Journal of Neurolinguistics 33 (2015) 14e28

Contents lists available at ScienceDirect

Journal of Neurolinguistics
journal homepage: www.elsevier.com/locate/
jneuroling

White-matter structure in the right hemisphere

predicts Mandarin Chinese learning success
Zhenghan Qi a, b, *, Michelle Han a, b, Keri Garel a, b,
Ee San Chen a, b, John D.E. Gabrieli a, b, c
a

Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139,
USA
McGovern Institute for Brain Research, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
c
Institute for Medical Engineering & Science, Massachusetts Institute of Technology, Cambridge, MA 02139,
USA
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 18 March 2014
Received in revised form 11 August 2014
Accepted 13 August 2014
Available online 2 September 2014

Second language learning becomes increasingly difcult with age,

but some adults learn more successfully than others. We examined
whether inter-subject variability in the microstructure of white
matter pathways, as measured by diffusion tensor imaging (DTI),
would predict native English speakers' outcomes in learning
Mandarin Chinese. Twenty-one adults were scanned before
participating in an intensive 4-week Mandarin course. At the end
of the Mandarin course, participants completed a nal exam that
assessed their skills in both spoken and written Mandarin. Individual participants' white-matter tracts were reconstructed from
their native DTI data and related to nal-exam performance. Superior language learning was correlated with DTI measures in the
right hemisphere, but not in the left hemisphere. In particular,
greater initial fractional anisotropy (FA) in both the right superior
longitudinal fasciculus (parietal bundle) and the right inferior
longitudinal fasciculus was associated with more successful Mandarin learning. The relation between white-matter structure in the
right hemisphere of native English speakers and successful initial
language learning may reect the tonal and visuo-spatial properties, respectively, of spoken and written Mandarin Chinese.
2014 The Authors. Published by Elsevier Ltd. This is an open
access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/3.0/).

Keywords:
Second-language acquisition
Mandarin
Diffusion tensor imaging
White matter
Right hemisphere

* Corresponding author. Massachusetts Institute of Technology, 43 Vassar St., 46-4033, Cambridge, MA 02139, USA.
Tel.: 1 617 324 4959.
E-mail address: [email protected] (Z. Qi).
http://dx.doi.org/10.1016/j.jneuroling.2014.08.004
0911-6044/ 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/3.0/).

