The Asian Journal of Applied Linguistics

Vol. 7 No. 1, 2020, pp. 18-31 AJAL

Chinese-to-English phonetic transfer of Chinese university ESL
students
Zheng Fu
School of English, Tianjin Foreign Studies University, China

Chang-Ho Ji
School of Education, La Sierra University, U.S.A.

Heidi Weiss-Krumm
Office of International Students, La Sierra University, U.S.A.

Geng Wang
School of Education, University of Glasgow, U.K.

Yunfei Ma
College of Humanities, Tianjin Agricultural University, China

Phonetic transfer is defined as an L1 influence on the acquisition of L2 phonetics.

Previous studies have investigated phonetic transfer in the area of articulation, but the
effects of L1 on L2 pronunciation measured by speech recognition technology have
been under-researched. This study aims to address the issue by focusing on a sample of
676 Chinese university ESL students. Drawing on quantitative data, it examined
whether the participants applied phonetic transfer to ESL learning and what factors
might have influenced the results of phonetic transfer. We assumed that Chinese-to-
English phonetic transfer occurs but that the extent of the transfer would be small
because Chinese and English belong to different language families. However, findings
from this study confirm that Chinese-to-English phonetic transfer occurs and the extent
is large. The findings regarding high transferability might be attributed to spelling
through phonics and the nature of pronunciation acquisition.

Keywords: Phonetic transfer; language transfer; second language learning; Chinese; English;
ESL

Introduction

Phonetic transfer and its extent

Phonetic transfer, widely accepted as a common phenomenon in second language
acquisition, refers to L1 influence on L2 phonetics acquisition (Eckman, 2004; Odlin,
2003; Ringbom, 2007). Rather than being a static phenomenon, a phonetic transfer is a
process in which learners transfer phonetic knowledge from one language to a different
language. Learners are likely to apply L1 techniques and mechanisms when learning an
L2 (Cook, 2003; Larsen-Freeman & Long, 2014). Phonetic transfer is defined as how the
sound system of a language can affect a user’s perception and production of speech in
another language (Jarvis & Pavlenko, 2008).
Results of phonetic transfer can be classified through Contrastive Analysis, an
application utilized to compare attributes and characteristics of L1 and L2 (Figueroa &

ISSN 2308-6262
http://caes.hku.hk/ajal
 The Asian Journal of Applied Linguistics 19

Gárate, 2005). This determines whether transfer results are correct or incorrect. A
distinction that is commonly made in the literature in connection with L2 learning is one
between positive transfer (facilitation) and negative transfer (interference). Transfer can
have either beneficial or negative consequences, depending on the distance between L1
and L2 (VanPatten & Williams, 2015).
Positive transfer is the facilitating influence of similarities between L1 and L2, which
indicates that the extent of L1 to L2 transfer could be substantial when the languages are
categorized into the same language family or different language families with a common
ancestor (Weinreich, 2010). The effects of positive transfer are determinable through
comparisons of success, and such comparisons often show that cross-linguistic
similarities can produce positive transfers in several ways (Odlin, 2012). The production
of speech sounds involves the formation of automatic motor skills, and this will result in
ingrained L1 speech habits that are hard to avoid in the L2 (Rogerson-Revell, 2011). L1
rules tend to transfer, with those rules that simplify pronunciation transferring more easily
than morphologically restricted rules (Hansen, 2006). As an example, it is common sense
to assume that the phonetic system of Italian would be facilitated by a previous knowledge
of Spanish and vice versa because the two languages are derived from a common source
(Angelis & Selinker, 2001; Gass & Selinker, 2001).
On the contrary, when something previously learned hinders performance or learning
of a new language, a negative transfer occurs. Negative interlanguage influence might be
possible if the language learner is fluent in a language typologically distant from the target
language (Angelis & Selinker, 2001), i.e., a negative transfer between two languages with
different sounds and structures is predicted resulting in learning difficulty and error. Since
a negative transfer involves divergences from norms in a target language, it will lead to
phonetic underproduction, overproduction, production errors, or misinterpretation (Odlin,
2012). If L1 and L2 belong to different language families and are geographically distant,
they are considered likely to produce more incorrect transfer results than correct ones.
For example, a common assumption is that if L1 and L2 are proximate (e.g., Dutch and
English) leaning will be easier and more rapid than if they are distant (e.g., Chinese and
English) (Ellis, 2015). Moreover, lack of a similar L1 sound and lack of equivalent
articulatory motor skills can make it hard to acquire an L2 sound and to remove
incorrectly transferred sounds (Rogerson-Revell, 2011).

