iou: 91.86 %, ͻ: 8.14 %
(continued)
67
68
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ, in iŋ: 20 %, in: 80 % iŋ, in iŋ: 54.55 %, in: 45.45 % iŋ
ͻŋ ͻŋ, ͻ ͻ once ͻŋ ͻŋ
uo uo uo uo
Tableau 2.7 OT tableau for tʂ *(+cont) *Retroflex *(+anterior)

the Indonesian girl: 5, 5 + 1,
and 5 + 2 years of age ☞ ts *
tʂ *!
2.5.8 The Six-Year-Old Indonesian Boy
The six-year-old Indonesian boy still had phonetic errors in aspirations, fricatives,
retroflexes, and final nasals. He occasionally deaspirated the aspirated velar stops
and alveopalatal affricates. He replaced the aspirated voiceless velar stop with the
unaspirated one; for example, /khǎ ū / ‘test’ was pronounced as [kǎ ū ] and /tɕhio iə̌ n/
‘earthworm’ was pronounced as [tɕī o iə̌ n]. He replaced the alveopalatal fricative
with an affricate or an alveolar stop. For example, /ɕī pī an/ ‘riverbank’ was pro-
nounced as [tɕī pī an], /ɕíe ɕíe/ ‘thank’ as [tɕhíe tɕhíe], and /ɕǐ au kǒ / ‘dog’ as [tíau
kǒ ]. The retroflexes were replaced by their alveolar counterparts. For example, /tʂ̄
táu/ ‘know’ was pronounced as [ts̩̄ táu], /tʂhàŋ tɕǐ n lú/ ‘giraffe’ as [tshàŋ tɕǐ n lu], /ʂ́
thò/ ‘stone’ as [s̩ ̀ thò], and /ʐə̀n/ ‘person’ as [lə̀n]. He mixed the place of articulation
of final nasals; for example, /sán kā i/ ‘spread’ was pronounced as [sáŋ kā i] and /tháŋ
ʂā ŋ/ ‘burn’ was pronounced as [thaŋ́ ʂaŋ́ ]. The substitution of retroflexes with
alveolars was also observed in his mother, but not other errors.
In the second year, the child continued to replace retroflex affricates with alveolar
affricates and replaced a labiodental fricative with a velar fricative followed by a
labiovelar glide. For example, /f/ was replaced by [xu] as in /fā n tɕhie/̀ ‘tomato.’ The
voiced retroflex fricative was pronounced correctly, but the voiceless one /ʂ/ was
still replaced by an alveolar one [s], e.g., /ʂueǐ kuǒ / ‘fruit’ ! [sueǐ kuǒ ]. In addi-
tion, the child replaced the voiced bilabial nasal stop /m/ with the lateral liquid [l],
e.g., [mień pei] ‘quilt’ ! [lień pei]́ . All the vowels and finals have been acquired
except the rounded high front vowel was still pronounced without lip rounding.
In the third year, when the child was eight years of age, all the phonemes were
pronounced correctly, except the vowel /u/ was dropped in /uo/ sequence, e.g., /phiŋ̀
kuǒ / ‘apple’ ! [phiŋ̀ kǒ ]. It was correct in the first two years. The phonological
development of the six-year-old child of an Indonesian mother is summarized in
Table 2.20.
70
Phoneme Phonetic representations, first year Phonetic representations, second year Phonetic representations, third
year
p p p p
ph ph ph ph
m m m m
f f f f
t t t, th t (88.89 %), th (11.11 %) t
th th th th
n n m n
l l l l
k k k k
kh kh, k k (16 %) kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh, tɕ tɕ (10 %) tɕh tɕh
ɕ ɕ, tɕ, t tɕ (50 %), t (10 %) ɕ ɕ
tʂ ts 100 % ts 100 % ts 100 %
tʂh tsh, s tsh (96 %), s (4 %) tʂh, tsh tsh (80 %) tah 100 %
ʂ s, ts, tsh s (50 %), ts (40 %), tsh (10 %) s 100 % s 100 %
ʐ l 100 % ʐ, l ʐ (33.33), l (66.67 %) l, z l (87.5 %), z (12.5 %)
ts ts ts ts
tsh tsh tsh tsh
s s s s
ŋ ŋ ŋ ŋ
(continued)
year
i i, ei ei (6 %) i, ø, y i (82.61 %), ø (15.22 %), y (2.17 %) i
u u u, ø u (68.18 %), ø (31.82 %) u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an, aŋ aŋ (18 %) an an
ən ən ən ən
aŋ aŋ, an an (7 %) aŋ aŋ
əŋ əŋ əŋ, ən əŋ (66.67 %), ən (33.33 %) əŋ
ə˞ ə˞ ə˞ ə˞
ia ia ia ia
ua ua ua ua
ͻŋ ͻŋ ͻŋ ͻŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
(continued)
71
72
year
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ iŋ, in iŋ (85.71 %), in (14.29 %) iŋ
ͻŋ ͻŋ ͻŋ ͻŋ
uo uo uo uo
2.5.9 The Six-Year-Old Indonesian Girl
2.5.9.1 Phonological development in the first year
The six-year-old Indonesian girl has acquired most phonemes in Mandarin except
alveopalatals, retroflexes, and final nasals. The child sometimes changed alveopa-
latal affricates to alveolar ones; for example, /tɕhyà n pu/́ ‘whole’ was pronounced as
[tshɔŋ̀ pu]́ . She mixed retroflexes and alveolars. She replaced retroflexes with
alveolars most of the time. For example, /tʂ̄ taú / ‘know’ was pronounced as [ts̄ taú ],
/piá n tʂhəŋ̀ / ‘become’ was pronounced as [piá n tshəŋ̀ ], /niá n ʂū / ‘study’ was pro-
nounced as [niá n su], and /ʐəǹ / ‘person’ was pronounced as [ləǹ ]. At other times,
alveolars were replaced by retroflexes. For example, /tś tɕǐ / ‘self’ was pronounced
as [tʂ́ tɕǐ ] and̄ /lan
̀ sə/́ ‘blue’ was pronounced as [laǹ ʂə]́ . She occasionally changed
the place of final nasals or even dropped it. For example, /tɕhyà n pu/́ ‘whole’ was
pronounced as [tshɔŋ̀ pu]́ , /xuɑŋ̀ sə/́ ‘yellow’ was pronounced as [xuà n sə]́ , and /sə̄ n
lin/ ‘forest’ was pronounced as [sə̄ n li]. The child did not seem to have difficulty
with particular place articulation, but was confused about when to use which one.
Only the replacement of retroflexes with alveolars was observed in the mother.
2.5.9.2 Phonological development in the second year
In the second year, the mistakes with low frequency in the first year disappeared but
some new errors were observed. There were substitutions of the bilabial nasal /m/
with the high back vowel /u/ and the substitution of the alveolar lateral /l/ with the
alveolar nasal [n]. The final velar nasal /ö/ was replaced by the alveolar nasal [n]
with higher frequency (30 %). The final alveolar nasal, which was dropped occa-
sionally in the first year, was pronounced correctly most of the time while replaced
by the velar nasal 25 % of the time. The retroflex mid central vowel /Ä/ was
pronounced without retroflexation. She replaced aspirated retroflex affricates and
fricatives with alveolars 100 % of the time. The unaspirated retroflex affricates / tÏ/
co-exist with the alveolar and alveolpalatal variants.
2.5.9.3 Phonological development in the third year
In the third year, the girl has acquired all the phonemes except replacing retroflexes
with alveolars consistently. The aspirated voiceless retroflex affricates /tÏh/ were
pronounced correctly 75 % of the time though. The acquisition and substitutions of
the sounds by the six-year-old Indonesian girl for three years were summarized in
Table 2.21.
The six-year-old Indonesian boy and girl both high-ranked *Retroflex and
changed the retroflex fricative to the alveolar fricative throughout the three years as
shown in Tableau 2.8.
74
Phoneme Phonetic representations, first Phonetic representations, second year Phonetic representations, third
year year
p p, f f (2 %) p p
ph ph ph ph
m m m, u m (87.5 %), u (12.5 %) m
f f f f f
t t, ø ø (2 %) t t
th th th th
n n, l l (5 %) m n
l l l, n l (67.74 %), n (32.26 %) l
k k, t t (8 %) k k
kh kh kh kh
x x, f f (3 %) x x
tɕ tɕ, ɕ ɕ (3 %) tɕ tɕ
tɕh tɕh, tsh tsh (13 %) tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ tʂ, ts ts (63 %) tʂ, ts, tɕ tʂ (8.33 %), ts (79.17 %), tɕ (12.5 %) ts 100 %
tʂh tʂh, tsh tsh (45 %) tsh 100 % tsh, ts tʂh (75 %), ts (25 %)
ʂ ʂ, s s (87 %) s 100 % s 100 %
ʐ ʐ, l l (80 %) n 100 % l 100 %
ts ts, tʂ tʂ (16 %) ts ts
tsh tsh tsh tsh
s s, ʂ ʂ (15 %) s, ʂ, ts s (88.24 %), ʂ (5.88 %), ts (5.88 %) s, ʂ ʂ once
ŋ ŋ ŋ ŋ
(continued)
year year
i i, y, ø y (4 %), ø (2 %) i i
u u u u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai, ei ai (96.3 %), ei (3.7 %) ai
e e e e
au au au, a au (94.44 %), a (5.56 %) au
ou ou ou ou
an an, ͻŋ ͻŋ (4 %) an an
ən ən ən ən
aŋ aŋ, an an (3 %) aŋ, an aŋ (70 %), an (30 %) aŋ
əŋ əŋ əŋ əŋ
ə˞ ə˞ ə 100 % ə˞
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
(continued)
75
76
year year
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in, i i (11 %) in, iŋ in (75 %), iŋ (25 %) in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ iŋ, in iŋ (83.33 %), in (16.67 %) iŋ
uo uo uo uo
Tableau 2.8 OT tableau for the Indonesian children: 6, 6 + 1, and 6 + 2 years of age
ʂ Ident (astrident) *Retroflex *Affricate *(+cont) *(+anterior)
☞s * *
ʂ *! *
2.5.10 The Two-Year-Old Taiwanese Boy
The Taiwanese two-year-old boy has acquired most of the phonemes and just had
mistakes occasionally (under 10 %). The only frequent error is the replacement of
retroflexes with alveolars. For example, he pronounced /tʂáu ɕiáŋ tɕī / ‘camera’ as
[tśau ɕiáŋ tɕi], /tʂhùan/ ‘boat’ as [tshùan], /ʂū / ‘book’ as [sū ], and /tɕī tɕhí ʐə̀n/
‘robot’ as [tɕī tɕhí lə̀n]. He also deretroflexed the retroflex vowel; for example, /ʂ̀ ə́˞/
‘twelve’ was pronounced as [s̩ ̀ ə́]. He dropped final nasals or changed the place of
articulation occasionally. He also simplified vowel sequence occasionally, but it
was not systematic in terms of which one was dropped. For example, /kǔ o tʂ̄ / ‘juice’
was pronounced as [kǒ ts̩̄ ] and /nà kúo lài/ ‘bring’ was pronounced as [nà kú lài].
The mother also replaced retroflexes with alveolars and deretroflexed retroflex
vowels. She also changed the final velar nasal following /i/ to the alveolar nasal.
In the second year, the boy has improved a lot in pronunciation. The only problem
remained was retroflex and final nasal. He still replaced retroflexes with alveolars
and deretroflexed the retroflex vowel. In addition, he changed the final velar nasal
following /i/ to the alveolar nasal.
The third year is similar to the second year. The errors in retroflex and final nasal
seemed to have fossilized. He still replaced retroflexes with alveolars and dere-
troflexed the retroflex vowel and still changed the final velar nasal following /i/ to
the alveolar nasal. The phonological development of the two-year-old Taiwanese
boy is summarized in Table 2.22.
Table 2.22 Acquisition and substitutions of the sounds from the Taiwanese boy: 2, 2 + 1, and 2 + 2 years of age
78
year
p p, f, m p (94.25 %), f (4.59 %), m (1.14 %) p p
ph ph, p ph (88.88 %), p (11.11 %) ph ph
m m, p m (98.87 %), p (1.12 %) m m
f f f f
t t, l t (99.1 %), l (0.89 %) t t, ø ø once
th th, tɕh th (97.91 %), tɕh (2.08 %) th th, ø ø once
n n, l n (97.61 %), l (2.38 %) m n, l n (97.18 %)
l (2.82 %)
l l, ʐ l (97.29 %), ʐ (2.7 %) l l, ø ø once
k k, ø k (99.15 %), k k
ø (0.84 %)
kh kh, x kh (97.5 %), kh kh
x (2.5 %)
x x, ø x (99.23 %) x x, ø ø once
ø (0.76 %)
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ tʂ, ts tʂ (41.98 %) tʂ, ts tʂ (6.25 %), ts (93.75 %) tʂ, ts tʂ (50.43 %), ts (49.57 %)
ts (58.01 %)
tʂh tʂh, tsh tʂh (35.71 %), tsh (64.28 %) tʂh, tʂh (14.29 %), tsh tʂh, tsh, tʂh (66.66 %), tsh (16.67 %), ts
tsh (85.71 %) ts (16.67 %)
(continued)
year
ʂ ʂ, s, ts, ʂ (63.46 %), s (35.25 %), ts (0.64 %), ɕ s 100 % ʂ, s ʂ (55.38 %), s (44.62 %)
ɕ (0.64 %)
ʐ ʐ, l ʐ (21.05 %), l (78.94 %) l 100 % ʐ, l ʐ (88.89 %), l (11.11 %)
ts ts ts ts
tsh tsh tsh tsh
s s, f s (97.95 %), f (2.04 %) s s, ʂ ʂ once
ŋ ŋ ŋ ŋ
i i i i
u u, ø u (98.43 %), ø (1.56 %) u u
y y y y
a a a a
o o o o
ə ə, o, əŋ ə (96.41 %), o (3.4 %), əŋ (0.3 %) ə, o ə (95 %), o (5 %) ə
ei ei, i ei (96.96 %), i (3.03 %) ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an, a an (98.41 %), a (1.58 %) an an

ən ən, ə, ən (89.47 %) ən ən
aŋ ə (5.26 %)
aŋ (5.26 %)
(continued)
79
80

year
aŋ aŋ, a, aŋ (90.9 %), a (6.06 %), an (3.03 %) aŋ aŋ
an
əŋ əŋ əŋ əŋ
ə˞ ə˞, ə ə˞ (84.21 %), ə (15.78 %) ə˞, ə ə˞ (83.33 %), ə ə˞, ə ə˞ (15.38 %), ə (84.62 %)
(16.67 %)
ia ia ia ia
ua ua ua ua
ŋ ŋ ŋ ŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien, ie ien (96.29 %), ie (3.7 %) ien ien
uan uan, ua uan (98.03 %), ua (1.96 %) uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
(continued)
year
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ, in iŋ (2.56 %), in (97.43 %) iŋ, in iŋ (40 %), in (60 %) iŋ, in iŋ (41.67 %), in (58.33 %)
ɔŋ ɔŋ ɔŋ ɔŋ
uo uo, o, u uo (97.95 %), o (1.02 %), u (1.02 %) uo uo
81
2.5.11 The Two-Year-Old Taiwanese Girl
In the first year, the child has acquired most phonemes in Mandarin while having
problems in producing alveolar nasals, alveopalatal fricatives, and retroflex affri-
cates. She tended to replace an alveolar nasal /n/ with an alveolar lateral [l] and
replace an alveopalatal fricative /ɕ/ with an unaspirated alveopalatal affricate [tɕ].
She pronounced retroflex affricates as alveolar affricates, which were also observed
in her parents’ speech. The father even pronounced the retroflex fricative /ʂ/ as an
alveolar fricative [s], e.g., /ʂi/́ ‘is’ ! [si]́ . The child had problems with vowel /o/
and /u/. She pronounced /o/ as [ej] and /u/ as [u], [ej] [i], or [ou] inconsistently, e.g.,
/tʂə́ kə̄ / ‘this’![tsé i kə], /khəì /̀ ‘can’ ! [khoǔ ǐ ]. Adults did not have vowel
problems.
In the second year, the alveolar nasal /n/ and an alveopalatal fricative /ɕ/ have been
acquired. However, the retroflex fricative /ʂ/ was still replaced by an alveolar
fricative [s], for example, /mǎ liŋ̀ ʂǔ / ‘potato’ ! [mǎ liŋ̀ sǔ ]. Retroflex affricates
were not observed in the data of the second year. A new error has been observed,
that is, the replacement of aspirated alveopalatal affricates with unaspirated ones
sometimes, e.g., /tɕhī ɛn pǐ / ‘pencil’ ! [tɕī ɛn pǐ ]. All the vowels have been
acquired. The child has acquired most of the phonemes except the retroflexes,
whose correct forms were not available in the input.
The child did not make any further progress in the third year. All the errors have
seemed to be fossilized. The retroflexes, fricatives, and affricates were replaced with
alveolar ones for example, /ʂaŋ́ miań / ‘on the top’ ! [saŋ́ miań ]. Aspirated
alveopalatal affricates were still replaced by unaspirated ones sometimes.
The phonological development from two-year-old Taiwanese girl for three years
is summarized in Table 2.23.
Putting in OT framework, like the Indonesian two-year-old girl, the Taiwanese
two-year-old boy also high-ranked * retroflex and so changed the retroflex affricate
to alveolar affricate in the first two years as shown in Tableau 2.9. In the third year,
the alveolar (49.57 %) and retroflex affricates (50.43 %) coexist. Those two kinds
of rankings coexist as shown in Tableau 2.9 and 2.10.
Table 2.23 Acquisition and substitutions of the sounds from the Taiwanese girl: 2, 2 + 1, and
2 + 2 years of age
Phoneme Phonetic representations, first year Phonetic Phonetic
representations, representations,
second year third year
p p p p
ph ph ph ph
m m m m
f f (50 %), x (25 %), u (25 %) f f
x, u
t t t t
th th, ɕ th th
n l n n
l l, ø ø (9 %) l l
k k k k
kh kh kh kh
x x, t t (8 %) x c
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh, tɕ (67 %) tɕh
tɕ
ɕ ɕ ɕ ɕ
tʂ ts, tɕ ts (14 %) ts ts (14 %) tʂ, ts (54 %)
ts
tʂh tʂh
ʂ ʂ s s (100 %) ʂ, s s (76 %)
ʑ ʑ
ts ts ts, tɕ (50 %) ts
tɕ
tsh tsh
s s s, ɕ ɕ (73 %) s
i i, u u (13 %) i i
u u u u
y y y
a a a a
o ei o o
ɤ ɤ, ei, i, ou, ei (6 %), ou (6 %), a ɤ ɤ
a (6 %)
e e
ai ai ai ai
ei ei ei ei
au au au au
ou ou, əŋ
(continued)

an aŋ aŋ (100 %) an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ aŋ əŋ əŋ
ɤɪ ɤɪ ɤɪ ɤɪ
Tableau 2.9 OT tableau for the Taiwanese male: 2, 2 + 1 years of age

tʂ *(+cont) *Retroflex *(+anterior)
☞ ts *
tʂ *!
Tableau 2.10 OT tableau for the Taiwanese male: 2 + 2 years of age