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

15

1. Introduction
In contrast to native language acquisition, which seems to be effortless during the rst years of
childhood, second language acquisition during adulthood is exceptionally challenging (Hull & Vaid,
2007; Johnson & Newport, 1989; Kuhl, 2004; Perani et al., 1998, pp. 1841e1852; Sakai, 2005).
Adults vary substantially in their second-language learning efcacy, with few learners achieving
native-like uency in either the spoken or written form of the new language. Little is known about
how adult language learning capacities are represented in the brain, but neuroimaging studies
have started to explore the structural and functional neural differences underlying individual
variation in adult language learning. Here, we asked whether structural differences in whitematter pathways are related to differences among English-speaking adults in learning spoken
and written Mandarin.
In studies of English and other alphabetic languages, two white-matter tracts have been most
strongly associated with language: the left arcuate fasciculus (AF) component of the superior longitudinal fasciculus (SLF) and the left inferior longitudinal fasciculus (ILF). The SLF connects a dorsal
language network (Hickok & Poeppel, 2004, 2007) and is thought to consist of a direct AF pathway
connecting posterior (superior temporal gyrus/Wernicke's area) and anterior (inferior frontal gyrus/
Broca's area) language cortices and an indirect pathway including the SLFp connecting the inferior
parietal cortex and anterior language cortices (Catani, Jones, & Ffytche, 2005). The ILF connects a
ventral language network that includes Broca's area and posterior occipitotemporal regions and, via
another pathway, the anterior temporal lobe with the uncinate fasciculus (Anwander, Tittgemeyer, von
sche, 2007).
Cramon, Friederici, & Kno
Among these language-related white-matter tracts, the left AF is most prominently implicated in
language function; disruption of the left AF leads to impairment in speech production, auditory
comprehension, and reading (Dronkers, Plaisant, Iba-Zizen, & Cabanis, 2007; Rauschecker et al., 2009;
Vandermosten et al., 2012; Yeatman, Dougherty, Ben-Shachar, & Wandell, 2012; Yeatman et al., 2011).
The left ILF has also been implicated more specically in visual (orthographic) aspects of reading,
perhaps by transmitting visual information about words between the fusiform gyrus at the temporaloccipital junction and anterior/interior temporal regions (Epelbaum et al., 2008; Yeatman et al., 2012).
Neuroimaging in relation to language learning has shown an association between individual
learning performance and both structural and functional connectivity in the left hemisphere. In one
study, diffusion measures of the left AF strongly associated with superior learning of English pseu pez-Barroso et al., 2013). Another study found a positive correlation between diffusion
dowords (Lo
measures of the left ILF and the learning of associations between Mandarin tones and English pseudowords (Wong, Chandrasekaran, Garibaldi, & Wong, 2011). A resting-state functional magnetic
resonance imaging (rsfMRI) study found that successful learning of foreign sound contrasts was
associated with greater functional connectivity between Broca's area and the left superior parietal area,
which are two regions connected by the left SLFp, a white matter tract in the dorsal pathway (VenturaCampos et al., 2013). Bilingualism also affects the left inferior frontal-occipital fasciculus, an arguably
indistinguishable tract from the left ILF (Wakana et al., 2007), such that fractional anisotropy (FA), one
of the diffusivity measures indexing white matter microstructure, is highest in early bilinguals and
lowest in monolinguals with late bilinguals being in between (Mohades et al., 2012).
In contrast to the attention given to the left hemisphere, only a few studies have explored relations
between structural or functional connectivity in the right hemisphere and language learning, and even
these studies have reported mixed ndings. One study found diffusion measures in the right AF
correlated with individual's performance in pitch-related grammar learning (Loui, Li, & Schlaug, 2011).
Conversely, a study of Japanese speakers learning English found no association between the learning
performance and the white-matter structure in either the left or right hemisphere before learning.
Instead, learning prociency correlated with the training-induced increases in the strength of the
structural connectivity between the right inferior frontal gyrus and caudate (Hosoda, Tanaka, Nariai,
Honda, & Hanakawa, 2013).
Because Mandarin differs from English and other alphabetic languages in both spoken and written
forms, the neural bases that support learning Mandarin may be different from those that support
learning European (alphabetic) languages. For spoken language, Mandarin has four lexical tones (high-