Chinese-to-English phonetic transfer

Chinese and English are classified into different language families and do not originate
from the same ancestor language. The Chinese language, a logogram, is a member of the
Sino-Tibetan language family (Byram & Hu, 2017; Doleželová-Velingerová & Wagner,
2013; Thurgood & LaPolla, 2006), while English, a phonogram, belongs to the Indo-
European language family (Blake, 2008; Woodard, 2008). Because Chinese and English
have many significant differences in terms of phonetics and phonology (Kim, 2019), it
has been suggested that Chinese-to-English negative transfer occurs more frequently than
positive transfer (Wang, 2015; Yu & Odlin, 2016; H. Zhang, 2018). Differences between
Chinese and English consonants and vowels may interfere with phonetic transfer
(Xiaorong & Jian, 2011). As English is prominence-related while Chinese is tone-
determined, a positive Chinese-to-English phonetic transfer is hard to form (Liu, 2001).
However, the Chinese language uses the International Phonetic Alphabet (IPA) to
transcribe sound, and this system shares some phonetic similarities to the English IPA
system. English conveys meaning with an alphabetic system, and Chinese marks sound
with the Pinyin system, a Romanized system, which transcribes Chinese pronunciation
20 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

into Roman letters. The Pinyin system is called a “pseudo-transcription”; that is, the
Pinyin system is not a “real” phonetic transcription system or a “real” orthographic
system, but it performs these functions in a partial sense (Heselwood, 2013).
A positive role of the Pinyin system is often indicated and emphasized in Chinese
ESL learning (Bialystok, Luk, & Kwan, 2005). The Pinyin symbols are regarded as useful
for encoding Chinese characters to improve the independent study of phonological
awareness, tone awareness, phonemic awareness, and onset-rhyme skills. Also, the Pinyin
system can greatly enhance English reading in a cross-language transfer (Mau, 2006). For
advanced English learners, Chinese Pinyin’s phonological processing skills are reportedly
a unique predictor of English reading performance (Gu, 2004). The Pinyin naming and
English reading skills may help each other when children learn Chinese characters with
the Pinyin symbols (X. Chen, Xu, Nguyen, Hong, & Wang, 2010).

Research questions
There is a lack of empirical studies on the effect of phonetic transfer that is measured by
speech recognition technology, especially among Chinese ESL students. This study takes
Chinese university ESL students as research participants to investigate the phenomenon
of L1 to L2 phonetic transfer. To be specific, it examines whether L1 (Chinese)
pronunciation affects L2 (English) pronunciation and the extent to which L1 to L2 transfer
occurs. In this context, this study explores two questions:
1. Does the L1 pronunciation of Chinese university ESL students correlate with their
L2 pronunciation performance?
2. To what extent does the L1 pronunciation of the participants influence their L2
pronunciation performance?

Methods

Participants
The data collection for this study took place at a four-year public university in northern
China. This university provides undergraduate and graduate instruction in the areas of
Humanities, Social Sciences, and Management, particularly specializing in the field of
language and culture education. In 2016, the researchers contacted about 30 teachers at
the research site who were teaching non-English-specialization courses during the fall
semester asking for their assistance to contact potential participants for the study. Seven
teachers agreed, and all the students (N = 676) in their classes agreed to participate in the
research. Of the sample, 57.5% were female and 42.5% male. The participants were all
Year-2 students taking a four-hour compulsory university-level English course every
week for 18 weeks per semester. The time they spent on after-class English learning
depended on personal interests and goals. They were requested to fill out a personal
information sheet that asked their gender, age, the number of years they had learned
English, and the number of hours they practiced English per week.

Instruments
All participants took three measurement tests: an L1 (Chinese) pronunciation test, an L2
(English) oral proficiency test, and an L2 pronunciation test. The first test used a Chinese
pronunciation word list to evaluate the level and extent of the participants’ L1
pronunciation. This word list was the Pinyin table in the Xinhua Dictionary’s index
including all the Chinese Pinyin syllables (Xinhua dictionary, 2015). This test was used
 The Asian Journal of Applied Linguistics 21

not for an assessment of eloquence, but as an assessment of the degree of language

standardization achieved by the participants. The participants read this word list aloud to
a speech recognition tool that transcribed their speech and assessed their performance of
Chinese pronunciation.
The second test used the speaking section of the Test of English as a Foreign
Language (TOEFL) to measure the participants’ English oral proficiency. The tasks of
the speaking section consist of two parts: (a) participants respond to questions, discuss
the passage heard, and give an opinion; (b) participants respond in a speech to what is
heard and solve a problem (TOEFL iBTquickprep, 2011). A native English teacher
administered the TOEFL speaking test and evaluated the participants’ performance
according to the TOEFL rubrics including general description, delivery, language use,
and topic development as described by the Educational Testing Service (2014).
For the third test the researchers developed an English word list to assess the
participants’ English pronunciation. To achieve this, syllabic structures of English and
Chinese were compared by Contrastive Analysis. Chinese words are monosyllabic while
English words are either monosyllabic or multisyllabic. English syllables have an optional
onset, a nucleus, and an optional coda (e.g., cat /kæt/: /k/ onset, /æ/ nucleus, and /t/ coda).
Similarly, Pinyin syllables contain an optional onset, an optional medial, a nucleus, and
an optional coda (e.g., the Pinyin kuang /kʊaŋ/: /k/ onset, /ʊ/ medial, /a/ nucleus, and /ŋ/
coda). Pinyin syllables have an optional medial that English syllables do not have. English
syllabic structure can be decoded as V, CV, VC, or CVC (C refers to consonant and V to
vowel). Non-medial Pinyin syllables, also decoded as V, CV, VC, and CVC, have the
same structure as English syllables. But medial-containing Pinyin syllables decoded as
CVV and CVVC have two Chinese vowels that can be processed and pronounced as a
diphthong; then they can be transcribed as CV and CVC. In a sense, the structures of
English and Pinyin syllables are the same. Thus, English and Chinese can be compared
at the level of syllables. We combined all the English single consonants with vowels (i.e.,
single vowels and diphthongs) to produce the tested syllables. The words containing the
syllables formed an English word list, and each word contained a tested syllable. Because
the minimal speech unit discerned by a speech recognition tool is a word, monosyllabic
or multisyllabic words are read for speech identification. The participants were requested
to read English words to a speech recognition tool which scored them when the words
were recognized as correct.