tʂ *(+cont) *(+anterior) *Retroflex
ts *!
☞ tʂ *
2.5.12 The Three-Year-Old Taiwanese Boy
The three-year-old Taiwanese boy had acquired all the vowels and finals in
Mandarin while had problems with some initial consonants. He had problems with
manner of articulations. Aspiration and frication seem to confuse him. He turned
aspirated voiceless alveolar stops /th/ to aspirated voiceless alveolar affricates /tsh/
while turned an unaspirated voiceless alveolar affricate /ts/ to an unaspirated
voiceless alveolar stop [t], for example, /tsaí laì / ‘again’ as [taí laì ]. He replaced an
alveopalatal fricative /ɕ/ with an unaspirated alveopalatal affricate [tɕ], for example,
/ɕǐ / ‘happy’ ! [tɕǐ ]. He pronounced the retroflex fricatives as alveolar fricatives or
stops and retroflex affricates as alveopalatal affricates, for example, /ʂi/́ ‘copula BE’
as [ɕi], /ʂou taú / ‘receive’ as [tou taú ], /tʂuē i tʂə/ ‘chase’ as [tsuē i tsə], and /tʂə̄ n tʂə̄ ŋ/
‘really’ as [tə̄ n təŋ́ ]. His mother replaced retroflexes with alveolars. No other errors
were observed in his mother’s speech.
In the second year, the child got rid of problems in aspiration and frication. He
pronounced aspirated stops and unaspirated affricates correctly. The alveopalatal
fricative /ɕ/ was also pronounced correctly. However, he still had problems with
retroflexes. Alveolar affricates coexist with the correct forms. The voiceless retro-
flex fricative /ʂ/ was consistently replaced by the alveolar one [s], e.g., /ʂu/̄
‘book’ ! [sū ], while the voiced one /ʐ/ was replaced by the lateral liquid [l], for
example, /ʐɤ́ kǒ u/ ‘hot dog’ ! [lɤ́ kǒ u].
In the third year, the child had acquired most phonemes except the typical Taiwan
Mandarin problems. They are retroflexes and the rounded high front vowel. He still
replaced the voiceless retroflex affricate /tʂ/ with voiceless alveolar affricate [ts] and
replaced the voiceless retroflex fricative /ʂ/ with the alveolar fricative [s], for
example, /ʂí/ ‘is’ as [sí]. The rounded high front vowel /y/ was pronounced correctly
in the first two years, but was replaced with the unrounded one [i] in the third year,
e.g., /ɕiǎ u ɕyɔ̀ŋ/ ‘baby bear’ ! [ɕiǎ u ɕiɔ̀ŋ]. The target forms that the child acquired
seem to be Taiwanese-interfered Mandarin. A summary of the child’s phonemes
and phonetic variants in three years is given in Table 2.24.
Table 2.24 Acquisition and substitutions of the sounds from the Taiwanese boy: 3, 3 + 1, and
3 + 2 years of age
Phoneme Phonetic representations, Phonetic representations, Phonetic representations,
first year second year third year
p p p p
ph ph ph ph
m m m m
f f f f
t t t t
th tsh th th
n n n n
l n l l
k k k k
kh kh kh kh
x x x, f x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ, tɕ, ʂ ʂ (29 %) ɕ ɕ
tʂ tʂ, tɕ, t, tʂ (23 %), tɕ tʂ, ts, s ts (97 %), s tʂ (50 %), ts
ts (3 %), (2 %) (50 %)
t (31 %), ts
(43 %)
(continued)

Phoneme Phonetic representations, Phonetic representations, Phonetic representations,
first year second year third year
tʂh tʂh, tɕh, th (17 %), ts tʂhtsh, ts tʂh (40 %), ts tʂh
th, ts (25 %) (60 %)
ʂ ʂ, s, t, ɕ s (79 %), t (4 %), s s (100 %)
ɕ (4 %) (100 %)
ʑ l l (100 %) l l (100 %)
ts ts, t ts, t (11 %) ts ts ts ts
tsh ts ts (100 %) tsh tsh tsh tsh
s ɕ ɕ s s s s
i i i i i i i
u u, i u, i (9 %) u u u u
y y y y y i (100 %) i (100 %)
a a a a a a a
o o o o o o o
ɤ ɤ ɤ ɤ ɤ
e a a e e e e
ai ai ai ai ai ai ai
ei ei ei ei ei ei ei
au au au au au au au
ou ou ou ou ou ou ou
an an, aŋ an, aŋ (5 %) an an an an
ən ən ən ən ən ən ən
ɑŋ ɑŋ ɑŋ ɑŋ ɑŋ ɑŋ ɑŋ
əŋ əŋ əŋ əŋ əŋ əŋ əŋ
ɤ.ɪ ɤ.ɪ ɤ.ɪ ɤ.ɪ ɤ.ɪ
in in in
2.5.13 The Three-Year-Old Taiwanese Girl
The three-year-old Taiwanese girl’s major problem is retroflexes with occasional

errors in bilabial stops and vowel sequences. She replaced retroflexes with alveo-
lars; for example, /tʂǔ / ‘cook’ was pronounced as [tsǔ ], /tʂhù faŋ̀ / ‘kitchen’ as [tshù
faŋ̀ ], /ʂū / ‘lose’ as [sū ], /ʐəǹ / ‘person’ as [̌ ləǹ ], and /ʐɔŋ̀ xuá / ‘melt’ as [̌ nɔŋ̀ xuá ].
Occasionally, she replaced a bilabial stop with a velar stop; for example, phú khə́
phaì / ‘poker’ was pronounced as [khú khə́ phaì ]. She occasionally omits the /u/ in /
uo/ sequence, for example, /tsū o í tsū o/ ‘rub’ as [ts_ō í ts_ō ]. The mother also
replaced retroflexes with alveolars, which is a typical phonetic feature of Taiwan
Mandarin.
In the second year, the only problems left were retroflexes and omission of /u/ in
/uo/ sequence. As in the first year, she still replaced retroflexes with alveolars. Other
errors have disappeared.
In the third year, the girl consistently replaced retroflexes with alveolars. They seem
to have fossilized. The aspirated retroflex affricate /tʂh/ was pronounced as alveolar
affricate with or without aspiration. The phonemes and phonetic variations of the
Taiwanese three-year-old girl for three years are summarized in Table 2.25.
Both Taiwanese three-year-old boy and girl high-ranked *Retroflex and thus
pronounced retroflex fricative as alveolar fricative throughout three years as in
Tableau 2.11.
2.5.14 The Four-Year-Old Taiwanese Boy
The four-year-old Taiwanese boy’s pronunciation was not clear. He replaced a

retroflex affricate with an aspirated alveolar stop and dropped the final nasal. The
dropping of final nasal was also observed in /thiŋdaw/ ‘hear’ ! [thi daw]. In
addition, he replaced a retroflex fricative [ş] with an alveolar fricative [s]. He
deleted the voiceless velar fricative before back vowels, e.g., /xau/ ‘ok’ ! [au]. We
also noticed cross-segmental nasal assimilation, e.g., /nali/ ‘where’ ! [nani].
In the second year, the omission of final nasals and voiceless velar fricatives was no
longer observed. Neither was nasal assimilation. The voiced retroflex fricative /ʐ/
was acquired. The only problem left was the replacement of voiceless retroflex
fricatives and affricates—/ʂ/, /tʂ/, and /tʂh/ with alveolar ones—[s], [ts], and [tsh],
e.g., /ʂā fā / ‘sofa’ ! [sā fā ], /tʂǐ / ‘paper’ ! [tsǐ ], and /tʂhǐ / ‘ruler’ ! [tshǐ ].
Table 2.25 Acquisition and substitutions of the sounds from the Taiwanese girl: 3, 3 + 1, and 3 + 2 years of age
88
p p p p
ph ph, kh kh (10 %) ph ph
m m m m
f f f f
t t t t
th th, x x (5 %) th th
n n m n
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ ts 100 % ts 100 % ts 100 %
tʂh tsh 100 % tʂh, tsh tʂh (20 %) tsh (80 %) tʂh, tsh tsh (57.14 %), ts (42.86 %)
ʂ s 100 % s 100 % s 100 %
ʐ l, n l (85 %), n (25 %) ʐ, l ʐ (33.33 %), l (66.67 %) l 100 %
ts ts ts ts
tsh tsh tsh tsh
s s s s
ŋ ŋ ŋ ŋ
i i, u, ian u (5 %), ian (5 %) i i
(continued)
u u, ø ø (10 %) u, ø u (69.23 %), ø (30.77 %) u
y y y y
a a a a
o o, ͻŋ ɔŋ (2 %) o o
ə ə ə ə
ei ei ei ei
ai ai, an an (6 %) ai ai
e e e e
au au au au
ou ou ou ou
an an, ai ai (3 %) an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ əŋ əŋ
ə˞ ə˞ ə 100 % ə˞
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
(continued)
89
90
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ iŋ iŋ
uo uo uo uo
Tableau 2.11 OT tableau for ʂ *Retroflex *(+anterior) *(+strident)

the Taiwanese children: 3,
3 + 1, and 3 + 2 years of age ☞s * *
ʂ *! *
In the third year, the child acquired the aspirated retroflex affricate /tʂh/, but still
replaced unaspirated voiceless retroflex affricates—/tʂ/ and the voiceless retroflex
fricative /ʂ/ with alveolar ones—[ts] and [s], e.g., /tʂə́ tʂā ŋ/ ‘this piece’ as [tsə́ tsā n]
and /ʂi/́ ‘is’ as [si]́ . A new error was observed. That is the replacement of final velar
nasal by alveolar nasal. The child’s phonological development for three years is
summarized in Table 2.26.
4 + 2 years of age
p p, ph p (86 %) p p
ph (14 %)
ph ph, tɕh ph (89 %) ph ph
tɕh (11 %)
m m m m
f f f f
t t, ts t t
th th, kh, th (56 %) th th
ph, t kh (27 %)
ph (9 %)
t (9 %)
n ŋ n n
l l, n l (87 %) l l
n (13 %)
k k, th th (9 %) k k
kh kh, th kh kh
x x, ø x (73 %) x x
ø (27 %)
tɕ tɕ, ɕ tɕ (92 %) tɕ tɕ
ɕ (8 %)
tɕh tɕh, tɕ, tɕ (25 %) tɕh tɕh
ts ts (13 %)
ɕ ɕ ɕ ɕ
tʂ tʂ, ts, t, tʂ (71 %), ts (12 %), t ts tʂ, tʂ (52 %),
th, ʂ (12 %), th (2 %), ʂ (2 %) (100 %) ts ts (48 %)
(continued)

tʂh tʂh, tsh, tʂh (6 %), tsh (59 %), t tʂh tʂh
t, th, kh (18 %), th (18 %), kh (100 %)
ʂ ʂ, s, ɕ, ts ʂ (14 %), s (79 %), ɕ (5 %), s ʂ, ʂ (50 %), s
ts (2 %) (100 %) s (50 %)
ʐ z ʐ ʐ
ts ts, tɕ, t, ts (62 %), s (15 %), ɕ (8 %), ts ts
kh ts (15 %)
tsh tsh, kh tsh, kh (75 %) tsh
s s s s
i i i i
u i u u
y j, w y y
a a a a
o ŋ o o
ɤ ɤ, ei ei (13 %) ɤ
e e e e e
ai ai, a a (15 %) ai ai
ei ei ei ei
au au au au
ou ou ou ou
an a a (100 %) an an
ən əŋ əŋ (100 %) ən ən
aŋ a a (100 %) aŋ aŋ
əŋ əŋ əŋ əŋ
ɤ.ɪ ɤ.ɪ ɤ.ɪ
2.5.15 The Four-Year-Old Taiwanese Girl
The Taiwanese four-year-old girl has acquired most of the phonemes in Mandarin
and just replaced retroflexes with alveolars and changed final velar nasals to
alveolar nasals. She replaced retroflexes with alveolars as /tʂə́ kə/ ‘this one’ was
pronounced as [tsə́ kə], [tʂ̄ h] ‘eat’ as [ts̩̄ h], and /ʂ/ ‘is’ as [s̩ ].
́ He changed the velar
nasal following high front vowel to the alveolar nasal; for example, /ly̌ ɕiŋ̀ / ‘travel’
was pronounced as [ly̌ ɕiǹ ]. The same errors were observed in the mother.
Tableau 2.12 OT tableau for the Taiwanese children: 4, 4 + 1, and 4 + 2 years of age
ʂ Ident (astrident) *Retroflex *(+anterior) Ident (aanterior)
ʂ *!
☞s * *
Phonological development was stabilized in the second year. She consistently

replaced retroflexes with alveolars. /tʂ/ was realized as [ts], /tʂh/ as [tsh], /ʂ/ as [s],
and /ʐ/ as [l] 100 % of time. She changed final velar nasals to alveolar nasals next to
/i/ sometimes.
In the third year, the correct target of retroflex coexisted with alveolar variations
with low frequency. The retroflex mid-central lax vowel /ə˞/ was consistently
pronounced without retroflex [ə], which was pronounced correctly in the last two
years. Final velar nasal next to /i/ was replaced by alveolar nasal with higher
frequency 62.96 % (Table 2.27). The child’s phonological development for three
years is summarized in Table 2.27.
The Taiwanese four-year-old boy and girl also high-ranked * retroflex and so
changed the retroflex fricative to alveolar fricative as shown in Tableau 2.12. The
high ranking of *retroflex by both Indonesian and Taiwanese children proves it to
be a universal constraint.
2.5.16 The Five-Year-Old Taiwanese Boy
The five-year-old Taiwanese boy had difficulties with some places of articulation.
He mixed alveolars and retroflexes, changed alveolar nasal to lateral, and changed
final velar nasal to alveolar. He used retroflexes and alveolars interchangeably. He
could pronounce both, but was not clear when to use which one. Retroflexes were
changed to alveolars in the following examples. /tʂə́ lǐ / ‘here’ was pronounced as
[tsə́ lǐ ]. /saí tʂhə̄ / ‘racing car’ was pronounced as [saí tsə̄ ]. /ʂǔ ei kǔ o/ ‘fruit’ was
pronounced as [sǔ ei kǔ o], and /ʐəǹ / ‘person’ was pronounced as [ləǹ ]. On the other
hand, alveolars were changed to retroflexes in the following examples. /tshəŋ̀ /
‘level’ was pronounced as [tʂhəǹ ], /tš sə/́ ‘purple’ as [tʂ̌ sə]́ , and /sā n/ ‘three’ as
[ʂā n]. He changed the alveolar nasal to a lateral as /naǹ / ‘difficult’ was pronounced
̀ ]. Velar nasals in coda were changed to alveolar nasals; for example, /t ī ŋ/
h
as [lan
‘listen’ was pronounced as [t ī n] and /tɕiŋ́ ts̩ / ‘mirror’ was pronounced as [tɕiń ts̩ ].
h
Table 2.27 Acquisition and substitutions of the sounds from the Taiwanese girl: 4, 4 + 1, and 4 + 2 years of age
94
year
p p p p
ph ph ph ph
m m, p, ø m (99.20 %), p (0.39 %), ø (0.39 %) m m
f f, x f (90 %), x (10 %) f f
t t t t
th th, ø th (96.77 %), ø (3.22 %) th th
n n m n, l l once
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ, ø tɕ (98.24 %), ø (1.75 %) tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ, tɕh ɕ (98.93 %), tɕh (1.06 %) ɕ ɕ
tʂ tʂ, ts, tɕ tʂ (18.23 %), ts (80.50 %), tɕ (1.25 %) ts 100 % tʂ, ts tʂ (6.85 %), ts (93.15 %)
tʂh tʂh, tsh, x tʂh (15.38 %), tsh (80.76 %), x (3.84 %) tsh 100 % tʂh, tsh tʂh (7.14 %), tsh (92.86 %)
ʂ ʂ, s, ɕ ʂ (21.66 %), s (77.22 %), ɕ (1.11 %) s 100 % ʂ, s ʂ (3.1 %), s (96.90 %)
ʐ ʐ, l ʐ (33.33 %), l (66.66 %) l 100 % ʐ, l ʐ (15.38 %), l (84.62 %)
ts ts ts ts
tsh tsh tsh tsh
s s s s
(continued)
year
ŋ ŋ ŋ ŋ
i i i i
u u, iau, ien u (97.72 %), iau (1.13 %), ien (1.13 %) u u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai, uei ai (97.05 %), uei (2.94 %) ai ai
e e e e
au au au au
ou ou ou ou
an an, iaŋ an (97.61 %), iaŋ (2.38 %) an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ əŋ əŋ
ə˞ ə˞ ə˞ ə 100 %
ia ia ia ia
ua ua ua ua
ͻŋ ͻŋ ͻŋ ͻŋ
ie ie ie ie
(continued)
95
96
year
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ, in iŋ (3.22 %), in (96.77 %) iŋ, in iŋ (57.14 %), in (42.86 %) iŋ, in iŋ (37.04 %), in (62.96 %)
uo uo, o uo (99.15 %), o (0.84 %) uo uo
His mother also replaced retroflexes with alveolars, but not the other way
around. For example, /tʂə̄ n tə/ ‘real’ was pronounced as [tsə̄ n tə]. The change of
velar nasals in coda to alveolar nasals was also observed in the mother. For
example, /ɕī en ʂə̄ ŋ/ ‘husband’ was pronounced as [ɕī en ʂə̄ n].
In the second year, the aspirated voiceless bilabial stop was pronounced without
aspiration. The frequency of the replacement of retroflexes with alveolars
decreased. The voiced retroflex fricative /ʐ/ was pronounced correctly. The retroflex
mid-central vowel was pronounced without retroflex all the time. Final velar nasals
were changed to alveolar nasals sometimes.
In the third year, the aspiration problem disappeared. The only problem left is
retroflex. Retroflex affricates and fricatives were replaced with alveolar counter-
parts. Voiced retroflex fricative /ʐ/ was replaced by [l] again as in the first year most
of the time. The phonetic variations of the Taiwanese five-year-old boy for three
years are summarized in Table 2.28.
5 + 2 years of age
Phoneme Phonetic representations, Phonetic Phonetic
first year representations, representations,
p p p p
ph ph p 100 % ph
m m m m
f f f f
t t t t
th th th th
n n n n
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
(continued)

tʂ ts 100 % tʂ, tʂ (30.77 %) ts 100 %
ts ts (69.23 %)
tʂh tʂh 100 % tʂh, tʂh (20 %) tsh 100 %
tsh tsh (80 %)
ʂ s 100 % s 100 % ʂ, s ʂ (1.11 %), s
(98.89 %)
ʐ ʐ, l, ʐ (4.76 %), l ʐ ʐ, ʐ (6.67 %), l
n, z (88.1 %), n l, z (86.66 %), z
(4.76 %), z (6.67 %)
(2.38 %)
ts ts, tʂ once ts ts
tʂ
tsh tsh, tʂh once tsh tsh
tʂh
s s s s
ŋ ŋ ŋ ŋ
i i i i
u u u u
y y, i y (95.24 %), i y y
(4.76 %)
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ ən 100 % əŋ, əŋ (50 %), əŋ
ən ən (50 %)
ə˞ ə˞, ə˞ (28.57 %), ə ə 100 % ə 100 %
ə (71.43 %)
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
(continued)

ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien ien ien
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ, iŋ (26.09 %), in iŋ, iŋ iŋ, iŋ (70 %), in
in (73.91 %) in (81.82 %), in (30 %)
in
(18.18 %)
uo uo uo uo
2.5.17 The Five-Year-Old Taiwanese Girl
The child can pronounce all the phonemes in Mandarin quite accurately except
retroflexes, which were sometimes replaced by alveolar counterparts. Voiced ret-
roflex fricatives /ʐ/ were also replaced by the lateral [l] sometimes, e.g., /xuá i ʐəǹ /
‘bad person’ as [xuá i ləǹ ]. The mother also replaced retroflexes with alveolars, but
never changed the manner of articulation.
In the second year, the child continued to replace retroflexes with alveolars. The
voiceless retroflex fricative /ʂ/ was consistently replaced with the voiceless alveolar
fricative [s]; for example, /ʂǔ ei kǔ o/ ‘fruit’ was pronounced as [sǔ ei kǔ o] and the
voiced retroflex fricative /ʐ/ was consistently replaced by the alveolar lateral /l/,
̀ / ‘human being’ ! [ləǹ ]. In addition to the aspirated voiceless alveolar
e.g., /ʑən
affricate [ts] used to replace the aspirated voiceless retroflex affricate /tʂh/ in the first
year, two other phonetic variants—/tɕh/ and /th/—showed up in the second year,
e.g., /tʂhuɑŋ̀ / ‘bed’ as [tshuɑŋ̀ ]. The child might have noticed his inaccurate pro-
nunciation and was trying other variants to approach the target.
All the phonemes were acquired in the third year except the unaspirated retroflex
affricate /tʂ/, which was replaced with the alveolar affricate [ts] sometimes. The
child’s phonological development for three years is summarized in Table 2.29.
Let us put /tʂ/ in OT framework. The five-year-old Taiwanese children
high-ranked * Retroflex and so changed retroflex affricate to alveolar affricates
throughout three years as in Tableau 2.13.
5 + 2 years of age
Phoneme Phonetic representations, Phonetic Phonetic representations,
first year representations, third year
second year
p p p p
ph ph ph ph
m m m m
f f f f
t t, th t (80 %), th (20 %) t t
th th th th
n n n n
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ ts, s ts (86 %), s (14 %) tʂ, ts, tɕ ts (93 %) tʂ, ts tʂ (46 %), ts (54 %)
tʂh tsh tsh (100 %) tsh, th tsh (50 %), tʂh
th (50 %)
ʂ ʂ, s ʂ (2 %), s (98 %) ʂ, ɕ s (100 %)
ʑ l l (100 %) l l (100 %)
ts ts, tɕ ts (10 %), tɕ (90 %) ts, tʂ ts
tsh tsh tsh
s s s, ʂ s
(continued)
Phoneme Phonetic representations, Phonetic Phonetic representations,
first year representations, third year
second year
i i i i
u u u u
y y y y
a a a a
o o o o
ɤ ɤ ɤ
e e e e
ai ai ai ai
ei ei ei ei
au au au au
ou uo ou ou
an an an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ əŋ əŋ
ɤ.ɪ ɤ.ɪ ɤ.ɪ
Tableau 2.13 OT tableau for tʂ *(+cont) *Retroflex *(+anterior)

the Taiwanese children: 5,
5 + 1, 5 + 2, and 5 + 3 years ☞ ts *
of age tʂ *!
2.5.18 The Six-Year-Old Taiwanese Boy
The child has acquired all the phonemes in Mandarin. The only errors sometimes
occurred were in retroflex consonants and final velar nasals. He sometimes pro-
nounced retroflex consonants as alveolar consonants just like his mother, e.g., /lǎ u
ʂī / ‘teacher’ as [lǎ u sī ]. This is actually quite common in Mandarin in Taiwan. Also,
he replaced the final velar nasal with an alveolar nasal, e.g., /thaŋ/ ‘soup’ ! [than].
In the second year, the child pronounced the aspirated retroflex affricate /tʂh/ and
voiced retroflex fricative /ʐ/ correctly, but still replaced unaspirated retroflex
affricate /tʂ/ and voiceless retroflex fricative /ʂ/ with their alveolar counterparts—
[ts] and [s] sometimes, e.g., /tʂǐ / ‘paper’ as [tsǐ ] and /xaì ʂǐ / ‘still’ as [xaí sǐ ]. The
final velar nasal was produced accurately.
No progress was made in the third year. The child consistently replaced the
unaspirated retroflex affricate /tʂ/ with the alveolar affricate [ts]. As for the voiceless
retroflex fricative /ʂ/, the correct phonetic representation coexisted with the alveolar
counterparts—[s]. A summary of the acquisition of sounds by the six-year-old boy
is presented in Table 2.30.
6 + 2 years of age
p p p p
ph ph ph ph
m m m m
f f f f
t t, th th (20 %) t t
th th th th
n n, l l (95 %) b b
l l l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ tɕ
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ tʂ, ts ts (94 %) tʂ, ts ts (88 %) tʂ, ts ts (41 %)
tʂh tʂh, tsh, tɕh tah (68 %), tɕh (18 %) tʂh, ts ts (20 %) tʂh
ʂ ʂ, s s (97 %) ʂ, s s (33 %) ʂ, s s (50 %)
ʑ ʐ, z z (81 %) ʑ
ts ts ts ts
tsh tʂh tʂh (100 %) tsh
s s s s
i i i i
u u u u
y y y y
a a, ai ai (25 %) a a
o o o o
ɤ ɤ ɤ
(continued)

e e e e
ai ai ai ai
ei ei ei ei
au au au au
ou ou ou ou
an an an an
ən ən ən ən
aŋ an an (20 %) aŋ aŋ
əŋ əŋ əŋ əŋ
ɤ.ɪ ɤ.ɪ ɤ.ɪ
2.5.19 The Six-Year-Old Taiwanese Girl
The Taiwanese six-year-old girl mixed alveolar laterals with alveolar nasals,
replaced retroflexes with alveolars, changed velar nasals in coda to alveolars, and
simplified vowel sequence. The child used alveolar laterals and nasals inter-
changeably. For example, /lǎ n tuó / ‘lazy’ was pronounced as [nǎ n tuó ] and /nioù
nǎ i/ ‘milk’ was pronounced as [liòu nǎ i]. She replaced retroflexes with alveolars.
For example, she pronounced /tʂ̩̌ / ‘paper’ as [tš̩ ], /xǔ o tʂhə̄ / ‘train’ as [xǔ o tshə̄ ], /ʂū /
‘book’ as [sū ], and /ʐə̀n/ ‘person’ as [lə̀n]. The velar nasal was changed to an
alveolar nasal as /mìŋ thiē n/ ‘tomorrow’ was pronounced as [mìn thiē n]. The /i/ in /
ie/ sequence was dropped sometimes as /ná pī en/ ‘there’ was pronounced as [ná
pēn].
Most of the child’s errors were also observed in the mother such as mixing
alveolar laterals with nasals, replacing retroflexes with alveolars, and changing
velar nasals to alveolar nasals. In addition, the mother sometimes pronounced the
voiceless labiodental fricative as a velar fricative. For example, /ī fù/ ‘clothes’ was
pronounced as [ī xù].
In the second year, the exchange of laterals and nasals disappeared. However, she
still replaced retroflexes with alveolars with lower frequency. The velar nasal was
changed to an alveolar nasal when followed by /i/ with lower frequency.
In the third year, the retroflexes were replaced by alveolars consistently. The
mid-central retroflex vowel was pronounced without retroflex all the time. The
problems with final nasals disappeared. The phonological development seems to
have stabilized. The acquisition and phonetic variations of the Taiwanese
six-year-old girl are summarized in Table 2.31.
Let us see the acquisition of retroflex in the OT framework. The six-year-old
Taiwanese girl high-ranked the faithfulness constraint ident (aretroflex) and
changed alveolar affricate to retroflex affricate. That could be due to overcorrection
when acquiring retroflexes. She demoted it and reached the target in the following
two years as shown in Tableau 2.14 and 2.15.
6 + 2 years of age
representations, second representations,
year third year
p p p p
ph ph ph ph
m m m m
f f f f
t t t t
th th th th
n n, l n (98 %), l (2 %) m n
l l, n l (99 %), n (1 %) l l
k k k k
kh kh kh kh
x x x x
tɕ tɕ tɕ, tɕ (92.86 %), tɕh tɕ
tɕh (7.14 %)
tɕh tɕh tɕh tɕh
ɕ ɕ ɕ ɕ
tʂ tʂ, tʂ (3 %), ts (97 %) tʂ, tʂ (22.22 %), ts ts 100 %
ts ts (77.78 %)
tʂh tʂh, tʂh (8 %), tsh (92 %) tʂh, tʂh (66.67 %), tsh 100 %
tsh tsh tsh (33.33 %)
ʂ ʂ, s ʂ (1.1 %), s (98.9 %) ʂ, s ʂ (50 %), s s 100 %
(50 %)
ʐ l, n ʐ (6.4 %), l (35.4 %), n ʐ l, n l (91.49 %), n
(41.9 %), z (16.1 %) (8.51 %)
(continued)

year third year
ts ts, ts (99 %), tʂ (1 %) ts ts
tʂ
tsh tsh tsh tsh
s s s s
ŋ ŋ ŋ ŋ
i i i i
u u u u
y y y y
a a a a
o o o o
ə ə ə ə
ei ei ei ei
ai ai ai ai
e e e e
au au au au
ou ou ou ou
an an an an
ən ən ən ən
aŋ aŋ aŋ aŋ
əŋ əŋ, əŋ (11.1 %), ən (88.9 %) əŋ əŋ
ən
ə ə ə ə 100 %
ia ia ia ia
ua ua ua ua
ɔŋ ɔŋ ɔŋ ɔŋ
ie ie ie ie
ye ye ye ye
uai uai uai uai
uei uei uei uei
iau iau iau iau
iou iou iou iou
ien ien, ien (96.8 %), en (3.2 %) ien ien
en
uan uan uan uan
yen yen yen yen
in in in in
uən uən uən uən
yn yn yn yn
iaŋ iaŋ iaŋ iaŋ
(continued)

year third year
uaŋ uaŋ uaŋ uaŋ
iŋ iŋ, iŋ (20.4 %), in (79.6 %) iŋ, iŋ (72.73 %), in iŋ
in in (27.27 %)
uo uo uo uo
Tableau 2.14 OT tableau for tsh *(+anterior) Ident (aretroflex) *(+strident)

the Taiwanese children: 6 + 1 h
years of age ts *! *
☞ tʂh * *
Tableau 2.15 OT tableau for tsh Ident (as.g) Ident (aretroflex) *(+anterior)
the Taiwanese children: 6 and
6 + 2 years of age tsh *
☞ tʂh *!
2.6 Discussion and Summary
Children of Indonesian immigrants have unique and common difficulties with the
children of Vietnamese or Taiwanese mothers. The phonetic variants of children of
Indonesian and Taiwanese mothers are summarized in Table 2.32. The first column
lists the target phonemes in Mandarin. The phonetic variants observed in the
children of Indonesian were listed in the second column and those produced by the
children of Taiwanese were listed in the last column. Omission is indicated by Ø.
Like the children of Vietnamese and Taiwanese, the children of Indonesian have
difficulties in retroflexes, aspirations, and nasals. They tended to replace retroflexes
with alveolars just like many adult Mandarin speakers in Taiwan. For example, the
three-year-old Indonesian boy pronounced /tʂə́/ ‘this’ as [tsə́], which is also
observed in Huang (2011) and Hsu (1996). *Retroflex outranked ident (aretroflex).
This could be due to the universal markedness of retroflex or the children’s input.
Retroflexes usually do not preserve in adult Mandarin in Taiwan. That could just be
their target rather than an error.
Children of Indonesian mothers failed to discriminate aspiration. For example,
unaspirated bilabial stop /p/ and unaspirated velar stop /k/ were replaced by their
aspirated counterparts. Deaspirations were also observed. Fuzzy discrimination of
aspiration was also found in the children of Taiwanese mothers. Hsu (1996) also
found the substitutions of aspirated stops with unaspirated stops in the children of
Taiwanese parents.
2.6 Discussion and Summary 107
Table 2.32 Comparison of phonetic variation between children of Indonesian and Taiwanese
mothers
Phoneme Phonetic variations
Indonesian Taiwanese
p p, ph, f, ø, m p, ph, f, m
ph ph, p, x, f, th ph, p, kh, tɕh
m m, l, u m, p, ø
f f, u, p, ph, w, x f, u, x
t t, k, l, th, ø t, l, thø, ts
th th, x, ø, tsh, tʃ th, tɕh, ø, ɕ, x, kh, ph, t
n n, l, m n, l, ŋ
l l, n, k n, l, ŋ
k k, kh, ø, t k, ø, th
kh kh kh, x, th
x x, ø, u, f x, ø, t
tɕ tɕ, tɕh, t, ɕ, ts, th, ø tɕ, ɕ, ø, tɕh
tɕh tɕh, tɕ, tsh, ɕ, ø tɕh, tɕ, ts
ɕ ɕ, tɕ, tɕh, ɕ, s, kh, ts, k, t ɕ, ts, ʂ, tɕh
tʂ ts, tɕ, t tʂ, ts, tɕ, t, th, t, x
tʂh tʂh, tsh, tɕh, kh, th, s tʂh, tsh, tɕh, kh, ts, th, t, x
ʂ ʂ, s, ts, ɕs, tsh, kh, x, k, th, tʂ, t, ɕ ʂ, s, ts, ɕ, t
ʐ ʐ, l ʐ, l, n, z
ts ts, tɕ, k, tʂ, t ts, tʂ, tɕ, t, kh
tsh tsh, kh, tɕh, tʂh, tʂ, k, ø tsh, tʂh, kh
s s, ts, ʂ, kh, t, ɕ s, f, ʂ, ɕ
ŋ ŋ ŋ
All the children have problems with nasals. Younger children tended to drop
nasals in coda, while older children changed the place of articulation. The
two-year-old Indonesian girl pronounced /tan/ ‘egg’ as [ta]. The four-year-old
Indonesian boy pronounced /tshū an/ ‘wear’ as [tshū aŋ] ‘window.’ The six-year-old
Indonesian boy pronounced /ʂā ŋ/ ‘hurt’ as [san] ‘three.’ Changing the place of final
nasals was also observed in Indonesian mothers. Children of Indonesian were often
confused by aspiration like their mothers. For example, the three-year-old
Indonesian boy said /tɕin/ ‘enter’ as [tɕhin]. Children in both groups were accu-
rate in tones, regardless of the Indonesian mothers’ tone errors. The Indonesian
mother of the four-year-old girl and five-year-old boy pronounced the high-level
tone as the falling tone, for example, /ɕiā ŋ kū / ‘mushroom’ as [ɕiáŋ kū ].
The unique problems observed in the children of Indonesian include simplifi-
cation of complex vowels and preference for affricates and palatals. Younger
children of Indonesian mothers tended to simplify complex vowels, which have not
been observed in the children of Taiwanese or Vietnamese mothers. For example,
the Indonesian two-year-old girl pronounced /tiau/ ‘lose’ as [ta]. Only one vowel
was left in the three-vowel sequence. The Indonesian boy pronounced /tuan/
‘broken’ as [tan]. Only one vowel was left in the two-vowel sequence. /a/ tended to
the vowel preserved.
The challenging alveopalatal affricates for the children of Vietnamese and
Taiwanese mothers (Kuo 2008) do not seem to be a problem for the children of
Indonesian mothers. They preferred affricates to fricatives. The two-year-old
Indonesian girl pronounced the alveolar fricative /s/ as the alveolar affricate [ts], for
example, /sə/ ‘color’ as [tsə]. The six-year-old Indonesian boy pronounced the
alveopalatal fricative /ɕ/ as alveopalatal affricate [tɕ]; for example, /ɕi/ ‘river’ was
pronounced as [tɕi].
Children of Indonesian mothers changed alveopalatals to palatals, which were
not observed in the children of Taiwanese or Vietnamese mothers. The
three-year-old and four-year-old Indonesian boys changed alveopalatal fricatives to
palatal fricatives. For example, the three-year-old Indonesian boy pronounced /tɔ̄ ŋ
ɕī / ‘thing’ as [tɔ̄ ŋ ҫī ]. The four-year-old Indonesian boy pronounced /ɕiaŋ tau/
‘think’ as [ҫiaŋ tau]. These kinds of errors were also observed in their Indonesian
mothers. For example, the mother of Indonesian three-year-old boy pronounced /tś
tɕǐ / ‘self’ as [tɕí tɕǐ ].
The results of the present study show that the Mandarin phonological devel-
opment by the children of immigrant mothers might be influenced by their mothers’
Mandarin (L2), because they made similar errors. Since the same errors were
observed among the children of mothers of different nationalities, they cannot be
attributed to the mothers’ L1 alone. Despite the similar exposure to Taiwan
Mandarin as native children, children of immigrants made unique errors. This
indicates the importance of primary interlocutors, their mothers. Children of
immigrants have received input mediated by some universal constraints, whose
rankings are dynamic in the process of development. Innatism and interactionism
may not be mutually exclusive as the proposal for ‘innately guided learning’ (Gould
and Marler 1987; Jusczyk and Bertoncini 1988; Marler 1991).
References
Bernstein Rater, N. (1984). Patterns of vowel modification in mother-child speech. Journal of

Child Language, 11, 557–578.
Boersma, P., & Hayes, B. (1999). Empirical tests of the gradual learning algorithm. ROA-348.
Cheng, C. C. (1973). A synchronic phonology of Mandarin Chinese, monographs on linguistic
analysis 4. The Hague: Mouton.
Cheng, R. L. (1997). Taiwanese and Mandarin structures and their developmental trends in
Taiwan I: Taiwanese phonology and morphology. Taiwan: Yuan-Liou Publishing Co. Ltd.
Dekkers, J., Leeuw, F., & Weijer, J. (Eds.). (2000). Optimality theory: phonology, syntax, and
acquisition. Oxford: Oxford University Press.
Duanmu, S. (2000). The phonology of standard Chinese. New York: Oxford University Press.
Gleason, J. B. (2012). The development of language (8th ed.). Boston: Pearson.
References 109
Gnanadesikan, A. E. (1996). Child phonology in optimality theory: Ranking markedness and

