Professional Documents
Culture Documents
To cite this article: Mei Jiang, Raymond J. Green, Tracy B. Henley & William G. Masten (2009)
Acculturation in relation to the acquisition of a second language, Journal of Multilingual and
Multicultural Development, 30:6, 481-492, DOI: 10.1080/01434630903147898
Department of Psychology and Special Education, Texas A&M University-Commerce, P.O. Box
3011, Commerce, TX 75429, USA
(Received 17 October 2008; final version received 19 June 2009)
During the 20072008 academic year, 110,128 Chinese students (including 29,001
students from Taiwan) were enrolled in US colleges and universities, constituting
17.6% of the international students studying in the USA (Institute of International
Education 2008). Due to the importance of language for learning and communica-
tion, second language (L2) acquisition is a significant issue for students arriving in
the USA after puberty. Based on the fact that these students’ age of arrival (AOA) in
the USA usually has long passed the critical period of language acquisition (e.g.
Lenneberg 1967), it is debatable whether they can acquire English at a near native-
like level. A further question then is, if anything approaching native-like English is
achievable for such Chinese students, what variables relate to such an attainment.
Studies supporting the critical period hypothesis (CPH; Lenneberg 1967) propose
that learners who begin to acquire a L2 in a naturalistic environment after puberty,
compared with early learners, are usually constrained by age-related maturational
factors (e.g. Birdsong 2005; Johnson and Newport 1989; Krashen 1985; Long 1990;
Pinker 1994; Scovel 1988). After the developmental window closes, the L2 can only
be acquired ‘through a conscious and labored effort’ (Lenneberg 1967, 176)
presumptively due to lateralisation effects and the loss in neuroplasticity. Despite
the strong position of CPH claiming that becoming a native speaker is virtually
impossible for late learners, some studies demonstrate that elements of native-like L2
are attainable (e.g. Birdsong and Molis 2001; Bongaerts 1999; Ioup et al. 1994).
Method
Participants
Forty-nine Chinese international students, 23 (46.9%) males and 26 (53.1%) females,
participated in this study. All participants were enrolled in graduate programmes at a
large Texas University at the time of the study. There were 24 (49.0%) business majors
(i.e. MBA, marketing and economics), 15 (30.6%) engineering majors (i.e. computer
science and industrial engineering) and 10 (20.4%) social science and education
majors (i.e. sociology, education and psychology). All participants’ AOA in the USA
were greater than 18 and their lengths of residence in USA were not more than five
years (M 2.00, SD 1.28). By the time of the study, their average length of formal
English education in the USA (including English as a second language [ESL] in the
USA) was 1.86 years (SD 1.15). Ten (20.4%) participants self-reported that they
received ESL training in the USA before getting admitted into the university. All
participants were foreign-born non-native English speakers, with no self-reported
learning or hearing disabilities. The participants’ Computer-Based TOEFL (CBT)
scores ranged between 173 and 260 (M 211.96, SD 22.07), with listening,
structure, reading and writing subtests included in the TOEFL test. None of the
participants majored in English in China or received intensive spoken English
training for more than six months in China. All of them were single at the time of the
study.
The demographic data indicated that students from Mainland China and Taiwan
had similar English education backgrounds. They all received in-class English
instruction from native Chinese-speaking teachers in middle schools and colleges. No
significant difference, t(47) 0.21, p0.84, was found for English in-class
instruction time in middle school between participants from Mainland China
(M5.80, SD2.10) and those from Taiwan (M 5.64, SD 2.15). Likewise, there
was no significant difference, t(47) 0.78, p 0.44, for English in-class instruction
time in college between participants from Mainland China (M 3.00, SD1.25) and
those from Taiwan (M 2.62, SD1.43).