Data classification
Contrastive Analysis was used to classify L1 to L2 phonetic transfer into three difficulty
levels: easy, somewhat difficult, and difficult. We designed a Chinese-to-English transfer
classification based on the assumption that English pronunciation difficulty represented
Chinese-to-English phonetic transfer difficulty. First, our classification identified 208
words from the English word list as easy, representing the low level of transfer difficulty.
For these syllables, the participants were assumed to be able to easily convert Chinese
syllables into English as many phonemes of Chinese and English are pronounced the same
or similarly. Table 1 shows the intersection of 21 phonemes belonging to this category:
14 consonants and seven vowels (Shei, 2014). They share some syllables made up of the
IPA symbols (e.g., the English word way and the Pinyin wei are marked as /weɪ/).
22 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

Table 1. Intersections between English and Chinese phonemes (after Shei, 2014)

English consonants Intersections of consonants Chinese consonants

/b/ /d/ /g/ /h/ /ʤ/ /p/ /t/ /f/ /k/ /l/ /tsh/ /ts/ /tʂh/ /tʂ/ /x/
/r/ /z/ /ʒ/ /ʃ/ /ʧ/ /m/ /n/ /ŋ/ /s/ /w/ /j/ /ʐ/ /ʂ/ /tɕ/ /tɕh/ /ɕ//ɥ/
/v/ /θ/ /ð/ (/kh/ /ph/ /th/)

English vowels Intersections of vowels Chinese vowels

/ɪ/ /ʌ/ /æ/ /ɒ/ /ɜ/ /i/ /u/ /ʊ/ /ɛ/ /ə/ /y/ /ɨ/ /ɣ/
/ɔ/ /a/

Moreover, some phonemes of the two languages sound similar, and the syllables
made up of these phonemes also sound similar. The participants might find it easy to
make a minor adjustment to pronounce the English sounds. In addition, some English
short vowels and Chinese long vowels also sound similar. The participants could shorten
long Chinese vowels to produce short English vowels. For example, the English word can
/kæn/ and the Pinyin kan /kan/ sound so similar that a speech recognition tool considers
them the same. Table 2 presents the pairs of English and Chinese phonemes that sound
similar.

Table 2. Pairs of similar phonemes of English and Chinese

English consonants Pinyin initials

b/b/ b /p/
d /d/ d /t/
g /g/ g /k/
h /h/ h /x/
r /r/ r /ʐ/
z /z/ z /ts/
ch/ʧ/ ch/tʂh/
sh/ʃ/ sh/ʂ/
j /ʤ/ zh/tʂ/
s /ʒ/ (e.g., vision) re /ʐə/

English vowels Pinyin finals

o /ɒ/ o /ɔ/
i/ɪ/ i/i/
u /ʊ/ u /u/
u /ʌ/ a /a/
or /ə/ e /ɣ/
 The Asian Journal of Applied Linguistics 23

Second, 168 words were counted as containing somewhat difficult pronunciation of

English syllables and represented the moderate level of transfer difficulty. The difference
between low-level and moderate-level transfer was identified by whether or not an
English syllable matched a Chinese character. The low-level syllables had one or more
counterpart Chinese characters, while the moderate-level did not. In this way, the
participants used the same or similar Pinyin symbols to spell the moderate-level words.
For example, the English word say and the sound of the Pinyin combination sei marked
by /seɪ/ does not match any Chinese character, but the participants might hear this sound
in a Chinese dialect. Although of moderate-level difficulty, they could still pronounce the
syllables by spelling through phonics. Therefore, the participants might pronounce 376
syllables (about 90%) of the English word list with the same or similar Pinyin symbols.
Third, 41 words in the word list were assigned the difficult level of English syllable
pronunciation because of their high-level transfer difficulty. The three English phonemes,
/v/, /θ/, and /ð/ do not exist or have a similar Pinyin symbol. The participants had difficulty
in pronouncing the words containing these phonemes; hence transfer difficulty at this
level was the highest.

Data collection and analysis procedures

The participants took the three tests as after-class activities, and several assessors
evaluated their performance. A speech recognition tool (HDecode tool HTKv3.4.1,
Young et al., 2009) with computer assistance was used to discern participants’ sounds to
obtain L1 and L2 pronunciation scores. An English-speaking teacher administered the
TOEFL speaking test to determine L2 oral proficiency scores. The analysis took into
account several independent variables, including gender, age, L1 (Chinese)
pronunciation, the number of years participants had learned L2 (English), the number of
hours they practiced L2 per week, and L2 oral proficiency, which were assumed to relate
to L2 pronunciation and the three transfer levels.