faithfulness constraints. In Proceedings of the 20th Annual Boston University Conference on
Language Development (Vol. 1, pp. 237–248).
Gould, J. L., & Marler, P. (1987). Learning by instinct. Scientific American, 225(1), 75–85.
Hsu, J. (1996). A study of the stages of development and acquisition of Mandarin Chinese by
children in Taiwan. Taipei: Crane.
Huang, L. (2011). Mandarin consonant acquisition of children of Vietnamese and Indonesian
Mothers in Taiwan (M.A. thesis). Chiayi: National Chiayi University.
Jusczyk, P. W., & Bertoncini, J. (1988). Viewing the development of speech perception as innately
guided learning process. Learning and Speech, 31(3), 217–238.
Kager, R. (1999). Optimality theory. Cambridge: Cambridge University Press.
Crane.
Ladefoged, P., & Johnson, K. (2015). A course in phonetics (7th ed.). Independence: Cengage
Learning.
Li, Y. J. (2006). A study of the segments in hearing-impaired preschoolers’ learning of Mandarin
Chinese (Unpublished M.A. thesis). Providence University.
Liu, H. C. (1999). An optimality-theoretic analysis of Taiwanese consonant acquisition
(Unpublished M.A. thesis). National Chung Cheng University.
Liu, H. C. (2009). Speech errors in Taiwan Southern Min: A phonological study (Ph.D. Thesis).
National Tsing Hua University.
Macdonald, R. R., & Darjowidjojo, S. (2007). Indonesian reference grammar. Georgetown:
University Press.
Malsheen, B. (1980). Two hypotheses for phonetic clarification in the speech of mothers to
children. In G. Yeni-Komshian, J. F. Kavanagh & C. A. Ferguson (Eds.), Child phonology:
Vol. 2. Perception. New York: Academic Press.
Marler, P. (1991). The instinct to learn. In S. Carey & R. Gelman (Eds.), The Epigenesist of mind:
Essays on biology and cognition. Hillsdale, NJ: Lawrence Erlbaum Associates.
McCarthy, J. (2008). Doing optimality theory: Applying theory to data. Hoboken: Blackwell
Publishing.
McCarthy, J., & Prince, A. (1994). The emergence of the unmarked: Optimality in prosodic
morphology. In Proceedings of the North East Linguistic Society (Vol. 24, pp. 333–379).
Amherst (Mass.): GLSA. ROA-013
Menn, L. (1971). Phonotactic rules in beginning speech. Lingua, 26, 225–241.
Prince, A., & Smolensky, P. (1997). Optimality: From neutral networks to universal grammar.
Science, 275, 1604–1610.
Prince, A., & Smolensky, P. (2004). Optimality theory: Constrain interaction in generative
grammar. Malden: Blackwell.
Rau, D. H., & Li, M, C. (1994). Phonological variation of /tʂ/, /tʂh/, /ʂ/ in Mandarin Chinese. In
Proceeding of the Fourth International Conference on Chinese teaching Pedagogy, Taiwan
(pp. 345–361). Taipei: World Chinese Language Association.
Tesar, B., & Smolensky, P. (2000). Learnability in optimality theory. Cambridge, MA: The MIT
Press.
Velleman, S. (1988). The role of linguistic perception in later phonological development. Journal
of Applied Psycholinguistics, 9, 221–236.
Vietnamese language. (2007, September 28). Wikipedia, the free encyclopedia. Retrieved
September 28, 2007 from http://en.wikipedia.ord/w/index.php
Chinese
Chung, Rong Fu. 鐘榮富. (2002).台語的語音基礎 “Taiyu de Yuyin Jichu” (The nature of
Taiwanese). Taipei: The Crane Publishing Company.
Lin, Ling-Ying. 林玲英. (2005). 越籍配偶漢語語音偏誤現象初探 “Yueji peiou hanyu yuyin
pianwu xianxiang chutan” (The study of Vietnamese foreign spouses’ phonetic errors in
learning Mandarin Chinese). National Conference on Linguistics. Hsinchu: National
Chiao-Tung University.
Chapter 3
Lexical Development
3.1 General Literature on Lexical Development
Vocabulary is crucial not only because words allow speakers to express themselves,
but also because of the strong association between vocabulary and reading. If
students do not have sufficient vocabulary to understand the texts at school, they
cannot learn the content-area material for academic success.
Very young children understand the pragmatic intent of adults’ utterances before
they can understand the words themselves. This earlier comprehension is at the
emotional, social, and contextual levels. Only very slowly do children come to
understand the use of words independent of context (Gleason 2012). In this chapter,
I describe the theories for lexical development, phenomena in lexical development
in general and in Mandarin. Then, I will discuss the present study on Mandarin
lexical development of children of Indonesian and Taiwanese mothers in Taiwan.
3.1.1 Theories of Lexical Development
There are two major theories of lexical development: learning theory and develop-
mental theories. According to learning theories, children learn the meanings of their
first words by associative learning. That is, by repeated exposure to a word (e.g.,
hearing doggie) paired with a particular experience (e.g., seeing the family dog), the
child makes connections between the word and meanings. Exclusively replying on
associative learning would be slow and result in many errors. However, children’s
learning of words is rapid, predictable, and remarkably accurate (Gleason 2012).
In contrast to the behavioral model, developmental theories consider semantic
development within the wider context of the child’s unfolding social, cognitive, and
linguistic skills (Gleason 2012). Clark (1993) claimed that by the time children start
to learn a language, they have developed a set of ontological categories including

DOI 10.1007/978-981-10-1035-4_3
112 3 Lexical Development
objects, actions, events, relations, states, and properties. They learn words to match
these categories. In order to perform this mapping, children rely on their rudi-
mentary understanding of speakers’ attention and intentions and how these relate to
what is likely to be communicated (Tomasello 1995). Other lexical principles have
been suggested to assist children’s word learning task, for example, whole object
principle (Markman and Wachtel 1988), principle of mutual exclusivity (Markman
1987; Markman and Wachtel 1988) and principle of contrast (Clark 1987).
However, these lexical principles can be overridden by linguistic and world
knowledge (e.g., Akhtar 2002; Hall 1994).
3.1.2 Phenomena of Lexical Development
Children as young as 18 months old start to make word-referent mapping after only
a few exposure to a new word, often also without explicit instruction (Houston-Price
et al. 2005). By early in their second year, most children have begun to produce
words themselves. They begin with words related to what is intellectually and
socially most meaningful to them (Anglin 1995). Also, early words tend to share
phonetic features. They are shorter and occur frequently in speech (Storkel 2004).
Children’s early productive vocabulary consists of words from a variety of
grammatical classes, while common nouns constitute nearly 40 % of the average
English—speaking child’s first 50 words (Bates et al. 1994). Among nouns, those
that are easiest to distinguish from the surroundings, such as animate beings or
things that move, are the earliest learned (Gentner 1999). This nouns bias has also
been found in several other languages such as Spanish, Dutch, French, Hebrew,
Italian, despite striking structural differences across languages (Bornstein and Cote
2004). However, some research shows less marked noun preference in Mandarin or
Korean (Tardif et al. 1999; Choi and Gopnik 1995). Several explanations about
noun bias have been proposed. One hypothesis is that children’s vocabulary reflects
the input directed to them. Studies have shown that labeling of objects dominates
the adult speech to English-speaking children (Goldfield 1993). An alternative
explanation is that nouns are linguistically simpler than verbs. In addition, the
concepts represented by nouns are clear and concrete (Gentner 1983, 1988). The
first proposal can account for differences across languages in acquisition, while the
second will predict universality in language acquisition. In my study, I will
investigate whether this noun bias exists in children acquiring Mandarin in Taiwan.
3.1.3 Adult Role in Lexical Development
Even before children begin producing words themselves, adults’ labeling and
gazing behaviors serve to focus children’s attention on objects. As children acquire
language, adults’ speech to them incorporate increasingly rich information about
3.1 General Literature on Lexical Development 113
the categories they are acquiring. For example, parents explain that ‘Bats live in
caves.’ (Gelman et al. 1998).
The labels adults provide for children seem to be based on children’s own ways
of categorizing the world (Golinkoff et al. 1995), which corresponds to so-called
basic-level categories. Basic level is the most general level at which objects are
similar because of their forms, functions, component parts (Poulin-Dubois 1995), or
motions. Mother’s speech has also been shown to have an effect on the ways that
children come to understand the use vocabulary relating to their own inner states
(Beeghly et al. 1986; Tingley et al. 1994).
In addition to the special vocabulary directed to children, some characteristics of
child-directed speech may facilitate semantic development. Input language, espe-
cially when young children begin to understand and use words, is more clearly and
slowly enunciated and is characterized by exaggerated intonation and clear pauses
between utterances (Sachs et al. 1976). In addition, the words being focused on tend
to be placed in sentence-final position with especially marked pitch and stress
(Fernald and Mazzie 1991). We wonder whether Indonesian mothers in Taiwan
with limited Mandarin proficiency are able to provide this kind of input to facilitate
their children’s Mandarin acquisition or not.
3.1.4 Lexicon of Mandarin, Vietnamese and Indonesian
Most of the input of children is from their primary caretakers, their mothers in our
study. Immigrant mothers are second language learners of Mandarin, whose
Mandarin is likely to be interfered by their first language like most second language
learners. Therefore, it is helpful to understand Indonesian vocabulary composition
when studying children of Indonesian mothers’ lexical development. A discussion
of Vietnamese is helpful for comparing with the previous study on children of
Vietnamese mothers (Kuo 2008).
As much as 60 % of Vietnamese vocabulary is derived from Chinese, especially
words relating to science and politics due to a thousand years of Chinese domi-
nation. Many compound words are Sino-Vietnamese, consisting of native
Vietnamese words combined with Chinese borrowings. A native Vietnamese word
can be distinguished from a Chinese borrowing by reduplication, which indicates
intensity. A native Vietnamese word can be reduplicated and the meaning does not
change when the tone is shifted. As a result of French colonization, Vietnamese also
has words borrowed from French. Recently many words are borrowed from
English, for example, TV (Alves 1999). Like Chinese, Vietnamese also has a
classifier system to indicate classes of nouns. Among the common categorization
bases are shape, animacy, material, and size. The classifier cái is used for most
inanimate objects, which can extend to all other classifiers.
On the other hand, the vast majority of Indonesian words, however, come from
the root lexical stock of its Austronesian (incl. Old Malay) heritage. Indonesian has
borrowed from many other languages, such as Sanskrit, Arabic, Persian,
Portuguese, Dutch, and Chinese. There are about 750 Sanskrit loan words in
modern Indonesian, 1000 Arabic loans, some of Persian and Hebrew origin, so 125
words of Portuguese origin and around 10,000 loan words from Dutch. The latter
also comprises many words from other European languages, which came via Dutch,
the so-called International Vocabulary. The Chinese loan words are usually con-
cerned with cuisine, trade, or often just exclusively things Chinese. There is a
considerable Chinese presence in the whole of Southeast Asia (Sneddon 2004).
3.2 Mandarin Lexical Development by Native Children
3.2.1 Vocabulary Composition
According to Hsu’s (1996) longitudinal study of twenty 1–5-year-old children’s

acquisition of Mandarin in Taiwan for three years, it is found that children began to
utter words shortly after the age of one. Some of the words were standard words and
others are word approximations such as [a:m] observed at 1;2 whenever the child
wanted to eat. Children’s utterances consist of one word at this stage. The major
group of the words was nouns and then followed by verbs, adjectives, adverbs, and
others. Among nouns, most of them were concrete nouns and then abstract nouns.
The second largest group of words were verbs. All types of verbs that appeared in
this stage were action verbs, process verbs, and stative verbs. There were very few
adjectives in this period, and they were all object properties such as da ‘big.’ The
only classifier used at this stage was ge. From the age of two years and up,
children’s vocabulary increased in both quantity and quality.
Basic-level nouns constitute most of the Mandarin-speaking children’s lexicon,
subordinate next and superordinate least (Jiang 2000). The frequency of superor-
dinate and subordinate words increased with age. Acquisition of verbs is relevant to
children’s perceptual development. She claimed the basic level of verbs is based on
five senses—seeing, hearing, smelling, tasting, and touching. According to her
findings, most of the verbs in children’s lexicon are based on seeing and touching.
Voice, smell, and taste are too abstract to describe.
3.2.2 Classifier Acquisition
Numeral classifiers are an important characteristic in Mandarin Chinese; thus, they

have drawn the attention of several acquisition researchers. Numeral classifier lan-
guages are those in which classifiers are obligatory in expressions of quantity, and
anaphoric or deictic expressions. Children first acquire the general classifier ge and
use it as a ‘syntactic place-holder’ (Fang 1985; Hu 1993). Children could use three
different types of numerical constructions (numeral + CL, numeral + CL + N,
3.2 Mandarin Lexical Development by Native Children 115
and numeral + CL + Adj + N) correctly. Surprisingly, children almost never use

the classifier in wrong syntactic positions. Younger children (age 3 and 4) overuse
and allow the general classifier ge to co-occur with almost any noun. For example,
children often use ge to replace lian and zhi in yi lian xiao cheche ‘one CL little car’
and yi zhi xiao gougou ‘one CL little puppy’ (Fang 1985; Hu 1993; Myers 2000;
Chien et al. 2003; Huang 2007).
Erbaugh (1986) observed two Chinese-speaking children’s (a one-year-old boy
and a two-year, ten-month-old girl) conversation with the experimenter and their
caretakers longitudinally. She found that the children’s use of classifiers could be
divided into three developmental stages. The first stage is the early use of classifiers
at the lexical stage. Children rarely use specific classifiers. The second stage is the
later use of classifiers at the prototypical stage. Children start applying specific
classifiers to their prototypical items. The third stage is the broadening use of
classifiers as generalization. Children generalized semantic categories based on
their prototypical referents and applied those rules to non-prototypical items.
Hu (1993) investigated the comprehension of 12 Chinese numeral classifiers with
24 Chinese-speaking children at age three to six in Boston. She claimed that children’s
perceptual and categorization abilities help classifier acquisition in general, but not in
the case of specific classifiers. Hu also concluded that there was a gap between
children’s comprehension and their production of these classifiers. Generally, chil-
dren at various ages show better understanding than production of these classifiers.
Chien et al. (2003) designed two comprehension tasks with objects of different
sizes, color, and shape to test Chinese children’s understanding of count-classifiers.
There were 80 participants at age three to eight and 16 adults. Fourteen
count-classifiers and four mass-classifiers were tested. The results indicated five
points. First, Chinese children, in early stages of language acquisition (even as
young as three years), honored the grammatical count–mass distinction which is
reflected at the level of classifiers. Second, Chinese children are able to distinguish
count- and mass-classifiers in a given set of classifiers. Children realized that the
relationship between a classifier and an entity is absolutely fixed. Third, Chinese
children’s abilities in dealing with mass-classifiers are comparable to their abilities
in dealing with count-classifiers. Fourth, although there are developmental differ-
ences across the classifiers tested (presumably due to lexical learning), these dif-
ferences tend to decline by age 4. Finally, the general classifier ge differs from other
classifiers in that it did not require that the entity denoted by the noun to be of a
particular type. This was seen even in adults to some degree.
Later, Huang (2007) also explored Chinese preschoolers’ acquisition of the
count- and mass-classifiers by two experiments: the comprehension and production
tasks. There were 45 participants at age three to five. Ten count-classifiers and eight
mass-classifiers were tested. The results indicated five points. First, children have
performed significantly better on count-classifiers than on mass-classifiers. Second,
the critical stage in Chinese children’s classifier development was from age three to
four. There was a significant difference on in children’s responses to overall count-
and mass-classifiers among the age groups. Third, there was a significant difference
between children’s comprehension and production tasks. Children showed better
ability in comprehension than in production of count- and mass-classifiers. Fourth,

ge was the earliest classifier acquired. Fifth, the misuse of the classifier ge indicated
overgeneralization. Children usually used ge to classify objects despite different
meanings. This means ge was considered ‘unmarked’ by children.
In the above studies, ge was considered as the default classifier for children when
they did not know which classifier to use. That is, the classifier ge is used for children
with unfamiliar objects. Children do overuse the general classifier ge widely at the
early period of their life. The overuse of the classifier ge decreases with age.
3.3 Previous Studies of Mandarin Lexical Development

by Children of Immigrants
Previous studies of Mandarin lexical development show no difference between native

and immigrant children. For example, Tsay et al. (2005) gave Peabody
Picture Vocabulary Test and a picture story telling task to school-age children and
found no difference between Taiwanese children and children of immigrants including
mothers from mainland China, whose Mandarin is native just like Taiwanese mothers.
In Kuo’s (2008) study of the lexical development of preschool children also show
that the lexical development of Vietnamese children is not slower than Taiwanese
children. Younger (age 2–4) children with Vietnamese mothers used more nouns,
verbs, and adjectives in spontaneous speech than their age mates with native
Mandarin-speaking parents. Children of Taiwanese mothers exceeded when they
were older (age 5–6). More classifiers were observed in the spontaneous and elicited
speech of children of Vietnamese mothers than children of Taiwanese mothers though.
Mandarin lexical development does not seem to cause difficulties for children of
Vietnamese, which may be due to the similarities between Vietnamese and
Mandarin. As much as 60 % of Vietnamese vocabulary is derived from Chinese,
especially words relating to science and politics due to a thousand years of Chinese
domination. Many compound words are Sino-Vietnamese, consisting of native
Vietnamese words combined with Chinese borrowings. Like Chinese, Vietnamese
also has a classifier system to indicate classes of nouns. Shape, animacy, material,
and size are common categorization bases (Alves 1999). It is a different case for
children of Indonesian immigrants, because Indonesian have no classifiers.
3.4 Mandarin Lexical Development by Children

In the present study, we collected one-hour spontaneous speech samples from two-,
three-, four-, five-, six-year-old children of Indonesian mothers and Taiwanese
mothers and designed elicitation tasks for classifiers, which were lack in
3.4 Mandarin Lexical Development by Children of Indonesian Mothers 117
Fig. 3.1 Vocabulary Vocabulary acquired by Indonesian and Taiwanese children

acquired by children of
Total number of vocabulary

Indonesian and Taiwanese 250
mothers 200
150 Indonesian
100 Taiwanese
50
0
2 3 4 5 6
Age
spontaneous speech. We focused on the acquisition of nouns, verbs, and adjectives,

which consist of most of children’s speech. In addition, classifiers, an important
feature in Chinese noun phrases, were also investigated.
The lexical development was given in Fig. 3.1. Children’s vocabulary size
increased by age in both groups. Across age groups, children of Indonesian mothers
have smaller productive vocabulary size than children of Taiwanese mothers. The
difference was largest for the two-year-olds and smallest when they were three.
There is great individual variation in vocabulary composition. Nouns dominate
most of children’s lexicon while verb bias was observed in two Indonesian children
and three Taiwanese children. The results support neither the universal noun bias
hypothesis (Gentner 1981) nor Tardif’s (1996, 1999) argument for a verb bias for
Mandarin-speaking children. The results are different from the lexical development
of children of Thai mothers, which show noun bias. Unlike native
Mandarin-speaking mothers, Thai mothers may know more nouns than verbs and
not familiar with the pro-drop structure in Mandarin, which constitute only 6.02 %
of Thai mothers’ utterances in the data, thus result in more nouns in their
child-directed speech (Kuo 2015).
3.4.1 Noun Acquisition
The nouns in the spontaneous speech of Indonesian children were more than those
of Taiwanese children. On average, Indonesian children have acquired 74.44 nouns
and Taiwanese children have acquired 69.6 nouns. The nouns are divided into
semantic categories such as animals, commodities, food, body parts, fruit, vehicle,
toys, clothing, furniture, stationary, color, plant, personal names, kinship, and place
names. The largest portion of Indonesian children’s nouns consists of animals
(12.30 %) while that of Taiwanese children consists of food (9.75 %). The details
are shown in Table 3.1. The first column indicates the semantic category in bold
and an example below. The first row indicates the nationality of the participant’s
mother. The percentage shows the proposition of the semantic category among the
Table 3.1 Semantic categories of nouns produced in child spontaneous speech

Indonesian N = 9 Taiwanese N = 10
Animal
Crocodile (%) 12.30 5.33
Body part
Head (%) 4.44 4.45
Book title
Chinese Sanzijing (%) 1.49 0
Commodities
Stove (%) 9.40 8.43
Class
Homework (%) 1.43 2
Clothes
Hat (%) 1.22 1.49
Construction
Home (%) 5.16 4.18
Cartoon character
Mickey (%) 3.64 0.93
Country
USA (%) 0.25 0
Day
Friday (%) 0.96 0
Food
Hamburger (%) 10.14 9.57
Furniture
Table (%) 0.24 0
Game
Heart attack (%) 0.79 0
Holiday
Birthday (%) 0.32 0
Horoscope
Libra (%) 0 0
Human
Volunteer (%) 11.09 6.50
Instrument
Piano (%) 0 1.54
Kinship
Uncle (%) 7.01 7.82
Language
English (%) 0.24 0
Money
One hundred dollar (%) 1.15 0.43
Monster
Monster (%) 0.22 0
(continued)