16

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

level, rising, dipping and falling) that confer prosodic features to speech at the syllable level. Together
with the consonantevowel combination, lexical tone is a crucial phonetic component signaling the
semantics of the syllable. The four tones can be distinguished by the shapes of the internal pitch
contour, and there is evidence that the right hemisphere is dominant in processing pitch-related
acoustic information (see review: Wong, 2002). Damage to the right auditory cortex or disruptions
to its connectivity pattern leads to impairment in pitch-related processing, including production and
perception of speech prosody (Baum & Pell, 1999; Loui, Alsop, & Schlaug, 2009; Pell, 1998; Warrier &
Zatorre, 2004). Further, the processing of pitch information in speech has been associated with
increased activation in both the right superior temporal gyrus and the right inferior frontal gyrus
(Meyer, Alter, Friederici, Lohmann, & von Cramon, 2002; Wildgruber, Pihan, Ackermann, Erb, & Grodd,
2002; Zatorre & Gandour, 2008). Thus, the role of pitch perception in spoken Mandarin may invoke
greater right-hemisphere participation than that occurring for non-tonal languages.
Despite the role of the right-hemisphere in acoustic processing of lexical tones that is important for
spoken Mandarin, there is evidence that phonological processing of lexical tones is subserved by the
left hemisphere in native Mandarin speakers (Gandour et al., 2004; Gu, Zhang, Hu, & Zhao, 2013; Tracy
et al., 2011; Wang, Wang, & Chen, 2013; Xu et al., 2006). Electrophysiological and fMRI evidence has
shown that processing lexical tones is left-lateralized in native Mandarin speakers, especially for pitch
contour information, while processing acoustic tones remains right-lateralized regardless of listeners'
language background (Gandour et al., 2004; Tong et al., 2005). Thus, there are two alternative possibilities in regards to the lateralization of early learning for nave Mandarin learners whose native
language is English. Early learning of spoken Mandarin could depend upon right-hemisphere networks
involved in pitch processing. Alternatively, learning Mandarin lexical tones, which is dominated by the
left hemisphere in native Mandarin speakers, may also depend upon left-hemisphere networks.
For written language, Mandarin employs a logographic system. Unlike alphabetic languages, which
map visual symbols onto phonemes, each Chinese character is a monosyllabic word. Chinese characters
usually consist of two or more spatially arranged radicles with a large variety of spatial relationships.
The pronunciation of characters is usually opaque with only 28% of the phonetic radicles carrying the
same phoneme information as the whole character (Perfetti, Liu, & Tan, 2005). Thus, Mandarin learners
cannot apply the same grapheme-to-phoneme scheme of alphabetic systems while reading Chinese,
but instead access the phonological information through visuo-orthographic information of the
character (Tan, Hoosain, & Peng, 1995; Tan, Laird, Li, & Fox, 2005). Despite widespread consensus on
left-hemisphere dominance in reading (Cohen & Dehaene, 2004; Pugh et al., 2001; Saygin et al., 2013;
Tan et al., 2011; Yeatman et al., 2012), a number of cross-linguistic studies have reported languagespecic effects on the recruited neural network. Chinese readers showed greater activation of the
right middle occipital and right fusiform regions than English readers when reading in their native
languages (Bolger, Perfetti, & Schneider, 2005; Tan et al., 2005). In Chinese-English bilingual readers,
processing Chinese characters with increased spacing between strokes elicited bilateral activation in
the cuneus, while processing English words with increased spacing did not (Sun, Yang, Desroches, Liu,
& Peng, 2011). Therefore, we hypothesized that the visuo-spatial demands of learning Chinese characters may invoke right-hemisphere networks specialized for visuo-spatial processing (Pisella et al.,
2011).
One DTI study directly examined Mandarin-learning induced change in white matter structure. This
longitudinal study measured white-matter microstructure over the course of nine months of Mandarin
instruction and found increased fractional anisotropy (FA) in frontal-lobe tracts crossing the genu of
the corpus callosum in the language-training group but not in the control group (Schlegel, Rudelson, &
Tse, 2012). The development of stronger inter-hemispheric connectivity suggested that Mandarin
training in native English speakers, in whom speech and reading is highly left-lateralized, might lead to
increased right-hemisphere involvement. This study did not examine whether initial variability in
white-matter pathways was associated with better or worse learning.
In the present study, we examined whether the characteristics of language-relevant white-matter
pathways (left and right AF, SLFp, and ILF) at baseline were correlated with variation in prociency of
learning spoken and written Mandarin in adults. Learners were students enrolled in an intensive onemonth Mandarin training program in the classroom, a traditional learning environment similar to reallife learning experiences. Using diffusion tensor imaging (DTI), we extracted diffusion measures of