Results

Descriptive analysis
Table 3 summarizes the demographic information of the participants. Their average age
was 20.06 years old. On average, they reported to have studied English for 9.15 years and
to spend 6.58 hours per week on English learning. Concerning proficiency and
pronunciation, the mean score of L1 pronunciation was 384.98 of the total 411 Chinese
Pinyin words. Their average performance was 93.44% of the maximum possible level,
which indicates that they were competent in pronouncing Chinese words. Second, the
mean score of L2 oral proficiency was 7.51 out of 30 points (25.03%), showing that their
oral English level was significantly low. Considering they reported having studied
English for over nine years on average, it is surprising to see their poor performance in
oral English. Third, the mean score of L2 pronunciation as a whole was 367.04 out of the
total 417 English words (88.02%). Viewed separately by transfer level, the mean of the
low-level transfer difficulty was 191.00 of 208 words (91.83%); the mean of the
moderate-level transfer difficulty was 144.25 of 168 words (85.86%); the mean of high-
level transfer difficulty was 31.80 of 41 words (77.56%). These results show that the
average accuracy rates of L2 pronunciation as a whole and at different levels were above
75%.
24 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

Table 3. Results of descriptive analysis (N = 676)

Measure M SD Min. Max.

Age 20.06 1.37 18 25

Years of L2 (English) learning 9.15 1.75 6 15
Hours of practicing L2 per week 6.58 2.46 4 17
L1 (Chinese) pronunciation 384.98 20.47 280 411
L2 oral proficiency 7.51 3.24 3 23
L2 pronunciation 367.04 27.09 234 410
Low-level transfer 191.00 12.60 120 206
Moderate-level transfer 144.25 12.07 93 166
High-level transfer 31.80 4.40 11 39

Bivariate correlation analysis

Pearson’s correlation analysis was employed to determine whether four pairs of
relationships were established for Chinese university ESL students and how strong they
were. L2 pronunciation, a dependent variable, was assumed to correlate with years of L2
learning, hours of practicing L2 per week, L1 pronunciation, and L2 oral proficiency, the
primary independent predictors. In Table 4, L2 pronunciation was correlated with years
of L2 learning and L1 pronunciation but uncorrelated with hours of practicing L2 per
week and L2 oral proficiency. The correlation between years of L2 learning and L2
pronunciation was positive in direction, but its dimension was weak (r = .14). For L1 and
L2 pronunciation, the analysis produced a strong and positive correlation (r = .95), which
indicated that the participants with good L1 pronunciation were capable of accurately
producing L2 sounds. On the contrary, hours of practicing L2 per week had no significant
relationship with L2 pronunciation. Lastly, on the correlation between L2 oral proficiency
and L2 pronunciation, the result was not statistically significant.

Table 4. Results of Pearson’s correlation analysis (N = 676)

Measure Measure Measure Measure Measure

1 2 3 4 5

Years of L2 (English) learning --

Hours of practicing L2 per week .21** --
L1 (Chinese) pronunciation .12** .07 --
L2 oral proficiency .06 .71** .08* --
L2 pronunciation .14** .06 .95** .07 --

*p < .05 **p < .01

 The Asian Journal of Applied Linguistics 25

Table 5. Results of multiple linear regression analysis (N = 676)

L2 (English) Pronunciation Low-Level Transfer Moderate-Level Transfer High-Level Transfer

Measure B SE ß B SE ß B SE ß B SE ß

Constant -108.54** 8.41 -19.73** 5.14 -61.45** 5.63 -27.34** 2.41

Gender -.57 .69 -.01 -.37 .42 -.02 .02 .46 .00 -.22 .20 -.02

Age -.63 .39 -.32 -.32 .24 -.04 .24 .26 .03 -.54** .11 -.17

Years of L2 (English) learning .78* .30 .05 .17 .19 .02 .41* .20 .06 .20* .09 .08

Hours of practicing L2 per week -.15 .20 -.01 -.01 .12 -.00 -.01 .13 -.00 -.12* .06 -.07

L1 (Chinese) pronunciation 1.25** .02 .95 .56** .01 .91 .51** .01 .87 .18** .01 .83

L2 oral proficiency .04 .15 .00 -.01 .09 -.00 .03 .10 .01 .02 .04 .01

R2/Adj. R2 .90/.90 .83/.83 .78/.78 .70/.69

F (df) 1025.55** (6, 669) 548.37** (6, 669) 393.01** (6, 669) 253.79** (6, 669)

Note: Gender (0 female+, 1 male). *p < .05. **p < .01.