Indonesian N = 9 Taiwanese N = 10
Music
Music (%) 0.57 0.22
Substance
Water (%) 1.89 0.51
Name
Name (%) 0.24 2
Number
One (%) 8.11 5.35
Place
Xitou (%) 3.32 3.19
Plant
Flower (%) 1.68 0
Score
Zero point (%) 0 0
Shape
Circle (%) 0 0.57
Song
Song (%) 0 0
Sound
Sound (%) 0.13 0.23
Sport
Soccer (%) 0.67 0
Stationary
Eraser (%) 2.67 3.11
Story
Story (%) 0.00 0
Time
Eight o’clock (%) 1.34 0
Toy
Toy (%) 2.53 3.66
Traffic
Road (%) 0.31 0.22
Vacation
Winter vacation (%) 0 0
Vehicle
Sailboat (%) 3.80 2.94
Weapon
Gun (%) 0.06 0
Weather
Typhoon (%) 0.47 0
Mean 74.44 69.6
children’s nouns. The last row gives the mean number of nouns produced by each
group.
The following lists errors in noun usages found in Indonesian children’s spon-
taneous speech. Most of their errors involve superordinate categories as in Example
1 and relational terms as in Example 2. Example 1–2 are from the eight-year-old
Indoneisan girl, who was confused about the superordinate categories animal and
plant and relational terms husband and wife.
1. *CHI: 還有花也是(*) 動物
%pho: xài iŏu xuā iĕ s̩ ́ tɔ́ŋ ú
%eng: The flowers are animals, too.
*INV: 花…花算是(*)….植物啦!
%pho: xuā xuā suán s̩ ́tʂ̀ ú la !
%eng: Flowers…are plants.
*CHI: 動物
%pho: tɔŋ́ ú
%eng: Animal.
*INV: 植物啦! 動物是(*) 會動的啊! 對不對**會講話的啊
%pho: tʂú la ! tɔ́ŋú s̩ ́ uéi tɔ́ŋ tə tuéi pù tuéi xuéi tɕiăŋ xuá tə
%eng: They are plants. The animals can move, right? They can speak.
*CHI: 動物**不會**
%pho: tɔ́ŋ ú pù xuéi
%eng: No, animals cannot.
2. *CHI: 還有紅太狼, 紅太狼是(*) 灰太狼的老公[老婆]
%pho: xài iŏu xɔŋ̀ thái làŋ xɔŋ̀ thái làŋ s̩ ́ xuēi thái làŋ tə lău kɔ̄ ŋ
%eng: Another character is Wolnie (Red Wolf). Wolnie is Wolffy’s husband [wife].
Examples 3–5 are from the Indonesian 4-year-old boy. He was confused with
nouns sharing some features. In Example 3, he was confused between cucumbers
and eggplants, which are both long.
3. *CHI: 小[*]黃瓜
%pho: tɕ[*]iău xuaŋ̀ kuā
%eng: Cucumber.
*INV: 不是[*],這[*]是[*]茄子
%pho: pù ś̝[*] ts[*]ə́ ś̝[*] tɕʰiè ts̟
%eng: No. This is an eggplant.
In Example 4, he took light for fire because they are both bright.
4. *INV: 沒有啦,那不是[*]火啦
%pho: mèi iŏu la ná pù ś̝[*] xuŏ la
%eng: Ni, that is not fire.
*CHI: 那個…電[*]燈[*]而[*]已喔?
%pho: ná kə teń [*] tən[*] ə[̀ *] ĭ o
%eng: That….is only a light.
Example 5 involves the overextension of the vegetable suffix cai to eggplant.
5. *INV: 那這[*]是[*]什[*]麼?
%pho: ná ts[*]ə́ ṣ[́ *] s[*]ə mə
%eng: What is this?
*CHI: 高麗[*]菜….嗯….
%pho: kāu léi[*] tshái ən
%eng: Cabbage….well…
*INV: 這[*]是[*]茄子阿
%pho: ts[*]ə́ ṣ́[*] tɕhiè tṣ a
%eng: This is an eggplant.
*CHI: 茄[*]子菜[*]
%pho: tɕhè[*] tṣ tshái
%eng: Eggplant.
In Example 6, he was confused between clothes and pants.
6. *INV: 好,那這[*]是[*]什[*]麼你知[*]道嗎?
%pho: xău ná ts[*]ə́ ṣ[́ *] s[*]ə mə n ĭ tṣ[*] táu ma
%eng: Ok, do you know what this is?
*CHI: 衣[*]服
%pho: í[*] fù
%eng: Clothing.
*INV: 這[*]是[*]褲子
%pho: ts[*]ə́ ṣ[́ *] khú tṣ
%eng: They are pants.
*CHI: 褲子…阿這[*]個…
%pho: khú tṣ a ts[*]ə́ kə
%eng: Pants…this…
On the other hand, most of the errors made by children of Taiwanese parents are
due to Taiwanese transfer and overextension. Example 7 was from the
three-year-old boy, who transferred the word ‘money’ from Taiwanese. The word /
kho/ usually follows a numeral in Taiwanese, but the child reduplicated it.
Examples 8–10 are from the six-year-old Taiwanese boy. Example 8 is also caused
by Taiwanese transfer.
7. 摳摳
%pho: kho kho
%eng: money
8. Child: 擦子[橡皮擦]
%pho: tshā ts1[ɕiáŋ phі̀ tshā]
%eng: eraser
Examples 9 and 10 were due to overextension. In Example 9, the boy coined the
terms xiaowangguo ‘small kingdom’ for the kingdom with dwarfs. In Example 10,
he overextended the rain umbrella for the sun umbrella.
9. Child: 是小王國
%pho: sí ɕiaǔ uàŋ kuò
%eng: It is a small kingdom.
Mother: 小人國,不是小王國
%pho: ɕiaǔ zəǹ kuò, pù sí ɕiaǔ uàŋ kuò
%eng: It is a country with dwarfs, not a small kingdom.
10. Child: 這張媽怎麼沒有撐雨傘,熱死了
%pho: tsə́ tsāŋ mā tsə̌ n mo meі̀ tshəŋ y̌ sǎ n, z̨ə́ s1̌ lə
%eng: Why didn’t you use an umbrella? It was hot.
Mother: 因為要照相所以我沒有撐雨傘
%pho: і̄ n ueі́ iaú tsaú ɕiaŋ́ suǒ і̌ uǒ meі̀ iǒ u tshəŋ y̌ sǎ n
%eng: Because I will take the picture, I don’t use the umbrella.
Mother: 這叫陽傘,那個,撐太陽,擋太陽的叫陽傘
%pho: tsə́ tɕiaú iàŋ sǎ n, ná kə tshəŋ ́ thaі́ iàŋ, tǎ ŋ thaі́ iàŋ tə ɕiaú iàŋ sǎ n
%eng: This is called a sun umbrella. It is used to block the sunshine.
3.4.2 Verb Acquisition
The verbs acquired by Indonesian children (mean = 48.44) were fewer than those
by Taiwanese children (mean = 61.6). The verbs are divided into weather verbs,
stative verbs, change of state verbs, body function verbs, motion verbs, position
verbs, action verbs, cognition verbs, perception verbs, emotion verbs, communi-
cation verbs, and social interaction verbs. Both groups produced mostly action
verbs, which constitute 47.24 % of the verbs by Indonesian children and 43.82 %
of Taiwanese children. The details are shown in Table 3.2. The first column indi-
cates the semantic category and an example below. The first row indicates the
nationality of the participant’s mother and the number of participants. The per-
centage shows the proposition of the semantic category among the children’s verbs.
The last row gives the mean number of verbs produced by each group.
Most of the errors of verb use were phonological errors, which have been
discussed in the last chapter, so will not be discussed again here. An example from
the Taiwanese 2-year-olds was given in Example 11. Most of the semantic errors
were due to Taiwanese interference. For example, the Taiwanese verb 惜惜 was
used by both the three- and six-year-olds.
Table 3.2 Semantic Indonesian N = 9 Taiwanese N = 10

categories of verbs produced
in child spontaneous speech Weather verbs
Rain (%) 0.83 0.15
Stative verbs
Sleep (%) 12.38 11.63
Change of state verbs
Disappear (%) 6.46 6.63
Body functions
Throw up (%) 0.69 0.93
Motion verbs
Walk (%) 9.02 9.08
Position verbs
Sit (%) 1.23 1.54
Action verbs
Play (%) 47.24 43.82
Cognitive verbs
Know (%) 5.51 5.78
Sensation verbs
See (%) 2.11 4.27
Emotion verbs
Like (%) 5.01 4.16
Communication verbs
Say (%) 3.13 4.65
Social interaction verbs
Help (%) 6.40 7.36
Mean 48.44 61.6
(11) Mother: 你要幫我撕嗎?要嗎?

%pho: nі̌ iaú paŋ̄ uǒ sі̄ mā ? iaú mā ?
%eng: Do you want to tear it for me? Do you?
Child: 我也掌要。
%pho: uǒ iiè tʂ*aŋ̌ iaú
%eng: I want to…
3.4.3 Adjective Acquisition
Indonesian children (mean = 18.22) also acquired fewer adjectives than Taiwanese
children (mean = 22.70). According to Blackwell (2005), we divided the adjectives
into color, evaluative, age, human propensity, physical property, and quantity.
Indonesian children acquired most evaluative adjectives (31.21 %) and Taiwanese
children acquired most physical properties (32.83 %). Table 3.3 presents a summary
of adjectives produced by the two groups. The first row indicates the nationality of
Table 3.3 Semantic Indonesian N = 9 Taiwanese N = 10

categories of adjectives
produced in child Color
spontaneous speech Red (%) 23.64 13.84
Value
Good (%) 31.21 26.48
Age
Old (%) 2.33 1.97
Human propensity
Happy (%) 3.66 14.13
Physical properties
Small (%) 30.96 32.83
Quantity
Many (%) 8.21 10.85
Mean 18.22 22.70
the participant’s mother and the number of participants. The percentage shows the
proposition of the semantic category among the children’s adjectives. The last row
gives the mean number of adjectives produced by each group.
3.4.4 Classifier Acquisition
3.4.4.1 Spontaneous Speech
Few classifiers were produced in the spontaneous speech. The first classifier use
appeared at age two for both groups. The two-year-old Indonesian girl produced only
one classifier—ge, while the two-year-old Taiwanese girl produced four classifiers—
ge, zhang, shao, and ben. Indonesian children on average produced 2 classifiers and
Taiwanese children produced 2.7 classifiers. The four-year-old Taiwanese boy pro-
duced most types of classifiers—zhang, ge, tiao, ben zhong, and ba. The
three-year-old Indonesian boy produced most tokens, but 31 out of the 34 were ge.
A list of classifiers occurred in the speech sample was given in Table 3.4. The
first column lists the classifiers. The first row lists the mothers’ nationality. The
numbers indicate the mean tokens of classifiers observed. The last row gives the
mean number of classifiers produced by each group.
3.4.4.2 Classifier Elicitation Task
Since classifiers did not occur a lot in spontaneous speech, we conducted a classifier
elicitation task (Kuo 1999) to elicit classifier production of the same group of children
in the second year. Therefore, the participants were one year older than in the spon-
taneous speech. The age of the participants was three, four, five, six and seven.
Table 3.4 Classifiers Classifier Indonesian N = 9 Taiwanese N = 10

produced in child
spontaneous speech 張 zhang 0.11 2.6
個 ge 9.22 10.4
艘 sou 0 0.6
本 ben 0.11 1.4
隻 zhi 0.67 2
台 tai 0.11 0
顆 ke 1.11 0.2
條 tiao 0 0.4
班 ban 0 0.2
碗 wan 0 0.4
Total tokens 11.33 18.2
Total types 7 10
Twelve pictures were specially designed to elicit the production of shape clas-
sifiers in Mandarin Chinese with two or more of the target objects present in the
pictures. The target objects were the referents of the collocating nouns of the target
classifiers. The target objects appeared with things that they often occur with in a
natural context. The presence of two or more target objects creates an obligatory
context for the use of classifiers (Kuo 1999).
Eight categories are identified as relevant cognitive categories in Chinese clas-
sifiers—material, shape, consistency, size, location, arrangement, quanta, and
attributes of parts (Tai 1994). The first seven categories were from Allan (1977)
based on the use of classifiers in more than fifty languages. The material category
has three subcategories: animacy, inanimacy, and abstract and verbal nouns. Shape
is divided into saliently one-dimensional, saliently two-dimensional, and saliently
three-dimensional. Consistency consists of three subtypes: flexible, hard or rigid,
and non-discrete. Size means big or small. Location refers to the place where the
object exists. Arrangement refers to the relation among members that a collective
noun denotes. Quanta refers to the function of quantification. Attributes of parts are
proposed as a basis for human categorization by Tversky and Hemenway (1984).
Shape classifiers were chosen because they are semantically more productive in a
sense that they can classify a larger number of objects on semantic bases (Loke
1996). In addition, shape is determined by visual perception. Also, shape is easier to
manipulate than other classifications.
The participants were tested individually at their homes. They were asked to
describe pictures provided by the researcher including naming and counting the
items in the pictures. After the instructions of the task, the experimenter demon-
strated by orally describing a picture. To make sure that the participant understood
the procedure, one picture was given as a warm-up. The participant was given
immediate feedback. Assistance without providing the classifiers was allowed. The
total procedures took about 15 min for each child. The process was tape-recorded
and transcribed.
Table 3.5 Means of children CL Indonesian Taiwanese

of Indonesian and Taiwanese
mothers in classifier ge (3D) (%) 93.83 87.5
elicitation task ke (3D) (%) 5.56 15
zhang (2D) (%) 15.56 28
pian (2D) (%) 11.11 20
tiao (1D) (%) 3.03 9.09
gen (1D) (%) 0 30
zhi (1D) (%) 14.81 20
jian (%) 16.67 30
kuai (%) 0 0
Mean 25.43 32.17
The results show that the children of Taiwanese mothers did a bit better than
children of Indonesian mothers. The mean accuracy rate for the Indonesian group is
25.43 %, and the mean accuracy rate for the Taiwanese group is 32.17 %. Their scores
for each classifier are given in Table 3.5. 1D stands for saliently one-dimensional, 2D
for saliently two-dimensional, and 3D for saliently three-dimensional.
Both groups did best for ge, with Indonesian even higher than Taiwanese.
Indonesian children used ge almost across the board, which may lead to the high
accuracy rate for ge. Here is an example of classifier errors. Example 12 was
produced by the four-year-old Indonesian boy, who used jiao ‘feet’ for pants and
then changed to ge failing to utter the conventional tiao.
(12). *INV: 有多少[*]長[*]褲呢?
%pho: iŏu tuō s[*]ău tsh[*]àŋ khú nə
%eng: How many trousers are there?
*CHI: 兩個腳…腳
%pho: liăŋ kə tɕiău tɕiău
%eng: Two…feet..feet.
*INV: 這[*]是[*]腳沒錯…可是[*]這[*]是[*]長[*]褲阿,對不對?有多少[*]
長[*]褲?
%pho: ts[*]ə́ ṣ́[*] tɕiău mèi tshuó khə̆ ṣ́[*] ts[*]ə́ ṣ́[*] tsh[*]àŋ khú a tuéi pù
tuéi iŏu tuō s[*]ău tsh[*]àŋ khú
%eng: Feet….these are trousers, right? How many trousers are there?
*CHI: 一個長[*]褲
%pho: i kə tsh[*]aŋ̀ khú
%eng: one ge trousers
3.4.4.3 Classifier Comprehension Task
A classifier comprehension task (Sera et al. 2013) was also given in the second
year. Four classifiers tiao 條, ge 個, zhi 枝, and zhi 隻 were chosen for the classifier
comprehension task because they are acquired early and have high agreement
among adults (Batteen et al. 2007). The task consisted of 18 unique trials, 3 shape
and 3 kind trials, for each classifier (ge, zhi, and tiao). All shape trials consisted of
three line drawings of non-living things (e.g., an apple, a crayon, and a scarf). All
kind trials consisted of three drawings that depicted living things (e.g., a girl, a bird,
and a snake). Structuring the trials in this fashion avoided potential problems
associated with the homophones of zhi. A total of nine shape and nine kind trials
were presented to each participant, with each different classifier being queried three
times. So, for example, a child would be presented with drawings depicting a ball, a
pencil, and a belt three times. One time the child would be asked to ‘Give me one
ge,’ the other time they would be asked to ‘Give me one zhi,’ and another time they
would be asked to ‘Give me one tiao.’ The pictures depicting each item were
slightly different for each query.
The Indonesian children on average answered 63.25 % correctly, which is lower
than the mean of children with Taiwanese parents (75.08 %). Taiwanese
3-year-olds’ responses were only reliably better than chance on the shape trials of
zhi (t(15) = 4.4, p < 0.001) and the kind trials of tiao (t(15) = 2.5, p < 0.05). The
responses of all other age groups were well-above chance on all trial types. (Sera
et al. 2013). On the other hand, Indonesian children did not perform above chance
till seven years old. There were also large individual differences. The five-year-old
girl did better than the six-year-old girl. The results of individual Indonesian par-
ticipant’s performance in classifier comprehension task are given in Table 3.6. The
first column indicates the target classifiers. The top row lists the participants’ age
and gender. In order to trace the data from the first year, +1 was used to indicate the
second-year data. For example, 2 + 1♀ means the three-year-old girl.
Children’s acquisition of classifiers is guided by both cognitive principles and
frequency. Shape classifiers were first acquired by children and then were followed
by classifiers of high frequency such as jian for clothing and nouns that are
semantically vague such as thing. The first cognitive basis they acquired is shape. In
Table 3.6 Results of classifier comprehension task