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

17

fractional anisotropy (FA) for each tract within each individual. Because previous research has reported
positive correlations between FA values and superior language learning, we hypothesized that higher
FA values would predict greater Mandarin prociency in the learners. In particular, if individual
sensitivity to pitch information and visuo-spatial information is important for early Mandarin learning,
the right-hemisphere pathways may be more associated with gaining early Mandarin prociency than
the left-hemisphere pathways.
2. Methods
2.1. Language training procedure and prociency tests
Participants signed up for one of two course sections taught by the same instructor. The course used
The New Practical Chinese Reader Volume 1 (Liu, 2010) as the main textbook. The course, which took
place on the MIT campus, provided students with intensive training in introductory Modern Standard
Mandarin, 3.5 h per day for 5 days per week over 4 weeks. Daily homework was assigned from the
companion workbook and was supplemented with additional material designed by the instructor.
Throughout the course, participants spent, on average, 62.3 h in the classroom and completed 11 assignments, 10 quizzes, 1 mid-term exam, and 1 nal exam. An anonymous mid-term survey (in which
14 of the 21 students responded) showed participants also spent, on average, 2.7 h every day outside of
class studying Mandarin.
The nal exam was composed of speech production, listening, and reading sections, mostly written
in Chinese characters, which tapped into the students' ability to effectively integrate orthographic
information with phonologic, semantic, and syntactic information. The nal exam was graded by the
instructor on a scale of 0e100 and was used as the overall measure for the students' Mandarin learning
performance.
In addition to measuring the participants' holistic usage of the language with the nal exam total
scores, we included two more tests that more specically measured individuals' phonological skill and
orthographic skill. Phonological skill was measured by participants' accuracy in the pinyin (an alphabetic phonetic system for Mandarin pronunciation) dictation part of the nal exam. The students wrote
down the pinyin after listening to each of 10 Mandarin phrases/sentences twice. Orthographic skill was
measured in a computer-based task at the laboratory administered within one week after the Mandarin course. In each of the 30 trials, participants saw an English word and were asked to choose the
correct translation from the two given Chinese characters. All the Chinese characters were taught in the
class and were paired based on their close resemblance in visual forms (e.g. STAR, or?). Behavioral
accuracy reected how well one could match the slight difference in the spatial arrangement of strokes
with different semantic meanings. For both tests, scores were percentage of correct answers.
2.2. Participants
Twenty-two participants were recruited for the study. Because outlier values can have strong inuences on correlational analyses, one participant was excluded based on diffusion measures that
deviated more than 2.5 standard deviations from the mean in 5 of the 6 tracts of interest. All analyses
proceeded with the remaining twenty-one participants (mean age 23.6, SD 3.74; 13 males and 8
females; 18 right-handed and 3 left-handed). All volunteers gave written consent that was approved by
the institutional review board at Massachusetts Institute of Technology prior to their participation.
Participants were screened for MRI compatibility and no history of neurological diseases. The participants had an average standard IQ of 117 (SD 12.84) measured by Kaufman Brief Intelligence Test
(KBIT-2, Kaufman & Kaufman, 2004). All 21 participants, as indicated by the Language History Questionnaire (Linck et al., 2013), were native American-English speakers, of which 17 were monolinguals
with no extensive exposure to languages other than English as a child. Four participants reported
growing up in a bilingual environment (Polish, Spanish, Hindi and Hebrew). Among these, three
participants spoke the other language uently and one participant spoke English exclusively during
childhood. Importantly, the four bilingual participants did not differ from the monolinguals in respect
to their average nal exam scores (77.5 vs. 76.8, out of 100). All participants were exposed to foreign

18

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

languages, ranging from 1 to 5 different languages, with large variability in the level of uency. The selfreport uency of the second language ranged between 1.0 (limited) and 5.0 (excellent), with an average
of 3.0 (standard deviation, 1.15). However, crucially, none of the participants had prior exposure to
Chinese, other tonal languages (e.g. Thai), or any other language with a logographic writing system (e.g.
Japanese Kanji).
2.3. Imaging acquisition and processing
Diffusion tensor images were acquired before Mandarin training began. Data were collected on a
Siemens Trio 3-Tesla MRI scanner with a standard 32-channel phased array head coil (TE 84 ms;
TR 8040 ms). The diffusion-weighted scan included 10 non-diffusion-weighted reference volumes
(b 0) and 60 diffusion-weighted volumes (b 700 s/mm2) with voxel resolution of 2.0 mm isotropic
(eld of view 256 mm; matrix 128 mm  128 mm) and 64 slices per volume. The b0 volumes were
collected in the rst 10 volumes of the entire scan and were later averaged. A high-resolution T1weighted magnetization prepared rapid gradient echo (MPRAGE) scan was also acquired
(TE 1.64 ms; TR 2530 ms; FOV 220 mm; matrix 220 mm  220 mm; 176 slices). Cortical and
white matter surfaces were generated according to individual landmarks using the FreeSurfer image
analysis suite (http://surfer.nmr.mgh.harvard.edu/), with further manual editing for accuracy.
The DTI data were pre-processed using DTIPrep's (Oguz et al., 2014) automated artifact detection
software, which automatically excluded extensively distorted volumes. The data were then processed
using TRACULA (Yendiki et al., 2011), which corrected for eddy currents and head motion; performed
both intra- and inter-registration of each subject's b0 diffusion images to their T1-weighted images and
the MNI template respectively; calculated the tensor t at each voxel to generate maps of diffusion
tensor eigenvalues; and estimated the endpoints and waypoints for each major white matter pathway
using subject-specic anatomical information. FSL's bedpostx tool was used to t a ball-and-stick
model of diffusion to each subject's DWIs. TRACULA's pathway reconstruction software generated
probability distributions of the trajectory of the 18 major white matter pathways. Of these, we
examined three bilateral tracts of interest, dened a priori: the ILF, the SLFp, and the AF (Fig. 1).
The diffusion tensor eigenvalues extracted from TRACULA outputs were used to compute axial
diffusivity (AD), radial diffusivity (RD), and fractional anisotropy (FA) for the three bilateral tracts of
interest. AD describes diffusion along the longest principal axis of a ber. RD describes diffusion
perpendicular to the principle diffusion direction. FA describes the ratio of normalized standard deviation and root mean square of the three diffusion directions. The FA values range from 0 (equal
diffusion in all directions) to 1, with FA values of 0.2e1 indicating white matter and greater values
connoting greater white matter integrity (Kunimatsu et al., 2004). In our analysis, FA was used as the
main measure because it summarizes white matter microstructure. AD and RD values were used in a