26 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

Multivariate analysis
Multiple regression analysis was conducted to estimate the importance of independent
variables to the prediction of overall L2 pronunciation, which was followed by three
parallel analyses using low-level, moderate-level, and high-level transfer scores as the
dependent variables. Four multiple linear regression models were respectively established
to discuss the influence of the predictors on the dependent variables.
As given in the first column of Table 5, the model of L2 pronunciation had an adjusted
R-squared value of .90, F (6,669) = 1025.55, p <.01. When it was divided into three levels
of difficulty, the model of the low-level transfer had an adjusted R-squared value of .83,
F (6,669) = 548.37, p <.01; the moderate-level transfer had an adjusted R-squared value
of .78, F (6,669) = 393.01, p <.01; the high-level transfer had an adjusted R-squared value
of .69, F (6,669) = 253.79, p <.01. The results are displayed in the second, third, and last
columns of Table 5. That is, all the models were significant and explained the substantial
amounts of the variance of the dependent variables.
It is clear that the participants’ L1 pronunciation had a strong and positive influence
on their L2 pronunciation. Across the three different transfer levels, the strong and
positive effects of L1 pronunciation were also found to be significant. The size of beta
coefficients affirmed that the level of L1 pronunciation influence was much larger than
the other predictors (i.e., age, years of L2 learning, hours of practicing L2 per week, and
L1 pronunciation). In addition, the length of L2 learning was positively associated with
L2 pronunciation except for the lower-level transferability scores. Age and hours of
practicing L2 per week were related to transferability scores of the high level, yet they
were not significant at the low or moderate levels.

Discussion

Unusual linguistic phenomena

According to the data, Chinese university ESL students’ L1 (Chinese) pronunciation is as
high as 93.44%, and that of L2 (English) oral proficiency is as low as 25.03%. That is,
the average performance for L1 pronunciation is rather high. This finding is somewhat
surprising given that the participants represented a variety of Chinese linguistic
backgrounds, speaking different varieties of Chinese. Their surprisingly high
performance might be attributable to the fact that Chinese university students are required
to take Putonghua Shuiping Ceshi (PSC), also called the Standard Chinese Proficiency
Test, which tests their abilities in reading and conversation (Lam, 2005). The PSC test
might help them improve Chinese pronunciation and oral proficiency to meet the official
national standard. The China National Common Language and Script Law (enacted in
2001) stipulates the citizens’ rights and obligations to learn and use the common language
and script, specifies the domains of their compulsory use, and requires proficiency
measurements in education, for example, the PSC test, and for certain professions, for
example government staff, teachers, and show hosts (Zhou, 2019).
On the other hand, the participants’ performance of L2 oral proficiency was rather
low despite their average nine-years of studying ESL. This finding may reflect the
phenomenon of “mute English” or “dumb English”, describing people who can read and
understand English as a second language but cannot speak it well (Xinhua News Agency,
2002). This is a common phenomenon among Chinese students learning English and is
perceived to be a serious flaw that exists in the current Chinese education of spoken
English teaching and learning (Liang & Xu, 2013). This phenomenon occurs in China
because English linguistic knowledge is overstressed, while communicative competence
 The Asian Journal of Applied Linguistics 27

is overlooked (Y. Zhang & Wang, 2011). The root of the problem results from English
tests in China which prompt ESL learners to focus on mechanical multiple-choice skills
but fail to direct attention to communicative competence (Lo Bianco, Orton, & Gao,
2009). The College English Test and the National College Entrance Examination that
purportedly measure students’ cumulative memorized knowledge about English language
are the bane of English teaching in China (Liao & Wolff, 2010). These English
examinations usually include listening, reading, translation, and short-letter writing but
exclude speaking and essay writing. They are designed to test students’ knowledge of
grammar, reading skills, and vocabulary, but ignore the practical ability of spoken and
written English. As China’s English education is examination-oriented but not ability-
oriented, students spend most time acquiring input (reading and listening) rather than
output (speaking and writing) skills. This would explain the finding that the time the
participants spent on English learning was not related to their oral English proficiency.

Method of spelling through phonics

The present findings demonstrate that L1 (Chinese) pronunciation of Chinese university
ESL students correlates with and has a strong effect on L2 (English) pronunciation.
Previous studies showed that in a transfer process, learners tend to transfer the forms and
meanings from L1 to L2 not only productively but also receptively (Arabski, 2006; Jarvis
& Pavlenko, 2008). In a similar vein, our data support the weight of phonetic transfer in
that L1 pronunciation performance is influential in determining L2 sound production.
Initially, we assumed that Chinese-to-English phonetic transfer might occur but its
extent might be small. However, the participants’ L1 impact on their L2 pronunciation
was more significant and stronger than we expected. Chinese students might learn English
by imitating foreign sounds they have heard. One popular idea in the second-language-
acquisition literature is that sounds are learned mainly through imitation (Breidegard &
Balkenius, 2003; Gass & Mackey, 2012), which provides some learning advantages,
especially in developing pronunciation and intonation (Lightbown & Spada, 2006).
However, our results seem to be consistent with the perspective that Chinese students
rely on spelling through phonics besides imitation. Most Chinese pupils are required to
learn Pinyin symbols first and pronounce words by spelling through phonics in early
school years. When they master Pinyin symbols and phonics rules, they learn how to read
and write Chinese characters marked by Pinyin symbols. Chinese examinations
throughout primary and high school include a pronunciation section to test students’
Pinyin mastery and correct pronunciation errors. The same is true for Chinese students in
learning English. Most students are taught English IPA symbols and required to
memorize the pronunciation of English words by spelling through phonics. Before they
completely master English IPA symbols, they always spell English words with Chinese
Pinyin symbols as most of English IPA and Pinyin symbols sound the same or similar.
Chinese ESL educators believe that spelling methods can enhance the accuracy of English
pronunciation. Spelling methods enable Chinese students to produce correct English
pronunciation even if they do not have an authentic language environment or they are
unable to imitate English pronunciation by listening to people talking. Spelling through
phonics leads to good performance of Chinese and English pronunciation, which suggest
the extension of phonetic transfer.
28 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