Materials
Personal Data Questionnaire
The Personal Data Questionnaire collected participants’ demographic information
on gender, current major, length of residence in the USA, length of education in the
USA, TOEFL score, marital status, and also the information concerning their birth
place (i.e. Mainland or Taiwan), AOA in the USA and major area of study in China.
was measured on a four-point Likert scale from 1 (false) to 4 (true) to generate one
DSI score and an ESI scores, respectively. As the first acculturation scale developed
for use across ethnic groups, SMAS showed impressive reliability across major ethnic
groups (African American, Asian American, European Americans and Hispanic
Americans) and countries of origins. Alpha reliability coefficients were 0.86 for the
entire scale, 0.97 for ESI and 0.90 for DSI. Strong validity was also demonstrated in
the scale’s ability to identify the relationship between generation status and the
performance on the subscales. In the current study, participants received both
English and back-translated Chinese versions of the SMAS to insure accurate
comprehension of the scale.
L2 proficiency interview
In the current study, a second L2 element, oral speaking proficiency, was assessed in
the L2 proficiency interview. Language speaking proficiency has been widely
recognised as an important aspect of L2 learning (Educational Testing Service
[ETS] 1992) such that starting from 2005 the speaking subtest was added to the CBT
test to assess test takers’ topic development and language delivery. Speech has been
used to assess L2 attainments in past studies as well (e.g. Hansen 1995). The L2
proficiency interview in the current study was modelled after the Simulated Oral
Proficiency Interview (SOPI) protocol originally designed by the Centre of Applied
Linguistics (Malone 2000). It consisted of two picture-based, one topic-based and
one situation-based tasks, as shown below:
. [picture-based]: Now, look at these cartoons. Each cartoon tells a funny story.
This is the first cartoon. Can you figure out what they are doing in the first
486 M. Jiang et al.
picture? (Pause) Now, tell me the whole story. (Pause) Interesting! Now, look
at the second cartoon. Tell me what they’re doing in the first picture. (Pause)
What happened after that?
. [picture-based]: Here is the campus map. Suppose I come to visit my good
friend Jenny who’s studying at this school. I’m supposed to meet her in the
Business Administration Building (SPOT B). But I’m lost at SPOT A and I see
you at this time. How would you give me the direction to the Business
Administration Building? (Pause)
. [topic-based]: Next, we are going to hear some of your opinions. As an
international student studying in America, what do you think are the
advantages and disadvantages of studying abroad? More specifically, please
tell me about the good side and the bad side of studying here. (Pause)
. [situation-based]: Suppose you’re absent for a class yesterday because your
roommate was sick and you had to drive him/her to the hospital. The
professor was very angry because you were supposed to show up and give a
presentation yesterday. Now, you’re going to the professor’s office to make the
apology. What are you going to say to him/her? (the end)
Data coding
A holistic scoring method (Elliot, Plata, and Zelhart 1990) was used to score both
participants’ pronunciation in the L2 sentences reading task and their oral
proficiency in the L2 proficiency interview. The holistic scoring method, that makes
use of norming rubrics and training manipulations, has demonstrated high
reliabilities (i.e. inter-rater agreements) and validities in the past (Elliot, Plata, and
Zelhart 1990; Taylor 1998). In the current study, the pronunciation category under
FLOSEM and ACTFL proficiency guidelines were used as the rubrics of the
assessments to evaluate participants’ pronunciation and oral proficiency, respectively.
Journal of Multilingual and Multicultural Development 487
Following the process suggested by Elliot, Plata, and Zelhart (1990), two key
raters (native English speakers) selected eight training samples out of the data pool
for pronunciation levels 2 to 6. There was at least one sample for each level, although
no training sample was identified for level 1 because no participant exhibited
‘difficulty in accurately reproducing the target language sounds and sound patterns’.
Another four raters (native English speakers) were then trained with those samples
before scoring the experimental data. Each piece of experimental data was assessed
by two raters. According to the holistic scoring guidelines (Elliot, Plata, and Zelhart
1990), an agreement was reached when the data received scores that were identical, or
separated by one point. The same procedure was used to evaluate participants’ oral
proficiency in the L2 proficiency interview. Training samples were selected for level 4
(Intermediate Low) to level 10 (Superior). Levels below 4 were omitted, as all
participants demonstrated the ability to fulfil communicative tasks.