I2 + 1 I3 + 1 I3 + 1 I4 + 1 I4 + 1 I5 + 1 I5 + 1 I6 + 1 I6 + 1
♀ ♀ ♂ ♀ ♂ ♀ ♂ ♀ ♂
Tiao 條 1 4 1 6 6 3 6 6 6
Ge 個 5 5 3 6 4 5 9 6 8
Zhi 枝/隻 4 6 3 9 7 4 8 9 8
Total 10 15 7 21 17 12 23 21 22
% 38.46 57.69 26.92 80.77 65.38 46.15 88.46 80.77 84.62
Total 148
% 63.25
Mean 16.4
addition to the overextension of the general classifier ge, children’s classifier errors
did not violate shape principles. They used a classifier for nouns denoting objects of
the same shape but different material.
It seems that the vocabulary growth by children of Indonesian immigrants is slower

than children of native Mandarin-speaking Taiwanese mothers. The finding is
different from that of previous studies on lexical development of children of
immigrants, whose lexical development is not different from native children (Tsay
et al. 2005; Kuo 2008). Tsay et al. (2005) included mothers from mainland China,
whose Mandarin is native just like Taiwanese mothers. They may have messed the
results. Kuo’s (2008) findings about no differences between children of Vietnamese
mothers could be due to the fact that Vietnamese has more Chinese-based lexicon
than Indonesian. Thus, innateness perspective alone could not explain lexical
development.
Imitation was not observed in the data and word coinage was observed instead;
thus, behaviorism was rejected. Meaning correction was observed only in native
mothers or investigators. Indonesian mothers were less capable of providing such
feedback with limited Mandarin proficiency. Interactionism may be the best theory
for explaining lexical development.
References
Akhtar, N. (2002). Relevance and early word learning. Journal of Child Language, 29, 677–686.
Allan, K. (1977). Classifiers. Language, 53, 285–311.
Alves, M. (1999). What’s so Chinese about Vietnamese? The Ninth Annual Meeting of the
Southeast Asian Linguistics Society. University of California, Berkeley.
Anglin, J. (1995). Classifying the world through language: Functional relevance, cultural
significance, and category name learning. International Journal of Intercultural Relations, 19,
161–181.
Bates, E., Marchman, V., Thal, D., Fenson, L., Dale, P., Reznick, S., et al. (1994). Developmental
and stylistic variation in the composition of early vocabulary. Journal of Child Language, 21,
85–123.
Batteen, C., Davis, N., Fuller, J., Kuo, J. Y.-C., Li, D., Sera, M., et al. (2007). Classifier use by
adult Mandarin speakers. The 5th Conference of the European Association of Chinese
Linguistics. Leipzig, Germany. September 4–7.
Beeghly, M., Bretherton, I., & Mervis, C. (1986). Mothers’ internal state language to toddlers: The
socialization of psychological understanding. British Journal of Developmental Psychology, 4,
247–260.
Bornstein, M., & Cote, L. (2004). Cross-linguistic analysis of vocabulary in young children:
Spanish, Dutch, French, Hebrew, Italian, Korean, and American English. Child Development,
75, 1115–1139.
References 129
Chien, Y.-C., Lust, B., & Chiang, C.-P. (2003). Chinese children’s comprehension of
count-classifiers and mass-classifiers. Journal of East Asian Linguistics, 12, 91–120.
Choi, S., & Gopnik, A. (1995). Early acquisition of verbs in Korean: A cross-linguistic study.
Journal of child language, 22, 497–529
Clark, E. (1987). The principle of contrast: A constraint on language acquisition. In B.
MacWhinney (Ed.), Mechanisms of language acquisition. Hillsdale, NJ: Erlbaum.
Clark, E. (1993). The lexicon in acquisition. Cambridge, UK: Cambridge University Press.
Erbaugh, M. S. (1986). Taking stock: The development of Chinese noun classifiers historically and
in young children. In C. Craig (Ed.), Noun classes and categorization (pp. 399–436).
Amsterdam: John Benjamins Publishing Company.
Fang, F. (1985). An experiment on the use of classifiers by 4-to-6-Year-Olds. Acta Psychologica
Sinica, 17, 384–392.
Fernald, A., & Mazzie, C. (1991). Prosody and focus in speech to infants and adults.
Developmental Psychology, 27, 209–221.
Gelman, S., Coley, J., Rosengran, K., Hartman, E., & Pappas, A. (1998). Beyond labeling: The
role of maternal input in the acquisition of richly structured categories. Monographs of the
Society for Research in Child Development, 63(253), 1.
Gentner, D. (1981). Some interesting differences between verbs and nouns. Cognition and Brain
Theory, 4(2), 161–178.
Gentner, D. (1983). Nouns and verbs. Symposium Presented at the Meeting of the New England
Child Language Association. Medford, MA: Tufts University.
Gentner, D. (1988). Cognitive determinism: Object reference and relational reference. Paper
presented at the Boston University Child Language Conference, Boston, MA.
Gentner, D. (1999). Individuality and early word meaning. Paper presented at the VIIIth
International Congress for the Study of Child Language, San Sebastian, Spain.
Gleason, J. B., & Ratner, N. B. (2012). The development of Language (8th ed.). Cambridge:
Pearson.
Goldfield, B. (1993). Noun bias in maternal speech to one-year-olds. Journal of Child Language,
20, 85–99.
Golinkoff, R., Shuff-Bailey, M., Olguin, R., & Ruan, W. (1995). Young children extend novel
words at the basic level: Evidence for the principle of categorical scope. Developmental
Psychology, 31, 494–507.
Hall, D. (1994). How mothers teach basic-level and situation-restricted count nouns. Journal of
Houston-Price, C., Plunkett, K., & Harris, P. (2005). “Word learning wizardry” at 1;6. Journal of
Hu, Q. (1993). The acquisition of Chinese classifiers by young Mandarin-speaking children. Ph.D.
Thesis, Boston University.
Huang, Y. V. (2007). The use of count and mass classifiers in Chinese preschoolers. MA thesis,
National Taiwan Normal University, Taipei.
Jiang, Y.-W. (2000). Basic-level effects in Chinese lexicon. MA thesis, National Tsing Hua
University, Hsinchu.
Kuo, J. Y.-C. (1999). Strategies for learning classifiers. In Proceedings of the 1999 Second
Language Research Forum, pp. 424–442.
Kuo, J. Y.-C. (2008). Mandarin acquisition by children of Vietnamese mothers in Taiwan. Taipei:
Crane.
Kuo, J. Y.-C. (2015). Lexical development by children of Thai immigrants in Taiwan. Studies in
English Language and Literature, 35, 1–12.
Loke, K. K. (1996). Norms and realities of Mandarin shape classifiers. Journal of Chinese
Language Teachers Association, 31(2), 1–22.
Markman, E. (1987). How children constrain the possible meanings of words. In U. Neisser (Ed.),
Concepts and conceptual development: Ecological and intellectual factors in categorization.
Cambridge, UK: Cambridge University Press.
Markman, E., & Wachtel, G. (1988). Children’s use of mutual exclusivity to constrain the
meanings of words. Cognitive Psychology, 20, 121–157.
Myers, J. (2000). Rules vs. analogy in Mandarin classifier selection. Language and Linguistics, 1
(2), 187–209.
Poulin-Dubois, D. (1995). Object parts and the acquisition of the meaning of names. In K. Nelson
& Z. Reger (Eds.), Children’s language (Vol. 8). Hillsdale, NJ: Erlbaum.
Sachs, J., Brown, R., & Salerno, R. (1976). Adults’ speech to children. In W. won Raffler Engel &
Y. Lebrun (Eds.), Baby talk and infant speech. Lisse, The Netherlands: Swets and Zeitlinger.
Sera, M. D., Johnson, K., & Kuo, J. Y.-C. (2013). Classifiers augment and maintain shape-based
categorization in Mandarin speakers. Language and Cognition, 5(1), 1–23.
Sneddon, J. N. (2004). The Indonesian Language: Its history and role in modern society. Sydney:
UNSW Press.
Storkel, H. (2004). Do children acquire dense neighborhoods? An investigation of similarity
neighborhoods in lexical acquisition. Applied Psycholinguistics, 25, 201–222.
Tai, J. (1994). Chinese classifier systems and human categorization. In M. Chen & T. Ovid (Eds.),
In honor of professor William S-Y. Wang: Interdisciplinary studies on language and language
change (pp. 479–494). Taipei: Pyramid.
Tardif, T. (1996). Nouns are not always learned before verbs: Evidence from Mandarin speaker’s
early vocabulary. Developmental Psychology, 32, 492–504.
Tardif, T., Gelman, S., & Xu, F. (1999). Putting the “noun bias” in context: A comparison of
English and Mandarin. Child Development, 70, 620–635.
Tingley, E., Gleason, J., & Hooshyar, N. (1994). Mothers’ lexicon of internal state words in speech
to children at mealtime. Journal of Communication Disorders, 27, 135–155.
Tomasello, M. (1995). Pragmatic contexts for early verb learning. In M. Tomasello & W.
Merriman (Eds.), Beyond names for things: Young children’s acquisition of verbs. Hillsdale,
NJ: Erlbaum.
Tsay, S., Ke, H., Tsai, J., Lin, H., & Kuo, Y. 蔡素娟、柯華葳、蔡志賢、林惠玲、郭怡君.
(2005). 新移民女性及其子女的發展與適應研究-新移民女性之子女的語言發展研究 I
“Xinyimin nuxing ji qi zinu de fazhan yu shiying yanjiu—xinyimin nuxing zhi zinu de
yuyan fazhan yanjiu” [Language development by children of new female immigrants]. NSC
report. (NSC 93-2420-H-194-003-KD)
Tversky, B., & Hemenway, K. (1984). Objects, parts and categories. Journal of Experimental
Chapter 4
Syntactic Development
Three major theories have been proposed for syntactic development: semantic
bootstrapping, syntactic bootstrapping (Valian 1986, 1991; Radford 1990), and
distributional learning (Maratsos 1988). The first two focus on innate mechanism.
Semantic bootstrapping matches syntactic structures to innate ontological cate-
gories. Syntactic bootstrapping assumes the innate universal grammar. These two
do not predict differences between children of native parents and immigrant parents.
However, distributional learning predicts delay in the development with limited
input in children of immigrants. Many problems have been found in semantic
bootstrapping (Mills 1986; Karmiloff-Smith 1979; Pérez-Pereira 1991; Levy 1988;
Bowerman 1990). On the other hand, distributional learning is supported in several
studies (Elman 1990; Finch and Chater 1992; Plunkett and Marchman 1991). The
goal of this chapter is to examine whether there is a difference in the syntactic
development between children of Taiwanese and Indonesian mothers. It will have
implication for language acquisition theories.
In the present study, syntactic development was measured by length (mean
length of utterance), sentence structures (Index of Productive Syntax), and a
question elicitation task. Since these measurements were first developed with
English-speaking children, I will first review this literature followed by literature on
Mandarin syntactic development.
4.1 General Literature on Syntactic Development
4.1.1 Measuring Syntactic Development by Length—MLU
One of the popular methods to measure the syntactic development of children is

mean length of utterance (MLU), introduced by Brown (1973). MLU is the average
length of a child’s utterances in the spontaneous speech based on the number of

DOI 10.1007/978-981-10-1035-4_4
132 4 Syntactic Development
morphemes. Brown (1973) divided the language development of English-speaking

children into five stages by mean length of utterance: MLU in Stage I begins from
1.0 to 2.0 and the range of MLU in the following stages increased by 0.5. Thus,
Stage II is from 2.0 to 2.5; Stage III goes from 2.5 to 3.0; Stage IV is between 3.0
and 3.5; and Stage V goes from 3.5 to 4.0. In Stage I, as children acquire English as
their native language, they usually produce two-word utterances, but some may
have outstanding performance and can produce longer sentences which include
three or four words. These two-word utterances produced by young children were
simple and composed of content words primarily such as nouns, verbs, and
adjectives. Function words such as prepositions, conjunctions, and articles rarely
occur in the speech of young children at Stage I. Thus, the two-word utterances are
called telegraphic speech (Brown and Fraser 1963) since mostly these sentences
consist of content words, which makes them resemble telegrams, for example,
mommy come, drive car, and box shiny. Gleitman and Wanner (1982) argued that
content words were acquired earlier than function words by young children since
their perceptual and prosodic features are so salient that help the children discover
basic language structures.
During Stage II, the MLU of children grows from 2.0 to 2.5 when the gram-
matical morphemes begin to appear; that is, young children start to use bound
inflectional morphemes in the sentences according to syntactic structure, for
example, -ed as a suffix to a verb to indicate past tense and -s for present tense when
the subject is third person.
Measuring MLU in Chinese has been done with syllables or words. The present
study measured MLU by words (MLUw), which was has been found to be a more
reliable index in Chinese (Cheung 1998).
4.1.2 Acquisition of Sentences Types
After Stage II, the MLU goes from 2.5 to 3.0 when the different types of sentences
such as negation, questions, and passive sentences are developed gradually by
young children since they attempt to acquire the sentence structures from what they
hear in the daily life in order to express themselves. According to Bassano and
Mendes-Maillochon (1994), children developed their different sentence types based
on different intonation patterns, which closely matched those used by adults to mark
different sentence modalities.
4.1.2.1 Negation
The development of negation reflects a complex interrelation of syntactic, semantic,

and input factors that may influence children in the early stages in the path of
reaching the full mastery of sentence structures.
4.1 General Literature on Syntactic Development 133
Bellugi (1967) studied the development of negation from longitudinal transcripts

of children. She discovered three main periods in the acquisition of the structure of
negation. In the first period, children usually put the negative marker, no or not in
the initial position of the sentence to express negation. For example, No go movies
and No Mommy do it. Next, children attempt to place the negative words next to the
main verb instead of preceding the whole sentence, but the use of the auxiliary is
not so productive, e.g., I no like it and I no want book. In the final period, the
sentence structure of negation produced by young children is more similar to that
used by adults with the appearance of auxiliaries such as You can’t have this and
I’m not sad now, which is not usually reached until Stage V. We can see that young
children in this stage seem to develop the knowledge of combination of an auxiliary
and a negative word. Since auxiliaries in English are more complex, children spend
more time reaching the mastery of negation in English.
Another study was conducted by de Villiers and de Villiers (1979), who
observed the development of negation of their own children and found two varieties
of negative sentences during the first period: no at sentence initial and sentence
internal negation. The first form, no + sentence, is to describe rejection, while the
internal no form to express denial. However, this generalization failed to account
for some of Bellugi’s data.
4.1.2.2 Questions
There are three ways of forming questions in English, rising intonation with a
statement, yes/no questions, and wh-questions. Children at earlier stage seem to use
rising intonation to express questions (Klima and Bellugi 1966). Yes/no question
appears much later about in Stage III, because speakers need to invert the subject
and the auxiliary verb. Wh-question is used for obtaining more important infor-
mation instead of a yes or no answer, so speakers need to place wh-words such as
what, who, where, and when in the initial position to form a question in addition to
subject auxiliary inversion, for example, What should we do now? or When does the
show start? In the first period, children usually omit the auxiliary when asking wh-
questions, e.g., What that? or Where Mommy go? Next, they start to insert the
auxiliary into the structure of questions but do not reach the full mastery of
expression of questions since they sometimes do not invert the subject and the
auxiliary like Where he is going? The wh-questions can become more complicated
when they are formed from complex sentences which may undergo long-distance
movement across sentences. However, children are able to incorporate all of the
syntactic rules necessary to produce well-formed wh-questions with highly complex
rules (De Villier et al. 1990).
Klima and Bellugi (1966) claimed that children can produce correct forms of
yes/no questions earlier than produce wh-questions since there is only one rule in
yes/no question in which the subject of the sentence and auxiliary verb should be
switched and wh-questions involve two rules to complete the sentence patterns: the
placement of wh-words and inversion of subject and auxiliary.
Wootten et al. (1979) explored the order of acquisition of the various wh-
questions by children and found that the sentence patterns beginning with wh-words
such as what, where, and who were acquired first, but questions about when, how,
and why were used later. Also, the same order was found in the comprehension
study conducted by Winzemer (1980) since the concepts of how, when, and why are
about manner, time, and causality, which are more abstract.
During the preschool years, Stage IV (MLU = 3.0–3.5), more complex gram-
matical structures begin to be produced by preschoolers such as passive con-
struction, coordinations, and relative clauses; however, their acquisition of certain
sentence structures does not reach the full mastery until some years later.
4.1.2.3 Passive
The structure of passive is for focusing on the recipient of an action, for example,
‘John was hit by a ball.’ which is a full passive sentence, which focuses on the
person, John. The other passive form without an agent specified in the sentence is
called truncated passive, e.g., ‘The vase is broken.’ Not surprisingly, passives
rarely occur in children’s spontaneous speech. Thus, if researchers would like to
investigate the development of structure of passive, they have to develop a task to
elicit these passive patterns in an experiment. In addition, because the order of the
agent and the object is reversed in passives in English, this particular construction
can reveal a great deal about how children acquire word order rules that play a
major role in English syntax.
Horgan (1978) elicited English-speaking children’s production of passives and
discovered that the truncated passives were produced more than full passives by
children in the earlier period. In addition, she found that the subjects in full passives
produced by young children are animate, while those in truncated passives are
inanimate subjects. Thus, Horgan concluded that full and truncated passives were
acquired separately by children.
Bever (1970) proposed that young children around three or four generalize the
order of words in English passives based on word order strategy in which English
use noun–verb–noun sequences in the active voice meaning agent–action–object.
Thus, when they hear a passive sentence, they ignore the copula and the word, by
and infer the meaning of the passive noun–verb–noun sequence to be active.
However, other studies which explored children learning languages other than
English showed that this strategy is not universal. It was indicated that passive
construction was acquired much earlier by non-Indo-European language-speaking
children than English-speaking children. Children in southern Africa used passives
in daily conversation by age two since they mostly marked the topic of a sentence in
the conversation (Pye 1988; Suzman 1987).
4.1 General Literature on Syntactic Development 135
4.1.2.4 Coordination
During Stage IV, young children attempt to combine two or more sentences to form
the expression of complex or compound proposition by simply using and. There are
two main forms of coordination: sentential coordinations, in which two or more
whole sentences are conjoined by means of conjunctions, e.g., ‘Mary is singing and
Mary is dancing,’ and phrasal coordinations, in which only phrases in the sentences
are combined, e.g., ‘Mary is singing and dancing.’ Thus, some researchers were
interested in the order of these two coordinations in the development of language of
children. Bloom et al. (1980) studied four children longitudinally and found that for
three of them, both forms appeared in their speech in the same period, but one child
produced phrasal coordination before sentential coordination. In addition, they
discovered that the development of coordination was influenced by semantic fac-
tors. They found that their subjects used conjunction and to express a variety of
meanings, and these meanings were developed in a fixed order. First, they devel-
oped the meaning of additive, which showed no dependency relation between the
conjoined clauses, e.g., ‘Mary is singing and Carrie is dancing.’ Few months later,
they began developing coordination which showed temporal relations, e.g., ‘She
goes home and takes a shower.’ After that, they used and to encode causal relations,
e.g., ‘Go straight on and you’ll see the church.’ Therefore, we can see that coor-
dinations encode a variety of meanings and children use conjunction word, and, in a
semantically limited way and develop in a fixed order.
4.1.2.5 Relative Clauses
In Stage IV, embedded sentences such as relative clauses begin to appear in young
children’s speech, but the full mastery of these constructions is not reached until
Stage V. In Bloom’s (1980) study, after development of coordination, relativization
began to occur in young children’s speech, but scarcely. Therefore, Hambruger and
Crain (1982) and Tager-Flusberg (1982) designed an elicitation task for relative
clauses and found that object relatives were produced first by children at age four.
For example, ‘John met a girl who wears a pink dress.’ The grammatical structure
which adds a clause at the end of a sentence is much easier than that which is
interrupted by a clause in the middle of a sentence.
4.1.3 Measuring Syntactic Development by Sentence Types
In addition to MLU, I used a Mandarin version of Index of Productive Syntax (Lin