Nature of pronunciation acquisition

Learning time is regarded as an important dimension of production performance because
the accumulation of knowledge takes time and effort in the classroom and after class
(Brown, 2014). The length of time matters in L2 performance because the processing time
is a significant factor (Papageorgiou & Bailey, 2019). Contrary to this popular notion, the
data in the current study show that the time the participants spent on learning and
practicing the L2 had no, or at best a weak, effect on L2 pronunciation. Also, while oral
English proficiency is considered an important factor in predicting English pronunciation
(Pennington & Rogerson-Revell, 2019), the data shows that was not the case in the current
study.
A possible explanation for the phenomena is that L2 pronunciation is more acquired
than learned. An important distinction made by linguists between language acquisition
and learning has to do with the individual’s internal processes of learning and the degree
of consciousness associated with a learning task (Cook, 2016; McKay, 2006). Acquisition
implies the language involvement of innate, species-specific linguistic knowledge in the
mastery of a language, whether it is native or foreign (Anderson, 2007). That L1 to L2
transfer can come about through differences, recognizes that transfer can be an
unconscious process (Han, 2004). Generally, young people acquire L2 pronunciation
through an unconscious process during which they are mostly unaware of phonetics,
which is similar to the way they acquired their L1 pronunciation in their childhood. Once
they have formed their L1 accent along with the Pinyin system, they spontaneously use
Pinyin as a supportive tool to read English words and convert similar Pinyin rules into
English through the transfer process. Thus, they might develop Chinese English as a new
variety of English, and it might be hard for them to change their accent once developed.
In this way, transfer and developmental effects may interact in L2 acquisition, with L1
transfer dominant in the early stages of acquisition and developmental effects increasing
as L1 transfer effects decrease. Both may also affect the production and acquisition of a
single segment; while L1 transfer may affect pronunciation when sounds are difficult,
developmental effects may affect substitutions. It is possible that positive L1 transfer
effects may override developmental effects (Hansen, 2006). As pronunciation is acquired,
L2 pronunciation performance might not depend on L2 learning time or L2 oral
proficiency, but on L1 pronunciation level. Because Chinese university ESL students’
Chinese has been highly standardized with long-term training and assessment, their
Chinese level is high. They can convert more Chinese sounds into English words than is
generally believed.

Conclusion
This study explored the impact of Chinese university ESL students’ L1 (Chinese) on their
L2 (English) pronunciation. The results of this empirical research answered two research
questions. First, this study demonstrated a strong correlation between L1 pronunciation
and L2 pronunciation. Second, the extent to which L1 pronunciation influences L2
pronunciation performance is large and that contrasts with popular perception which is
based on the great discrepancy between the sound systems of Chinese and English.
The reasons why Chinese university ESL students’ L1 pronunciation has strongly
influenced their L2 pronunciation are probably because L1 to L2 phonetic transfer is
related to the way they have acquired English. Chinese children learn English
pronunciation mainly by spelling through phonics as they use English IPA symbols to
spell English words and correct English pronunciation. Moreover, they obtain L2
pronunciation by speech acquisition, and “L1 use does affect L2 accent regardless of
 The Asian Journal of Applied Linguistics 29

whether the L2 was acquired as a child or an adult” (Edwards & Zampini, 2008, p. 257).
Because the participants have taken standardized Chinese pronunciation examinations
throughout their education, their Chinese pronunciation level might be rather high, which
leads to a high L1 to L2 phonetic conversion.
Although the results show that participants presented high levels of English
pronunciation, their oral English proficiency was poor. It is important to note that ESL
teaching in China is problematic. Examination-oriented ESL teaching may contribute to
mute English or dumb English for most Chinese university students. So long as the
College English Test is the gatekeeper for university graduation, advanced studies, and
post-graduation employment, China will continue to produce functional illiterates (Liao
& Wolff, 2010). Measures are needed to cope with the problem, one of which is the use
of speech technology tools and technology development in the areas of: automatic speech
recognition technology, speech synthesis, spoken interaction, natural language
understanding, speech recognition, emotion in spoken dialogue systems, expressive
speech synthesis, affective computing, multimodal communication, interaction
technologies, and animated agents (F. Chen & Jokinen, 2010), which can be introduced
to ESL teaching and learning for improving learners’ oral English proficiency. For
instance, one-on-one individualized teaching effects can distinguish automatic speech
assessment technology, an intelligent tutoring system based on speech assessment for
English (Liang & Xu, 2013).