Combining two raters’ scores, participants’ pronunciation performance in the L2
sentences reading task had an average score of 6.65 (SD 1.64) with a possible score
of 12; and oral speaking proficiency in the L2 proficiency interview had an average
score of 11.86 (SD 2.70) with a possible score of 20. There was a positive
relationship between pronunciation scores and speaking proficiency scores, r(49)
0.613, p B0.001. The reliability of the holistic scoring method, in the form of inter-
rater agreement, was 94% (n 46) for pronunciation scores and 90% agreement (n
44) for speaking proficiency scores, respectively.
Results
On the measure of acculturation, the ESI subtest of the SMAS had an average score
of 61.88 (SD 4.28) with a possible score of 68, and the DSI subtest had an average
of 33.33 (SD 5.66) with a possible score of 60. Their corresponding percentile
scores then ranged from 74 to 100% for ESI but merely from 33 to 75% for the DSI,
suggesting that participants were unanimously more immersed in their original
culture than in the dominant American culture. On the relationship of ESI scores and
DSI scores, participants’ DSI scores had a significantly lower mean than ESI scores,
t(48) 21.96, p B0.001, but no significant relationship between ESI scores and
DSI scores was found, r(49) 0.01, p 0.96. Such results strongly suggested that the
participants in the current study were uniformly immersed in their original culture
(i.e. all above 74%), yet demonstrated much more variability in their acculturation to
the dominant society (3375%). Given this, only DSI scores were included in further
analyses.
Multiple regression analyses were conducted to see if the DSI scores had an
association with oral proficiency scores and pronunciation scores, respectively. Based
on the literature review, gender, length of residence in the USA, length of education
in the USA and TOEFL scores were planned for the regression analyses as possible
contributors to pronunciation and oral proficiency. Pre-analysis data screening
indicated that the length of education had a slight redundant effect (tolerance
0.051) with other variables so it was deleted from the further analyses.
The first sequential multiple regression analysis was performed to determine
whether oral proficiency would be affected by DSI scores and/or other variables (i.e.
gender, length of residence and TOEFL scores). That is, whether the addition of
information regarding gender, length of residence and TOEFL scores improved
prediction of speech beyond that afforded by differences in DSI scores. Data
488 M. Jiang et al.
demonstrated that R was significantly different from zero at the end of each step. As
the only variable in the first step, DSI accounted for 8% of the variance of the speech
scores and was a significant contributor to speech scores, b 0.28, t(47) 1.98,
p 0.05, with a small-to-medium (adjusted R2 0.06) effect size (Cohen 1988). In the
second step of the regression analysis, gender, length of residence in the USA and
TOEFL were entered, accounting for another 18% of the variance. When the two
steps were combined, the four variables significantly accounted for 26% of the
variance, with a medium-to-large effect size (adjusted R2 0.20). In addition to DSI
scores, TOEFL scores were found to be a significant contributor to speech scores,
accounting for 17% of the variance, b 0.42, t(47) 3.14, p 0.003.
A similar sequential multiple regression analysis was performed on pronunciation
scores to determine if the addition of information regarding gender, length of
residence and TOEFL score improved prediction of pronunciation beyond that
afforded by differences in DSI score. Data demonstrated that R was not significantly
different from zero after step one but was after step two. After step one, as the only
variable, DSI score was not a significant predictor of pronunciation scores, b 0.25,
t(47) 1.78, p0.08, accounting for only 6% of the variance in pronunciation scores.
The effect size was small (adjusted R2 0.04). However, when all four variables were
combined, they accounted for 24% of the variance and showed a medium to large
effect size (adjusted R2 0.17). Gender was a significant predictor of pronunciation
scores, accounting for 11% of the variance, b 0.342, t(44) 2.52, p 0.015. There
was a significant effect for gender on pronunciation scores, with females receiving
higher scores than males (Mfemale 7.23, Mmale 6.00, t(47) 2.80, p0.007).