Fig. 1. Illustration of tracts of interest in axial (left) and sagittal (right) views. Tracts were extracted from an example subject and
were registered to Montreal Neurological Imaging (MNI) template space. The tracts of interest include: Inferior Longitudinal
Fasciculus (ILF, red); Superior Longitudinal Fasciculus parietal bundle (SLFp, blue) and Arcuate Fasciculus (AF, purple).

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

19

post hoc analysis after we located tracts that were important for Mandarin learning, in order to better
characterize the underlying biological factors contributing to the inter-subject variability of the FA
values.
2.4. Statistical methods
We examined the relations between the participants' nal exam scores and three measures derived
from the FA values: 1) average FA across all three tracts of interest within each hemisphere; 2) average
FA within each tract of interest; and 3) laterality scores of each pair of tracts calculated as Laterality
Score (FAleft  FAright)/(FAleft FAright). More positive laterality scores indicate higher average FA
values in the left tract, i.e. left-lateralized. More negative laterality scores indicate higher average FA
values in the right tract, i.e. right-lateralized. Because FA values are bounded between 0 and 1 with
considerable skew and kurtosis (Alexander, Lee, Lazar, & Field, 2007), data normality was not assumed.
Therefore, for all the following analyses, we adopted a non-parametric method, Spearman's rank
correlation. All reported p values have been false detection rate (FDR) corrected for multiple comparisons (FDR < 0.05).
For analyses with unidirectional hypotheses (greater FA in the tracts for analysis are associated with
better learning performance), one-tailed statistical tests were used. For analyses where both directions
of hypothesis were possible (correlation between tract lateralization and learning performance), twotailed tests were used.
3. Results
3.1. Right but not left hemisphere tracts predicted Mandarin learning ability
The mean baseline FA of the three left-hemisphere tracts did not signicantly correlate with nal
exam scores (Spearman's r .24, one-tailed p .14; Fig. 2a), but there was a signicant correlation
between the mean baseline FA of the three right-hemisphere tracts and nal exam scores (Spearman's r .55, one-tailed p .005; Fig. 2b). The selective correlation with right-hemisphere pathways
occurred despite strong overall correlations between the means of the left and right pathways across
individuals (Spearman's r .65, two-tailed p .001). We directly compared the strengths of the leftand right-hemisphere correlations using a bootstrap procedure, which generated a condence interval based on 105 random resamplings in order to test whether they were statistically different

Fig. 2. Correlation between the average FA across three tracts of interest (ILF, SLFp and AF) and nal exam scores. The average FA
value of (a) the left tracts did not correlate with subsequent Mandarin learning success, while the average FA value of (b) the right
tracts did correlate signicantly and positively with subsequent Mandarin learning success. Black lines represent the linear trend of
the correlation and grey shades represent the 95% condence interval of the data.