Acknowledgment
This research was supported by Scientific Research Projects of Tianjin Municipal Education Commission:
“Research on the Training Strategy of Talents Internationalization in Agricultural Universities under the
Context of Modern Urban Agriculture” (2018SK116).

About the authors

Zheng Fu is an associate professor in the School of English at Tianjin Foreign Studies University, China.
She received a Doctor of Education degree from La Sierra University. Her research interests include EFL
methodology and technology, instructional methods of Chinese as a second language, and Sino-US
education issues.

Chang-Ho Ji is chair and professor of the Department of School Psychology & Counselling in the School
of Education at La Sierra University. He received his Ph.D. in education from Andrews University and a
Ph.D. in political science from the University of California, Riverside. His research interests are
psychometric testing, moral development, and educational politics and policy.

Heidi Weiss-Krumm is a principal designated school official of the Office of International Students at La
Sierra University. She is a doctoral student in the School of Education at La Sierra University. Her interests
are international education and administration.

Geng Wang received a Ph.D. in educational sociology from the School of Education, University of
Glasgow, UK. Her research interests are vocational education, skill formation, youth transition, the youth
labour market, and educational skills and policy.

Yunfei Ma is a lecturer at the College of Humanities at Tianjin Agricultural University. She holds a master’s
degree in English language and literature from Tianjin Foreign Studies University. Her research interests
are in English language and literature and English education.

References
Anderson, D. (2007). Li2-language variation: Child language. from
http://www.ling.cam.ac.uk/li2/childlang_da.pdf
30 Zheng Fu, Chang-Ho Ji, Heidi Weiss-Krumm, Geng Wang, Yunfei Ma

Angelis, G. D., & Selinker, L. (2001). Interlanguage transfer and competing linguistic systems in the
multilingual mind. In J. Ceñoz, B. Hufeisen, & U. Jessner (Eds.), Cross-linguistic Influence in
third language acquisition (pp. 42-58). Clevedon: Multilingual Matters.
Arabski, J. (2006). Cross-linguistic influences in the second language lexicon. Clevedon: Multilingual
Matters.
Bialystok, E., Luk, G., & Kwan, E. (2005). Bilingualism, biliteracy, and learning to read: Interactions
among languages and writing systems. Scientific Studies of Reading, 9(1), 43-61.
https://doi.org/10.1207/s1532799xssr0901_4
Blake, B. J. (2008). All about language. Oxford: Oxford University Press.
Breidegard, B., & Balkenius, C. (2003). Speech development by imitation. from
http://cogprints.org/3328/
Brown, H. D. (2014). Principles of language learning and teaching (6th ed.). White Plains, MY: Pearson
Education.
Byram, M., & Hu, A. (Eds.). (2017). Routledge encyclopaedia of language teaching and learning (2nd
ed.). London: Routledge.
Chen, F., & Jokinen, K. (2010). Speech technology: Theory and applications. Boston, MA: Springer.
Chen, X., Xu, F., Nguyen, T.-K., Hong, G., & Wang, Y. (2010). Effects of cross-language transfer on
first-language phonological awareness and literacy skills in Chinese children receiving English
instruction. Journal of Educational Psychology, 102(3), 712-728.
https://doi.org/10.1037/a0018802
Cook, V. (2003). Effects of the second language on the first. Bristol: Multilingual Matters.
Cook, V. (2016). Second language learning and language teaching (5th ed.). London: Routledge.
Doleželová-Velingerová, M., & Wagner, R. G. (2013). Chinese encyclopaedias of new global knowledge
(1870-1930): Changing ways of thought. Berlin: Springer.
Eckman, F. R. (2004). From phonemic differences to constraint rankings: Research on second language
phonology. Studies in Second Language Acquisition, 26(4), 513-549.
https://doi.org/10.1017/S027226310404001X
Edwards, J. G. H., & Zampini, M. L. (Eds.). (2008). Phonology and second language acquisition.
Amsterdam: John Benjamins.
Ellis, R. (2015). Understanding second language acquisition (Second edition. ed.). Oxford: Oxford
University Press.
Figueroa, C. M., & Gárate, T. I. M. (Eds.). (2005). Studies in contrastive linguistics: Proceedings of the
4th international contrastive linguistics conference, Universidade de Santiago de Compostela,
Spain. September, 2005. University of Santiago de Compostela.
Gass, S. M., & Mackey, A. (2012). The Routledge handbook of second language acquisition. Abingdon:
Routledge.
Gass, S. M., & Selinker, L. (2001). Second language acquisition : An introductory course (2nd ed. ed.).
Mahwah, N.J.: Lawrence Erlbaum.
Gu, Y. (2004). Investigating age factors in cross-language transfer of phonological processing across
English and Chinese languages (Unpubklished master’s thesis), Wilfrid Laurier University,
Waterloo, Canada. Retrieved from http://scholars.wlu.ca/etd/748
Han, Z. (2004). Fossilization in adult second language acquisition. Bristol: Multilingual Matters.
Hansen, J. G. (2006). Acquiring a non-native phonology: Linguistic constraints and social barriers.
London: Continuum.
Heselwood, B. (2013). Phonetic transcription in theory and practice. Edinburgh: Edinburgh University
Press.
Jarvis, S., & Pavlenko, A. (2008). Crosslinguistic influence in language and cognition (1st ed.). New
York: Routledge.
Kim, T. E. (2019). Mandarin loanwords. New York, NY: Routledge.
Lam, A. (2005). Language education in China: Policy and experience from 1949. Hong Kong: Hong
Kong University Press.
Larsen-Freeman, D., & Long, M. H. (2014). An introduction to second language acquisition research.
London: Routledge.
Liang, Y., & Xu, J. (2013). An intelligent tutoring system based on speech assessment for spoken English
learning in China. In D. Liu, C. Alippi, D. Zhao, & A. Hussain (Eds.), Advances in brain inspired
cognitive systems (pp. 358-365). Heidelberg, Germany: Springer.
Liao, Y., & Wolff, M. (2010). Mute English—The Latin of China. In M. Wolff (Ed.), Of the students, by
the students, and for the students: Time for another revolution (pp. 19-43). Newcastle upon Tyne:
Cambridge Scholars Publishing.
 The Asian Journal of Applied Linguistics 31