2009) to evaluate children’s syntactic growth based on The Index of Productive
Syntax (IPSyn) developed by Scarborough (1989), which measures the occurrence
of various grammatical forms in English. It investigates what kinds of grammatical
structures the young child use, but the misused grammatical forms are not taken for
examination. Thus, the IPSyn serves as a measure of grammatical types instead of

grammatical tokens by using the score sheet to mark the use of grammatical forms
which occur up to twice. There are four subscales on the sheet: noun phrases (NP),
verb phrases (VP), questions and negations (Q/N), and sentence structures (SS).
The IPSyn score is made from counting the total number of the points of the items
in each syntactic category.
Scarborough (1989) claimed that although both IPSyn and MLU were significant
in the ages between 2 and 3 , the scores in IPSyn became more significant between
beyond 3 years of age. IPSyn scores were highly affected by corpus size since
IPSyn is a measure of types in language development. The scores of the age
differences across five ages were significant in the smaller corpus size as well as in
the corpus of 100 utterances.
Scarborough (1989) also found the differences by age in the scores of four
subscales in the IPSyn. When children were between age 2 and 3, there was a
significant increases in the all four categories. After 3 years of age, the development
of verb phrases and sentence structures grew significantly, while the other two
subscales, noun phrases and questions, did not significantly increase until children
were between 3;5 and 4 years of age.
The reliability in two measures, IPSyn and MLU, was high. Although the IPSyn
has correlation with MLU, IPSyn can measure the syntactic development of chil-
dren when their MLU goes up over 4.0, which is the limitation of MLU. For
validity, IPSyn has been found to be used for examining age differences in the
language samples collected longitudinally for four years, while the other measure,
MLU, failed to perform as IPSyn did (Scarborough 1989). Therefore, the IPSyn not
only can provide a summary scale of the development of syntactic complexity of
individual differences in age changes but can also serve as a reliable measure for
language development.
4.2 Mandarin Syntactic Development by Native Children
Hsu (1996) proposed the stages of the syntactical development in Mandarin similar
to the one in English proposed by Brown (1973). In Stage I, young children around
1;0–1;6 produced one-word utterances that were usually composed of nouns, for
example, 車車 che che ‘a car,’ and since young children in this stage did not have
enough syntactic knowledge and vocabulary, they began to use the same word with
different intonation contours to convey their expression. In addition, Hsu general-
ized that the participants in his study tried to select the word conveying the
important information, e.g., 帽 mao for maozi ‘a hat.’ In Stage II, simple sentences
began to appear in children’s speech, e.g., 超人給你 chaoren gei ni ‘superman give
you’ ‘(I) give you the superman.’ During Stage III, Hsu found that children at age
two started to produce embedded sentences in the speech, e.g., 這是小阿姨綁的
zhe si xiao a yi bang de ‘little aunt tie de’ ‘This is tied by my little aunt,’ in which
de is a relativization marker. After two years at Stage IV, they began to produce
4.2 Mandarin Syntactic Development by Native Children 137
longer sentences. Hsu divided compound sentences used in Stage IV into two kinds.
First, children around 2;7–4;2 produced compound sentences without any con-
junctions, e.g., 你給我,我就給你 ni gei wo, wo jiu gei ni ‘(If) you give me (one
thing), I’ll give you (another thing).’ The second kind of compound sentences with
conjunctions appeared after age four, e.g., 如果你認識我弟弟,就行了 Ru kuo ni
ren shi wo di di, jiu xing le ‘If you know my younger brother, that’s fine.’
Comparing with the development of English-speaking children, we can see that the
syntactic development of Chinese-speaking children is faster since Chinese does
not have grammatical morphemes like English. Therefore, after Stage I, Chinese
children directly developed simple sentences instead of grammatical morphemes
like English-speaking children. In addition, negations and questions do not involve
any inversion of subjects and auxiliaries in Mandarin, so children around two
attempt to produce embedded clauses; however, for English-speaking children,
embedded sentences appeared at age three. Thus, we can see that Chinese children
develop syntactic constructions much earlier than English children do. Since
embedded sentences appear earlier in children’s speech, some researchers continued
to explore the mastery of complex sentences, which is an important developmental
milestone in language acquisition (Chang 1984; Cheung 2006).
Chang (1984) explored the comprehension of different types of relative clauses
by children in grade 1, 2, 4, and 6. There are four types of relative sentences
depending on the grammatical function of the head noun in the main sentence and
relative clause: SS, SO, OS, and OO. S stands for subject and O for object. She
found that SS and SO were much easier to be produced by children than OS and
OO since in Chinese, SS and SO contain no interruption in the main clause, while
the main clauses of OS and OO are interrupted.
Cheung (2006) studied longitudinally the order of development of different types
of complex sentences by four children around two to four. In his study, relative
clauses, serial verb constructions, descriptive clauses, pivotal constructions, and
subject/object complement sentences were investigated. He found that generally
serial verb constructions and descriptive clause appeared earliest among other types
of complex constructions by children at age two. Several months later, children
began to produce pivotal constructions and relative clauses. After the development
of relative clauses, subject/object complements were developed much later and
object complement appeared in children speech earlier than subject complements.
4.3 Mandarin Syntactic Development by Children

of Vietnamese Mothers
Kuo (2008) collected one-hour spontaneous speech of ten children of Vietnamese

and Taiwanese aged 2–6 and measure their syntactic development by mean length
of utterance and adapted version of Index of Productive Syntax (Scarborough 1989)
adding some peculiar structures in Mandarin. There were few differences between
Table 4.1 MLU of children Age (years) Vietnamese Taiwanese

of Vietnamese and Taiwanese
by age 2 2.84 2.16
3 2.69 2.73
4 4.37 3.88
5 3.78 3.87
6 3.62 6.37*
Mean 3.46 3.16
STDEV 0.70 0.86
children of Vietnamese and Taiwanese mothers in the MLU for age 2–5. The
difference showed up at the age of 6. The MLU for the six-year-old of Taiwanese
mother reached 6.37, while the Vietnamese age-mate remained 3.62 as shown in
Table 4.1.
In addition, Scarborough’s (1989) Index of Productive Syntax was adapted with
particular structures in Mandarin to measure children’s syntactic growth. It was
found that the sentence patterns used by children of Vietnamese mothers were not
fewer than those used by children of Taiwanese mothers as shown in Table 4.2.
However, different error patterns revealed. Both groups of children omitted or
misused function words. The younger children tended to omit function words. The
older children tended to misuse function words. Errors in complement selection
were found in the two-year-old of the Taiwanese mother, who used the intransitive
verb buxing ‘cannot’ as a transitive verb. The word order errors were observed only
in children of Vietnamese mothers as in Example 1 from a six-year-old child of a
Vietnamese mother. The letter stands for the mother’s nationality, and the number
stands for the child’s age. For example, V6 stands for a six-year-old of a
Vietnamese mother.
(1) V6: *媽媽買電池給我明天
%pho: mama mai tiàn t́ʂh gě i uǒ miə́ŋ tian
%Pinyin: mama mai dianchi gei wo mingtian
%eng gloss: mother buy battery give me tomorrow
%eng: Mother will buy me batteries tomorrow.
Table 4.2 Index of Age (years) Vietnamese Taiwanese

Productive Syntax by children
of Vietnamese and Taiwanese 2 25 21
3 43 23
4 22 34
5 39 28
6 35 38
Mean 32.8 28.8
St. Dev. 9.11 7.19
4.4 Mandarin Syntactic Development by Children of Indonesian Mothers 139
4.4 Mandarin Syntactic Development by Children

The present study investigated the Mandarin syntactic development of children of

Indonesian and Taiwanese mothers in Taiwan with both spontaneous speech and
elicited production data. We examined the mean length of utterances and Index of
Productive Syntax in spontaneous speech and analyzed their errors.
4.4.1 Spontaneous Speech
We collected one-hour spontaneous speech samples from two, three, four, five, and
six-year-old children of Indonesian mothers and Taiwanese mothers. The partici-
pants were the same as in the study of phonological development in Chap. 2 and
lexical development in Chap. 3. We visited the participants at their homes and
recorded their conversions, transcribed with IPA, and then analyzed the syntax by
length, sentence structures, and syntactic errors. A question elicitation task was
given in the third year for questions are rare in spontaneous speech.
The mean length of utterance was used to measure the length of utterances. The
Indonesian group seemed to delay in the syntactic development. The mean MLU
(M = 2.95) was smaller than the Taiwanese group (M = 3.61), and such difference
showed up as early as age 2 as shown in Table 4.3.
The sentence types were measured by Mandarin Index of Productive Syntax (Lin
2009), which adapted Scarborough’s (1989) Index of Productive Syntax with
particular structures in Mandarin. The ba construction, verb reduplication, aspect
markers (i.e., le, zheng, zhe), and sentence final particles observed in children’s
speech were added. It was found that the sentence patterns used by the Indonesian
group (M = 34.70) were fewer than those used by the Taiwanese group
(M = 39.10) as shown in Table 4.4.
Table 4.3 MLU for children of Indonesian and Taiwanese mothers

Age (years) Indonesian Taiwanese
Gender Male Female Male Female
2 2.28 1.99 2.81 2.16
3 3.14 3.00 4.40 2.73
4 2.33 4.25 3.88 3.75
5 3.76 3.16 2.80 3.87
6 2.64 2.93 6.37 3.30
Mean 2.83 3.07 4.05 3.16
Mean 2.95 3.61
Table 4.4 Mandarin Index of Productive Syntax by children of Indonesian and Taiwanese
Age (years) Indonesian Taiwanese
Gender Male Female Male Female
2 22 23 43 18
3 45 42 28 44
4 19 42 38 42
5 38 40 47 39
6 32 44 46 46
Mean 31.20 38.20 40.40 37.80
Mean 34.70 39.10
However, different error patterns revealed. Both groups of children omitted or

misused function words. The errors found in the young Taiwanese children were
found in older Indonesian children. In the following examples, the letter stands for
the mother’s nationality and the number stands for the child’s age. For example, T3
stands for a three-year-old Taiwanese child. Three-year-old Taiwanese and four-
and five-year-old Indonesian children omitted function words placed in brackets in
the following examples.
(2) T3: 摳摳 [在] 袋袋裡
%pho: khou khou [tsài] tài tài lǐ
%Pinyin: koukou [zai] daidai li
%eng gloss: money [in] pocket inside
%eng: Money is in the pocket.
(3) I4: 阿嬤家 [的] 狗
%pho: ā mà tɕiā [tə] kǒ u
%Pinyin: ama jia [de] gou
%eng gloss: Grandmother home [of] dog
%eng: Grandmother’s dog.
(4) I5: 那 [是] 什麼顏色?
%pho: nà [ʂ] ʂə́ mə ień sə̀
%Pinyin: na [shi] sheme yanse
%eng gloss: That [is] what color
%eng: What color is that?
Like Taiwanese children, Indonesian children transferred functions words from
Taiwanese. For example, the five-year-old Taiwanese transferred the function word
gei ‘give’ from Taiwanese (5). The six-year-old Indonesian transferred yong ‘use’
from Taiwanese. (6) Children of Vietnamese also did the same thing (Kuo 2008)
due to the exposure of Taiwanese.
(5) T5: 他給 *[把] 我的腳拿開1

%pho: thā kě i pǎ uɔ̌ tə tɕiǎ u ná khā i
%Pinyin: ta gei ba wode jiao nakai
%eng gloss: He give *[ba] I de foot take away
%eng: He pushed my leg away.
(6) I6: 用 *騎摩托車
%pho: yúŋ tɕh1́ ḿɔ thuɔ tʂə̄
%Pinyin: yong qi motuoche
%eng gloss: use *ride motorcycle
%eng: by motorcycle.
Errors in complement selection were found in the two-year-old of the Taiwanese
mother, who used the intransitive verb buxing ‘cannot’ as a transitive verb (7).
(7) T2: 我不行這顆
%pho: uɔ̌ pù ɕ1́ŋ tʂə̀ khə̄
%Pinyin: wo buxing zheke
%eng gloss: I cannot this CL
%eng: I can’t eat this one.
It was also found in the five-year-old Indonesian child, who used the path verb
‘jin’ alone as a transitive verb (8).
(8) I5: *我進這一個。[投進一個籃球]
%pho: uɔ̌ tɕ1ń tʂəі̀ ̄ kə̀ [thoú tɕ1ń і̄ kə̀ lɑn ́ tɕhioú ]
%Pinyin: wo jin zhe yi ge toujin yi ge lanqiu
%eng gloss: I shoot this one CL [shoot in one CL basketball]
%eng: I will shoot this one.
The word order errors were observed only in children of Indonesian mothers as
shown in the following examples.
(9) INV: 你在哪裡上課?
%pho: nǐ tsai nǎ lǐ ʂaŋ̀ khə̀
%Pinyin: ni zai nail shangke
%eng gloss: you at where attend class
%eng: Where do you attend class?
I6: * 上這裡。[在這裡上]
%pho: ʂaŋ̀ tʂəì ̌ [tasi tʂə̀ lǐ ʂaŋ̀ ]
%Pinyin: shang zheli zai zheli shang
%eng gloss: *go here [at here go]
%eng: Right here.
1
With kind permission from Crane Publishing: Mandarin Acquisition by Children of Vietnamese
Mothers in Taiwan (2008), pp. 99, 103, by Jenny Yichun Kuo, Examples (2, 6, 15).
(10) INV: 你功課寫完了沒?
%pho: nǐ kƱ̄ ŋ khə̀ ɕiě uań lə meí
%Pinyin: ni gongke xie wan le mei
%eng gloss: You class write-Asp Asp not
%eng: Do you have done your homework?
I5: *去學校寫 [在學校寫完了]
%pho: tɕʰỳ ɕyé ɕiaù ɕiě [tsai ɕyé ɕiaù ɕiě uań lə]
%Pinyin: qu xuexiao xie zai xuexiao xiewan le
%eng gloss: *go school write [at school write-Asp Asp]
%eng: I have finished it in school.
Indonesian children also misused the ba construction as in Example (11).
(11) I6: *把我割 [割到我]
%pho: pǎ uɔ̌ kə̄ [kə̄ taù uɔ̌ ]
%Pinyin: ba wo ge [gedao wo]
%eng gloss: *ba I cut [cut me]
%eng: I got a cut.
Although Taiwanese children omitted function words, they never omitted con-
tent words. Indonesian children omitted content words as in Example (12).
A five-year-old Indonesian child omitted you ‘have.’
(12) I5: 我 [有] 很多鉛筆。
%pho: uɔ̌ iǒ u xə̌ n tuɔ̄ tɕʰiɛ̄ n pǐ
%Pinyin: wo you henduo qianbi
%eng gloss: I [have] many pencil
%eng: I have a lot of penciles.
4.4.2 Question Elicitation Task
Since questions were lacking in children’s spontaneous speech and Mandarin

questions are different from English questions, a question elicitation task (Kuo
2008) was given.
English yes/no questions are acquired before wh-questions, because yes/no
questions were formed by auxiliary inversion or do support and the formation
wh-questions involve both what and auxiliary movement (Gleason 2012). Mandarin
wh-question is in situ. Questions in Mandarin were formed by adding a question
word without inversion or an auxiliary. Yes/ no questions can be formed by adding
the question final particle ma, or using the A-not-A pattern. The A can be a verb, a
copula an auxiliary or an adjective.
4.4.2.1 Participants
The participants were the same with those whose natural speech sample were
collected. There were nine children of Indonesian mothers and ten children of
Taiwanese mothers. It was two years after the collection of natural speech sample,
so the participants were two years older. The age of the children was four, five, six,
seven, and eight.
4.4.2.2 Materials
The same set of four colored pictures with minor differences used in Kuo’s (2008)
study was used. There were four or five animals holding different objects in each
picture.
4.4.2.3 Procedures
The participants were tested individually at their homes. The children were given
four pictures, and their mother had of one of them. The children were required to
ask questions to find out which of the four pictures their mothers was holding.
When the child found out, the mother took another picture. The game continued till
the mother had all the pictures. The total procedures took about 15 min for each
child. The process was tape recorded and transcribed.
4.4.2.4 Results and Discussion
The mean accuracy rate of questions by children with Indonesian mother is a little
bit lower than their age-mates with Taiwanese mothers. Accuracy is defined by the
tokens of correct use out of the total use of questions. The mean of Indonesian
children is 77.39 %, while that of Taiwanese children is 79.73 %. The error types
will be discussed later. The accuracy rate of individual children’s questions is listed
in Table 4.5. NA stands for not available. The child refused to participate in the
task.
It is unusual the older children did not ask better questions. The older children
were able to reach the communicative goal with fragments, that is, the limitation of
the task. Taiwanese children produced both A-not-A and particle questions as in
(12–14). They used question particles most of the time.
(12) T4: 小豬頭上有沒有蘋果?
%pho: ɕiaǔ tʂū thouʂ̀ aŋ́ iǒ u meì iǒ u phiŋk̀ uǒ
%Pinyin: xiaozhu toushang youmeiyou pinguo
%eng gloss: Little pig head above A-not-A apple
%eng: Does the pig have an apple?
Table 4.5 Accuracy rate of children’s questions

Yr/nationality Indonesian Taiwanese Mean (%)
Gender Male (%) Female (%) Male (%) Female (%)
4 NA 86.27 100 94 93.42
5 87.50 64.29 100 93.55 86.34
6 62.07 60.53 83 44.83 62.61
7 81.25 67.86 66.67 85 75.20
8 96.55 85.71 62.5 67.74 78.13
Mean (%) 81.84 72.93 82.43 77.02
Total mean (%) 77.39 79.73
(13) T4: 貓咪是不是在下面啊?

%pho: maū mі̄ sі́ pú sі́ tasі́ ɕiá miań a?
%Pinyin: maomi shibushi zai xiamian a
%eng gloss: cat A-not-A at below PART
%eng: Is the cat at the bottom of the picture?
(14) T4: 妳的是四隻嗎?
%pho: nі̌ tə ʂі́ sі́ tʂі̄ mā ?
%Pinyin: nide shi si zhi ma
%eng gloss: you de is four CL PART
%eng: Are there four animals in your picture?
The four-year-old sometimes used the particle redundantly in A-not-A question
as in Example (15). Function words were omitted in (16) and (17).
(15) T4: 媽咪, 馬馬的旁邊是不是熊熊嗎?
%pho: mǎ mі̀, mǎ mà tə phàŋ piē n sі́ pù sі́ ɕyɔ̌ ŋ ɕyɔ̀ŋ mā ?
%Pinyin: mami mama de pangbian shibushi xiongxiong ma[*]
%eng gloss: mom horse de nearby A-not-A bear PART
%eng: Mommy, is the horse next to the bear?
(16) T4: 媽咪, 兔子 [的] 旁邊是豬豬嗎?
%pho: mǎ mі,̀ thú tsi phaŋ̀ piē n sі́ tʂū tʂū mā ?
%Pinyin: mami tuzi [de] pangbian shi zhuzhu ma
%eng gloss: mom rabbit nearby is pig PART
%eng: Mommy, is the pig next to the rabbit?
(17) T5: 媽媽妳 [的] 卡片有沒有小豬?
%pho: ma mà nі̌ ka phián iǒ u mèi iǒ u ɕiǎ u tʂū
%Pinyin: mama ni [de] kapian youmeiyou xiaozhu
%eng gloss: mom you card A-not-A little pig
%eng: Mom, is there a pig in your picture card?
Transferring the experiential marker you from Taiwanese is also common such
as the six-year-old child.
(18) T6: 你的兔子有* 拿紅蘿蔔[*] 嗎?

%pho: nі̌ tə thú ts̩ ioŭ*̌ nà xɔŋ̀ lɔ̀ pɔ̄ [*] mā
%Pinyin: ni de tuzi you* na hongluobo ma
%eng gloss: you de rabbit have take carrot PART
%eng: Does your rabbit have a carrot?
The errors of the eight-year-old Taiwanese were caused by misinterpretation of
pictures as in Examples (19 and 20).
(19) T8: 牛 *[馬] 在中間嗎?
%pho: nioù tsaі́ tsɔ̄ ŋ tɕiē n mā
%Pinyin: niu ma zai zhongjian ma
%eng gloss: cow *[horse] in middle PART
%eng: Is the cow* [horse] in the middle?
(20) T8: 嗯…. 熊熊抱枕[*]頭 [罐子] 嗎?
%pho: um ɕyɔ̌ ŋ ɕyɔ̀ŋ paú tsə̌ n[*] thɔ̀ mā
%Pinyin: en xiongxiong baozhentou guanzi ma
%eng gloss: PART bear hold pillow bottle PART
%eng: Does the bear have a pillow?
The Indonesian five-year-olds omitted the question final particle ma as in
Example (21). He also transferred young ‘use’ from Taiwanese.
(21) I5: 這(*) 是(*) 用* 看的?
%pho:ts(*)əʂ́ (́ *)jɔŋ́ * khań tə
%Pinyin: zhe shi yong* kan de
%eng gloss: this(*) is(*) by look de
%eng: Is this for view?
Omission of other function words such as shi ‘is’ and fragment questions are
also common in Indonesian children as in Examples (22–24).
(22) I5: 阿這(*) 個勒?
%pho: a ts(*)ə-́ kəlei
%Pinyin: a zhe ge le
%eng gloss:PART this CL PART
%eng: What is this?
(23) I5: 這(*) 什(*)麼東東阿?
%pho.:ts(*)ə ́s(*)ə̌ -mə tɔ̄ ŋ-tɔ̄ ŋ a
%Pinyin: zhe sheme dongdong a
%eng gloss: this what thing PART
%eng: What is this?
(24) I5: 吊飾(*) 嗎?