20

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

from each other (DiCiccio & Efron, 1996). There was a signicantly stronger correlation between the
nal exam scores and FA values in the right hemisphere than those in the left hemisphere (one-tailed
p < .02).

3.2. Right ILF and right SLFp predicted Mandarin learning ability
Individual nal exam scores correlated positively with FA of the right ILF (Spearman's r .50, onetailed FDR-corrected, p .03; Fig. 3a) and the right SLFp (Spearman's r .60, one-tailed FDR-corrected,
p .01; Fig. 3b). Because nal exam scores were also marginally correlated with IQ (Spearman's r .32,
one-tailed p .08), we conducted a partial correlation analysis to control for individual IQ. The correlation between nal exam scores and FA of both tracts remained signicant (p's < .02). Individual IQ
scores and tract FA values (arcsin transformed in order to meet the normality assumption of the linear
regression analyses) were t into linear regression models for the prediction of the nal exam scores.
Compared with the model using IQ as the only predictor, adding either right SLFp or right ILF FA values
as a second predictor increased the variance of data explained by the model from 15.1% to 25.1% and
22.4%, respectively (model comparison p .04; p .06). There were also marginally signicant correlations between the FA values in the left AF, right AF, and left ILF and participants' nal exam scores
(uncorrected p's < .10, Table 1).
In order to further investigate the possible anatomical structure differences that contributed to the
inter-subject variability of FA values, we analyzed two additional diffusion parameters, AD and RD, in
the right ILF and right SLFp. The average AD did not correlate with individual nal exam scores in either
tract (Spearman's r's < .06, one-tailed p's > .28). However, the average RD signicantly correlated with

Fig. 3. Greater initial FA values of (a) the right ILF and (b) the right SLFp correlate with subsequent Mandarin learning success. In
both (a) and (b), black lines represent the linear trend of the correlation and grey shades represent the 95% condence interval of the
data. (c) Right ILF and (d) right SLFp were rendered in two example subjects (circled points) to illustrate the FA difference between a
less successful (Subj. 1) and a more successful (Subj. 2) learner. The tracts are colored according to the FA values.

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

21

individual nal exam score in both the right ILF (Spearman's r 0.56, one-tailed p 0.005) and the
right SLFp (Spearman's r 0.57, one-tailed p 0.004).
We also examined the specic relations of the right ILF and SLFp tracts to Mandarin phonology
learning (as measured by the pinyin dictation task in the nal exam) and orthography learning (as
measured by the two-alternative forced-choice task). Participants' performance on the nal phonology
measure correlated signicantly with baseline average FA values of both the right ILF (Spearman's
r 0.42, one-tailed p 0.03) and the right SLFp (Spearman's r 0.45, one-tailed p 0.02). Participants' performance on the nal orthography measure, however, correlated signicantly only with the
average FA values of the right ILF (Spearman's r 0.54, one-tailed p 0.006) and not the right SLFp
(Spearman's r 0.21, one-tailed p 0.18).
3.3. Laterality of ILF and SLFp predicted Mandarin learning ability
Given that right-hemisphere tracts played a greater role in predicting Mandarin learning efcacy
than left-hemisphere tracts, we further examined the relation between hemispheric asymmetry of
white-matter microstructure and Mandarin learning performance. By comparing the FA values between the two hemispheres for each pair of tracts of interest, we calculated laterality scores for each
tract in each participant. Participants displayed great variability in lateralization. More than 50% of the
participants showed right-lateralized ILF, whereas more than 50% showed left-lateralized SLFp (Fig. 4).
Greater right-lateralization of ILF FA correlated positively with superior nal exam scores (Spearman's
r .53, two-tailed FDR-corrected p .04; Fig. 4a) and greater right-lateralization of SLFp FA tended to
correlate with superior nal exam scores (Spearman's r .45, two-tailed FDR-corrected p .06;
Fig. 4b). There was no relation between nal exam scores and AF lateralization (Spearman's r .16,
two-tailed FDR-corrected p .49).
4. Discussion
In this study, we investigated the relations between initial white-matter microstructure and subsequent Mandarin learning success. Native English speakers enrolled in a one-month introductory
Mandarin course and reached a wide range of prociency at the end of the training. We found a
signicantly stronger positive correlation between students' nal exam scores and initial FA in righthemisphere tracts than in left-hemisphere tracts. In particular, higher FA in the right ILF and right SLFp
predicted greater success in learning Mandarin. As revealed by separate behavioral subtests, FA in both
right-hemisphere tracts was associated with phonology learning success, while FA in the right ILF was
associated with only orthography learning success. Analysis of hemispheric asymmetries of FA further
revealed that more rightward ILF and less leftward SLFp were associated with higher nal exam scores.
These results provide converging evidence for the hypothesis that the right-hemisphere tracts play a
more important role than left hemisphere tracts in preparing adults for initial learning of Mandarin as a
second language.
4.1. Tract laterality and handedness
The FA characterization of white-matter tracts is consistent with other studies. Previous research,
like our study, has shown that FA in homologous white matter structures are highly correlated in individuals (Wahl et al., 2010). Our laterality analysis of tract FA values showed left-lateralized SLF tracts