Discussion
The current study considered whether acculturation relates to second language
acquisition among the Chinese international students studying in the USA after
puberty. Based on the fact that the participants in this study arrived in the USA
within the past five years, they were all found to be still deeply immersed in their
original culture, but participants varied on their acculturation process towards
American society. The ubiquitously high ESI scores suggested that immersion was a
gradual and slow process and that five years had relatively little impact on
participants’ bond with Chinese society. Participants were still closely aligned with
their ethnic heritage, eating Chinese food, socialising with Chinese friends and
speaking Chinese whenever English was not required. This may also have been due to
the fact that most of the participants in the university had little interest in staying in
the US long term, thus had little motive to distance themselves from their native
culture.
In contrast, scores varied widely on the DSI portion of SMAS, reflecting that
participants had achieved various degrees of acculturation to American society.
Compared with the ESI pattern, this finding supports Berry’s (1980) argument that
getting closer to the dominant society is not necessarily achieved by separating from
the ethnic society. Since ESI scores showed little variance, they were excluded from
further analyses in the current study. That is, rather than defining acculturation in
terms of psychological immersion within Chinese society, the current study focused
on the participants’ degrees of immersion within American society.
Multiple regression analyses on two elements of L2 attainments, pronunciation
and oral speaking proficiency, demonstrated that the degree of immersion in
Journal of Multilingual and Multicultural Development 489
show greater interest in the target culture. As such, the result in the current study
supports the idea that gender socialisation is an important variable for L2
acquisition. Moreover, it is noticeable that the TOEFL score was found to predict
the participant’s speaking proficiency scores in this study. Although this version of
TOEFL test for this study did not include a speaking subtest, a significant
relationship between TOEFL scores and speaking proficiency scores was evident.
In summary, the current study investigated the relationship between immersion
within American society and L2-speaking proficiency and pronunciation with the
ChineseEnglish population, a group that speaks a non-Indo-European language
and has been numerically increasing for years but somehow has been neglected in L2
research. The results demonstrated a strong relationship between acculturation
towards the dominant society and English-speaking proficiency, suggesting that
English oral proficiency could be achieved through immersion within American
society, although accents were more problematic. Since the participants in the current
study were ubiquitously immersed into their original Chinese society, only their
acculturation progress towards the dominant society has been examined related to L2
acquisition, thus implying that L2 learners could reach a high L2 proficiency level
while keeping their original culture. Acquiring a L2 is not necessarily achieved by
distancing from the ethnic society.
As such, in order to enhance L2 oral proficiency, late life L2 learners should
consider getting more immersed within the American culture. For example, in
addition to English language learning per se, learners could benefit from obtaining
more knowledge about American history and customs, socialising with American
people outside of the workplace, and also try more unique elements of the American
lifestyle. Along the same line, immersion activities perhaps should be coordinated
into formal L2 education programmes to facilitate L2-speaking proficiency for
educational purposes. Real life-related activities and cultural events could be
introduced into the L2 classroom to better expose learners to the American culture,
such that the conventional boundary between formal schooling and the real world
could be softened as a way to enhance L2 learning. At the same time, it should be
emphasised that this does not require L2 learners to move away from their original
ethnic culture. All of these implications may be most useful when cultural and
linguistic differences are great such as is the case with ChineseEnglish bilinguals
who study and work in America.
References
Bacon, S., and M. Finnemann. 1992. Sex differences in self-reported beliefs about foreign
language learning and authentic oral and written input. Language Learning 42: 47195.
Baker, S.C., and P.D. MacIntyre. 2000. The role of gender and immersion in communication
and second language orientations. Language Learning 50: 31141.
Berry, J.W. 1980. Acculturation as varieties of adaptation. In Acculturation: Theory, models and
some new findings, ed. A.M. Padilla, 925. Boulder, CO: Westview.
Berry, J.W. 1995. Psychology of acculturation. In The culture and psychology, ed. N.R.
Goldberger and J.B. Veroff, 45788. New York: New York University Press.
Bialystok, E. 1997. The structure of age: In search of barriers to second language acquisition.
Second Language Research 13: 11637.
Bialystok, E., and B. Miller. 1999. The problem of age in second language acquisition:
Influences from language, structure, and task. Bilingualism: Language and Cognition 2:
12745.
Journal of Multilingual and Multicultural Development 491