Lightbown, P. M., & Spada, N. (2006). How languages are learned (3rd ed.). Oxford Oxford University
Press.
Liu, S. (2001). Studies on transfer in second language acquisition. Guangxi Normal University Journal, 3,
1-29.
Lo Bianco, J., Orton, J., & Gao, Y. (2009). China and English: Globalisation and the dilemmas of
identity. Bristol: Multilingual Matters.
Mau, P. (2006). Cross-language transfer of phonological awareness in Chinese-English bilinguals.
(Unpublished master’s thesis), University of Hong Kong, Hong Kong.
https://doi.org/10.5353/th_b3688930
McKay, P. (2006). Assessing young language learners. Cambridge: Cambridge University Press.
Odlin, T. (2003). Cross-linguistic influence. In C. J. Doughty & M. H. Long (Eds.), Handbook of second
language acquisition (pp. 436-486). Malden, MA: Blackwell.
Odlin, T. (2012). Language transfer: Cross-linguistic influence in language learning. Cambridge:
Cambridge University Press.
Papageorgiou, S., & Bailey, K. M. (Eds.). (2019). Global perspectives on language assessment research,
theory, and practice. New York, NY: Routledge.
Pennington, M. C., & Rogerson-Revell, P. (2019). English pronunciation teaching and research:
Contemporary perspectives. London: Palgrave Macmillan.
Ringbom, H. (2007). Cross-linguistic similarity in foreign language learning. Clevedon: Multilingual
Matters.
Rogerson-Revell, P. (2011). English phonology and pronunciation teaching. London: Continuum.
Shei, C. (2014). Understanding the Chinese language: A comprehensive linguistic introduction. New
York, NY: Routledge.
Thurgood, G., & LaPolla, R. J. (2006). The Sino-Tibetan languages. London: Routledge.
TOEFL iBTquickprep. (2011). Retrieved from
https://www.ets.org/s/toefl/pdf/accessible_toefl_ibt_quick_prep.doc
VanPatten, B., & Williams, J. (2015). Theories in second language acquisition: An introduction. London:
Routledge.
Wang, D. (2015). An analysis on negative transfer of native language in second language acquisition. In
D. Zheng (Ed.), Education management and management science (pp. 217-220). London: Taylor
& Francis Group.
Weinreich, U. (2010). Languages in contact: Findings and problems. The Hague, the Netherlands:
Mouton Publisher.
Woodard, R. D. (Ed.). (2008). The Ancient languages of Europe. Cambridge: Cambridge University
Press. https://doi.org/10.1017/CBO9780511486814
Xiaorong, L., & Jian, G. (2011). On phonetic negative transfer from Chinese to English. In J. Arabski &
A. Wojtaszek (Eds.), The acquisition of L2 phonology (pp. 16-26). Bristol: Multilingual Matters.
Xinhua dictionary. (2015). (3rd ed.). Beijing, China: The Commercial Press.
Xinhua News Agency. (2002, January 28, 2002). New generation of children dumping “Dumb English”.
Retrieved from http://www.china.org.cn/english/25990.htm
Young, S., Evermann, G., Gales, M., Hain, T., Kershaw, D., Liu, X., . . . Woodland, P. (2009). The HTK
book (for HTK Version 3.4). Cambridge: Cambridge University Engineering Department.
Yu, L., & Odlin, T. (2016). New perspectives on transfer in second language learning. Bristol:
Multilingual Matters.
Zhang, H. (2018). Second language acquisition of Mandarin Chinese tones: Beyond first-language
transfer. Leiden, Netherlands: Brill Rodopi.
Zhang, Y., & Wang, J. (2011). Primary school English language teaching in South China: Past, present,
and future. In F. Anwei (Ed.), English language education across greater China: Bilingual
education and bilingualism. Bristol: Multilingual Matters.
Zhou, M. (2019). Language ideology and order in rising China. Singapore: Palgrave Macmillan.