Table 1
Spearman correlations between white matter anisotropy and the nal exam score.

Correlation
p (Uncorrected)
p (FDR)

Left ILF

Left SLFp

Left AF

Right ILF

Right SLFp

Right AF

.29
.10
.12

.05
.41
.51

.35
.06
.12

.50
.01
.03

.60
.002
.01

.32
.08
.12

FDR false-discovery rate; Numbers in bold fonts indicate statistical signicance.

22

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

Fig. 4. Initial hemispheric asymmetry of tract FA values correlates with subsequent Mandarin learning success. (a) More rightlateralized ILF and (b) less left-lateralized SLFp were associated with higher nal exam scores. Black lines represent the linear
trend of the correlation and grey shades represent the 95% condence interval of the data.

in most participants, consistent with ndings of volume asymmetry of the tracts (Ellmore et al., 2010;
Glasser & Rilling, 2008; Saur et al., 2010). Because the incidence of right functional laterality in language increases with the degree of left-handedness (Basic et al., 2004; Lorenz et al., 2008), one possible
concern regarding the current data set is that our sample included three left-handers. However, the
tract laterality scores of all three left-handers showed left lateralized SLFp, similar to the majority of the
right-handers. We also reanalyzed the data excluding the three left-handers and found the same
signicant correlations between the right-hemisphere tracts and Mandarin learning. This result is
consistent with previous ndings of an overall leftward asymmetry of AF regardless of handedness or
functional language lateralization, as determined by the hemispheric difference in response to a verb
production task (Vernooij et al., 2007).
4.2. DTI analyses of white-matter microstructure
Higher FA values likely reect a number of different tract-specic properties, including ber
sche, &
diameter, ber density, membrane permeability, and myelination (Beaulieu, 2002; Jones, Kno
Turner, 2012), which can be associated with higher AD and/or lower RD (Song et al., 2002;
Takahashi et al., 2002). Our ndings showed it was RD, but not AD, that carried the correlation between the white matter microstructure and Mandarin learning success. This nding suggests that
increased axonal myelination or denser axonal tracts (lower RD), rather than axonal integrity measured
by AD, supported superior Mandarin learning.
There are limitations to DTI analyses of white-matter tracts. First, diffusion tensor modeling is not
able to solve the problem of intra-voxel crossing bers. Reduced RD and increased FA might result from
fewer crossing bers in successful learners. Therefore, the precise neurobiological basis of the FA
correlations cannot be determined. Second, there are limitations to the accurate delineation of tracts
by TRACULA. TRACULA reconstructs white matter into 18 tracts, based on the anatomical regions of
interest dened in Makris et al. (2005) (Yendiki et al., 2011). The SLF is divided into AF and SLFp. The AF
in this study corresponds best to the classically dened direct pathway of the arcuate fasciculus,
while the SLFp is closest to the anterior segment of the indirect pathway (Catani et al., 2005).
TRACULA does not, however, differentiate the posterior segment of the indirect pathway from the
posterior part of the AF, both of which connect the posterior parietal area with the posterior temporal
area. Likewise, TRACULA identies ILF as one of the ventral stream pathways. It is possible, however
that the ILF includes bers from other language-related ventral pathways, such as the middle