́ ʂ mā
%pho: tiau
%Pinyin: diaoshi ma
%eng gloss: ornament PART
%eng: Ornament?
The Indonesian children also transferred Taiwanese structure wule, you zai in
Mandarin ‘have at’ for the present progressive.
(25) I5: 小貓咪有在唱(*)歌嗎?
%pho: ɕі̌ aū mau-mi і̌ ou tsaí tsh(*)aŋ́ -kə̄ mā
%Pinyin: xiaomaomi you zai changge ma
%eng gloss: Little cat have Asp sing PART
%eng: Is the cat singing?
Here are more examples from the oldest Indonesian children. All questions types—
wh-questions (26), disjunctive questions (27), particles questions (28), and A-not-A
questions (29) have been observed.
They sometimes asked incomplete questions though understandable. In Example
(26), the possessive marker de is omitted.
(26) I8: 媽媽 [de] 動物有幾隻(*)。[<]
%pho: mā -mā tuŋ́ -uі́ ̌ outɕі̌ tʂ(*)
%Pinyin: mama de dongwu you ji zhi
%eng gloss: mom de animal have how many CL
%eng: Mother, how many animals?
In Example (27), whose animals were not specified?
(27) I8: 動物是(*)四隻(*) 還是五隻(*)
%pho: tuŋ́ -u ʂ́ (́ *) ʂ-́ tʂ̄ (*)haì ʂ(́ *) ǔ -tʂ̄ (*)
%Pinyin: dongwu shi si zhi haishi wu zhi
%eng gloss: animal is four CL or five CL
%eng: Are these animals four or five?
In Example (28), both the verb and question particle are omitted.
(28) I8: 紫色格子吧? 黃色?
%pho: tʂ̌ -sə́kə̀-tspaxùaŋ-sə́
%Pinyin: zise gezi ba huangse
%eng gloss: purple check PART yellow
%eng: No, it is the purple check. Yellow?
(29) I8: 是(*)不是(*) 這(*) 一張(*)
%pho: ʂ(́ *)pu-̀ ʂ(́ *)ts(*)ə-́ і-́ ts(*)ɑ̄ ŋ
%Pinyin: shibushi zhe yi zhang
%eng gloss: A-not-A this one CL
%eng: Is this one?
4.5 Discussion and Summary 147
The syntactic development of children of Indonesian mothers was slower than

children of Taiwanese parents in length, sentence types, and question accuracy.
They have shorter mean length of utterances, lower Mandarin Index of Productive
Syntax, and lower accuracy rate in questions. The results are different from children
of Vietnamese mothers, whose MLU shows little difference from children of
Taiwanese mothers in the MLU till age six with even more sentence types (Kuo
2008).
Overall, the syntactic development of Indonesian children is slower than native
Taiwanese children. Indonesian children do have difficulties in acquiring Mandarin
syntax. The results of this study fail to support innatism since children did not pick
up the correct language forms in their environment. However, common errors
support that some innate constraints may be guiding their syntactic development.
The creation of sentences that they had never heard before rejected the possibility of
imitation. In addition, feedback on syntactic errors was given by neither Indonesian
nor native mothers. Thus, behaviorism cannot explain syntactic development.
Interaction plays an important role. Children who have more opportunities to
interact with peers seem to develop better. Language acquisition is the result of
interaction of innate constraints and input factors.
References
Bassano, D., & Mendes-Maillochon, I. (1994). Early grammatical and prosodic marking of
utterance modality in French: A longitudinal case study. Journal of Child Language, 21, 649–
675.
Bellugi, U. (1967). The acquisition of negation (Unpublished doctoral dissertation). Cambridge,
MA: Harvard University.
Bever, T. G. (1970). The cognitive basis for linguistics structure. In J. R. Hayes (Ed.), Cognition
and the development of language. New York: Wiley.
Bloom, L., Lahey, J., Hood, Lifter, K., & Fiess, K. (1980). Complex sentences: Acquisition of
syntactic connectives and the semantic relations they encode. Journal of Child Language, 7,
235–261.
Bowerman, M. (1990). Mapping thematic roles onto syntactic functions: Are children helped by
innate linking rules? Linguistics 28, 1253–1289.
Brown, R. (1973). A first language: The early stages. Cambridge.
Brown, R., & Fraser, C. (1963). The acquisition of syntax. In C. N. Cofer & B. Musgrave (Eds.),
Verbal behavior and learning: Problems and processes. New York: McGraw-Hill.
Chang, H. (1984). The comprehension of complex Chinese sentences by children: Relative
clauses. Chinese Journal of Psychology, 26, 57–66.
Cheung, H. (2006). The uses of complex sentences in Mandarin-speaking children. NSC report.
de Villiers, P. A., & de Villiers, J. G. (1979). Early language. Massachusetts: Harvard University
Press.
de Villiers, J. G., Reoper, T., & Vainikka, A. (1990). The acquisition of long distance rules. In
L. Frazier & J. G. de Villiers (Eds.), Language processing and acquisition. Dordrecht, the
Netherlands: Kluwer Academic.
Elman, J. L. (1990). Finding structure in time. Cognitive Science, 14, 179–211.
Finch, S., & Chater, N. (1992). Bootstrapping syntactic categories. In Proceedings of the 14th
Annual Conference of the Cognitive Science Society.
Gleason, J. B., & Ratner, N. B. (2012). The development of language (8th ed.). Cambridge:
Pearson.
Gleitman, L. R., & Wanner, E. (1982). Language acquisition: The state of the art. In E. Wanner &
L. R. Gleitman (Eds.), Language acquisition: The state of art (pp. 3–48). Cambridge, MA:
Harvard University Press.
Hambruger, H., & Crain, S. (1982). Relative acquisition. In S. A. Kuczaj (Ed.), Language
development, Vol. 1: Syntax and semantics. Hillsdale, NJ: Erlbaum.
Horgan, D. (1978). The development of the full passive. Journal of Child Language, 5, 65–80.
Karmiloff-Smith, A. (1979). A functional approach to child language: A study of determiners and
reference. Cambridge: Cambridge University Press.
Klima, E., & Bellugi, U. (1966). Syntactic regularities in the speech of children. In J. Lyonsand &
R. Wales (Eds.), Psycholinguistic papers. Edinburgh: Edinburgh University Press.
Kuo, J. Y. (2008). Mandarin acquisition by children of Vietnamese Mothers in Taiwan. Taipei:
Crane.
Levy, Y. (1988). The nature of early language: Evidence from the development of Hebrew
morphology. In Y. Levy, I. Schlesinger, & M. D. S. Braine (Eds.), Categories and processes in
language acquisition (pp. 73–98). Hillsdale, NJ: Erlbaum.
Lin, H-Y. (2009). Toward an index of productive syntax in Mandarin. MA thesis. National Chiayi
University.
Maratsos, M. (1988). The acquisition of formal word classes. In Y. Levy, I. M. Schlesinger & M.
Braine (Eds.), Categories and processes in language acquisition (pp. 31–44). Hillsdale, NJ:
Erlbaum.
Mills, A. E. (1986). The acquisition of gender: A study of English and German. Berlin:
Spinger-Verlag.
Pérez-Pereira, M. (1991). The acquisition of gender: What Spanish children tell us. Journal of
Plunkett, K., & Marchman, V. (1991). U-shaped learning and frequency effects in a multi-layered
perception: Implications for child language acquisition. Cognition, 38, 43–102.
Pye, C. (1988). Precocious passives (and antipasives) in Quiche Mayan. Paper presented at the
Child Language Research Forum, Stanford, CA.
Radford, A. (1990). Syntactic theory and the acquisition of English syntax. Oxford: Basil
Blackwell.
Scarborough, H. (1989). Index of productive syntax. Applied Psycholinguistics, 11, 1–22.
Suzman, S. (1987). Passives and prototypes in Zulu children’s speech. African Studies, 46,
241–254.
Tager-Flusberg, H. (1982). The development of relative clauses in child speech. Child Language
Development, 18, 104–111.
Valian, V. (1986). Syntactic categories in the speech of young children. Developmental
References 149
Valian, V. (1991). Syntactic subject in the early speech of American and Italian children.
Cognition, 40, 21–81.
Winzemer, J. A. (1980). A lexical expectation model for children’s comprehension o wh-questions.
Paper presented at the Fifth Annual Boston University Conference on Language Development.
Wootten, J., Merkin, S., Hood, L., & Bloom, L. (1979). Wh-questions: Linguistic evidence to
explain the sequence of acquisition. Paper presented at the biennial meeting of the Society for
Research in Child Development, San Francisco.
Chinese
Cheung, H. 張顯達. (1998) 平均語句長度在中文的應用 Pingjun yuju zhangdu zai zhongwen de

yingyong [The application of MLU in Chinese]. Speech-Language-Hearing Association of the
Republic of China, 13, 36–48.
Chapter 5
Conclusion
This study has investigated the acquisition of Mandarin by children of Indonesian

mothers and Taiwanese mothers in Taiwan. Generally speaking, the language
development of children of Indonesian mothers is slower than children of
Taiwanese parents. The findings are summarized in phonological, lexical, and
syntactic aspects with limitations and implications.
5.1 Phonological Development
Children of Indonesian immigrants have unique and common difficulties in

phonological development. Younger children of Indonesian mothers tended to
simplify complex vowels, which have not been observed in children of Taiwanese
or Vietnamese mothers. For example, the two-year-old Indonesian child changed /
tiao/ ‘lose’ to [ta]. On the other hand, the challenging alveopalatal affricates for
children of Vietnamese and Taiwanese mothers (Kuo 2008) do not seem to be a
problem for them. They preferred affricates to fricatives. For example, the
two-year-old Indonesian changed /sə/ ‘color’ to [tsə] and the six-year-olds changed
/ɕi/ ‘river’ to [tɕi].
Like children of Vietnamese and Taiwanese, children of Indonesian have diffi-
culties in retroflex, nasals, and aspiration. They tended to replace retroflexes with
alveolars just like many adult Mandarin speakers in Taiwan. For example, the
three-year-old Indonesian pronounced /tʂə/ ‘this’ as [tsə]. Younger children tended
to drop nasals in coda while older children changed the place of articulation. The
two-year-olds pronounced /tan/ ‘egg’ as [ta], while the six-year-olds pronounced /
ʂaŋ/ ‘hurt’ as [san]. Children of Indonesian were often confused by aspiration like
their mothers. For example, the three-year-olds said /tɕin/ ‘enter’ as [tɕhin].
However, even though the Indonesian mothers had problems with Mandarin tones,
children’s tones were not affected.

DOI 10.1007/978-981-10-1035-4_5
152 5 Conclusion
Children of immigrants have no difficulty in the acquisition of Mandarin tones,

but have difficulties in acquiring segments. Some are also difficult for native chil-
dren while others are unique to children of immigrants.
5.2 Lexical Development
Across age-groups, children of Indonesian mothers have smaller productive

vocabulary size than children of Taiwanese mothers except nouns. On average,
Indonesian children have acquired 74.44 nouns and Taiwanese children have
acquired 69.6 nouns. The verbs acquired by Indonesian children (mean = 48.44)
were fewer than those of Taiwanese children (mean = 61.6). There is great indi-
vidual variation in vocabulary composition. Nouns dominate most of children’s
lexicon while verb bias was observed in two Indonesian children and three
Taiwanese children. The results support neither the universal noun bias hypothesis
(Gentner 1981) nor Tardif’s (1996) and Tardif’s et al. (1999) argument for a verb
bias for Mandarin-speaking children. Other factors may have contributed to lexical
development.
It seems that the vocabulary growth by children of Indonesian immigrants is
slower than children of native Mandarin-speaking Taiwanese mothers. The finding
is different from that of children of Vietnamese immigrants, whose lexical devel-
opment is not different from native children (Kuo 2008). This could be due to the
fact that Vietnamese has more Chinese-based lexicon than Indonesian.
5.3 Syntactic Development
The syntactic development of children of Indonesian mothers was slower than that
of children of Taiwanese parents in length, sentence types, and questions accuracy.
They have shorter mean length of utterances, lower Mandarin Index of Productive
Syntax, and lower accuracy rate in questions. The mean MLU (M = 3.02) was
smaller than the Taiwanese group (M = 3.61) and such difference showed up as
early as age 2. The sentence patterns used by the Indonesian group (M = 36.11)
were also fewer than those used by the Taiwanese group (M = 39.10). The mean
accuracy rate of questions by children with Indonesian mother is a little bit lower
than their age-mates with Taiwanese mothers. The mean of Indonesian children is
77.39 % while that of Taiwanese children is 79.73 %. Over all, the syntactic
development of Indonesian children is slower than native Taiwanese children.
The results are different from children of Vietnamese mothers, whose MLU
shows little difference from children of Taiwanese mothers in the MLU till age six
with even more sentence types (Kuo 2008).
5.4 Limitations 153
5.4 Limitations
There are two limitations of the study. First, the children in the two groups did not
match exact age in months. The age difference could have caused differences for
younger groups. Second, due to the small sample size, the results may not be
representative for all the children of immigrants. I was unable to control all the
factors that affect language acquisition such as the children’s IQ, siblings, and
socioeconomic status. Although IQ was not measured, none of them was reported
abnormal. Individual differences were not the focus of the present study. Families
with Indonesian mothers tend to have lower socioeconomic status than native
Taiwanese families. I have done my best to match the status of Taiwanese families
with them.
5.5 Implications
Children of Indonesian mothers in Taiwan were acquiring Mandarin as their first

language while interacting with their mothers with Mandarin as a second language.
They also received native Mandarin and Taiwanese input from other family
members or media like children of native Taiwan parents. Their delay in the lan-
guage development indicates the importance of child-directed speech and rejected
innatism. According to behaviorism, Indonesian mothers did not provide a good
model for imitation or negative reinforcements. Corrective feedback was not usu-
ally present in native Taiwanese mothers, either. Only a few cases were observed
about word meanings. Interactionism seems to be the best explanation. Interaction
plays an important role. Children who have more opportunities to interact with
peers seem to develop better.
On the other hand, many of the difficulties faced by children of Indonesian
mothers were also faced by children of Taiwanese mothers. They can be due to
innate biological constraints or poverty of stimulus. Biological constraints play the
most important role in phonological development. Conceptual development guides
semantic development and input determines the vocabulary size. Interaction is
necessary for syntactic development. Language acquisition is the result of inter-
action of innate constraints and input factors.
Children of mixed marriages have the advantage to become bilinguals like many
successful simultaneous bilingual cases all over the world (Yip and Matthews
2010). Immigrant mothers should be encouraged to speak their mother tongues to
their children and children can get input of Mandarin or Taiwanese from other
family members. Also, nursery schools can be an excellent place for children to
interact with peers. However, preschool programs can be expensive for family of
immigrants. The government should help establish more affordable preschool
programs to support children’s development.
154 5 Conclusion
References
Gentner, D. (1981). Some interesting differences between verbs and nouns. Cognition and Brain
Theory, 4(2), 161–178.
Crane.
Tardif, T. (1996). Nouns are not always learned before verbs: Evidence from Mandarin speaker’s
early vocabulary. Developmental Psychology, 32, 492–504.
Tardif, T., Gelman, S., & Xu, F. (1999). Putting the “noun bias” in context: A comparison of
English and Mandarin. Child Development, 70, 620–635.
Yip, V., & Matthews, S. (2010). The acquisition of Chinese in bilingual and multilingual contexts.
International Journal of Bilingualism, 14(1), 127–146.

Jenny Yi-Chun Kuo (Auth.) - Mandarin Development of Indonesian Immigrants' Children - A Longitudinal Study in Taiwan-Springer Singapore (2016)

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

Jenny Yi-Chun Kuo (Auth.) - Mandarin Development of Indonesian Immigrants' Children - A Longitudinal Study in Taiwan-Springer Singapore (2016)

Uploaded by

Copyright:

Available Formats

Jenny Yi-Chun Kuo

ISBN 978-981-10-1033-0 ISBN 978-981-10-1035-4 (eBook)

Library of Congress Control Number: 2016938656

© Springer Science+Business Media Singapore 2016

Printed on acid-free paper

This Springer imprint is published by Springer Nature

This book summarizes the results of a 3-year longitudinal project on Mandarin

3 Lexical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Figure 2.1 The mechanism of OT (Kager 1999: 22) . . . . . . . . . . . . . . 19

Table 1.1 A comparison among Mandarin, Vietnamese,

Table 2.19 Acquisition and substitutions of the sounds from the

Table 4.3 MLU for children of Indonesian and Taiwanese

Tableau 2.1 OT tableau (Dekkers et al. 2000: 3) . . . . . . . . . . . . ..... 20

Immigrants or children of immigrants constitute most of the population in Taiwan.

1.2 Previous Research on Children of Immigrants

© Springer Science+Business Media Singapore 2016 1

However, previous studies on the language development of children of immi-

1.3 Mandarin Versus Vietnamese Versus Indonesian

Vietnamese is more similar to Mandarin than Indonesian. Both Mandarin and

Table 1.1 A comparison among Mandarin, Vietnamese, and Indonesian

1.4 Research on Immigrants in Other Countries

1.5 Learning Chinese as a Second Language

1.6 Language Acquisition Theories

inadequacy or helpful, because it is closer to learner’s level. According to Krashen’s

1.7 The Present Study

Following the study on the Mandarin acquisition of children of the largest

1.7.1 The Two-Year-Old Indonesian Girl

1.7.2 The Three-Year-Old Indonesian Boy

The three-year-old boy of an Indonesian mother is extroverted. His father has a

1.7.3 The Three-Year-Old Indonesian Girl

1.7.4 The Four-Year-Old Indonesian Boy

The four-year-old boy of an Indonesian mother is extroverted. His father has a

1.7.5 The Four-Year-Old Indonesian Girl

1.7.6 The Five-Year-Old Indonesian Boy

1.7.7 The Five-Year-Old Indonesian Girl

The ﬁve-year-old girl of an Indonesian mother is not obviously extroverted or

1.7.8 The Six-Year-Old Indonesian Boy

The six-year-old boy of an Indonesian mother is a very extroverted boy. He is the

1.7.9 The Six-Year-Old Indonesian Girl

The six-year-old girl of an Indonesian mother is an introverted girl. She is the

1.7.10 The Two-Year-Old Taiwanese Boy

1.7.11 The Two-Year-Old Taiwanese Girl

The two-year-old girl of Taiwanese parents is not apparently introverted or extro-

1.7.12 The Three-Year-Old Taiwanese Boy

The three-year-old boy with Taiwanese parents is extroverted. He has a

1.7.13 The Three-Year-Old Taiwanese Girl

1.7.14 The Four-Year-Old Taiwanese Boy

The four-year-old child of Taiwanese parents is an extroverted boy. He has a sister.

1.7.15 The Four-Year-Old Taiwanese Girl

1.7.16 The Five-Year-Old Taiwanese Boy

1.7.17 The Five-Year-Old Taiwanese Girl

1.7.18 The Six-Year-Old Taiwanese Boy

1.7.19 The Six-Year-Old Taiwanese Girl

Table 1.2 Participant background information

1.8 Organization of the Book

Chen, X.-L., & Lin, Y.-X. 陳璽琳, 林月仙. (2006). 關於外籍配偶子女語言能力研究的一些看

Wang, R.-X. 王瑞勳 (2004). 大陸和外籍新娘婚生子女適應與學習能力之探究 “Dalu he waiji

2.1 General Literature on Phonological Development

2.1.1 Theories of the Acquisition of Phonology

Before understanding children’s phonological development, the theories of the

© Springer Science+Business Media Singapore 2016 17

2.1.1.1 Introduction to Optimality Theory (OT)

Optimality theory (OT), proposed by Prince and Smolensky (2004, 1997), is a