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

23

longitudinal fasciculus and inferior fronto-occipital fasciculus (Dick & Tremblay, 2012; Saur et al.,
2008; Wong et al., 2011).
4.3. Right-hemisphere contribution to Mandarin learning
In the present study, white-matter diffusivity measures in the right hemisphere predicted Mandarin
learning. On the surface, these ndings may seem to contradict the dominant involvement of left
el, de Vries, Scholz,
hemisphere tracts in learning new phonemes, words, and grammar in adults (Flo
 pez-Barroso et al., 2013; Wong et al., 2011; see review:
Breitenstein, & Johansen-Berg, 2009; Lo
Friederici, 2009). Prior studies found associations between diffusivity measures in the left hemisphere,
but not in the right hemisphere, and language learning performance. However, those studies focused
on European languages without the tonal properties of Mandarin or the non-alphabetic visuo-spatial
properties of Chinese characters.
Neuroimaging studies have found evidence suggesting that the brain areas connected by the ILF
and the SLFp in the right hemisphere participate in processing pitch information of Mandarin
speech. In a Chinese character naming task, native Mandarin speakers recruited the right inferior
frontal gyrus more strongly in the production of characters with different tones than the production
of characters with different vowels (Liu et al., 2006). Moreover, the engagement of the right
hemisphere in early attention to pitch information in Mandarin appears to be a shared mechanism
irrespective of language experience. A rightward asymmetry of activation in the superior temporal
sulcus and the middle frontal gyrus was found in both Mandarin and English native speakers performing same-different judgment on Mandarin intonation across syllables and on tones within
syllables (Gandour et al., 2004). Similarly, a rightward asymmetry in the middle frontal gyrus was
reported in both Mandarin and English native speakers processing sentence-level prosody of
Mandarin (Tong et al., 2005). The most relevant evidence comes from a training study in native
English speakers who were nave to Mandarin. The level of activation in the right temporal gyrus, as
well as the left temporal gyrus, during a tone discrimination task predicted whether an individual
was successful in learning new pseudowords with Mandarin lexical tones (Wong, Perrachione, &
Parrish, 2007).
White-matter studies provide evidence for a structural counterpart to the right-hemisphere
functional specialization for pitch processing. A DTI-identied pathway between superior/middle
temporal gyrus and the frontal lobe is located in right-hemisphere regions associated with prosodic
processing in European languages (Glasser & Rilling, 2008). The reduction of right SLF volume was
found to be associated with the behavioral deciencies of tone-deafness (Loui et al., 2009). Moreover,
there was a positive correlation between FA of the right temporal-parietal conjunction and individual's ability to learn an articial pitch-related grammar (Loui et al., 2011). Further, crosssectional studies discovered right-hemisphere structural characteristics specic to Chinese
speakers. Higher FA in the right superior temporal and the right lingual gyrus white-matter regions
were positively correlated with faster reading time of English words in a Chinese-English bilingual
group, but not in an English monolingual group (Cummine & Boliek, 2013). Increased white-matter
volume in the tract arising from the right Heschl's gyrus was found in multilingual Mandarin
speakers and native Mandarin speakers, but not in European multilinguals who did not speak
Mandarin (Crinion et al., 2009). Overall, these studies support the idea that due to the tonal nature of
Mandarin speech, there is greater right-hemisphere involvement for Mandarin speech than for nontonal European languages.
The other challenge Mandarin learners usually face is learning to read Chinese characters, which
have great visuo-spatial complexity. Previous studies showed greater activation of the right middle
occipital and right fusiform regions in Chinese readers than English readers when reading in their
native languages (Bolger et al., 2005; Tan et al., 2005). English speakers learning Mandarin increased
recruitment of the right fusiform regions when viewing Chinese characters (Nelson, Liu, Fiez, &
Perfetti, 2009). Our nding that learners' performance in the orthography task positively associated
with their FA values of the right ILF tract before training is consistent with these ndings. It is noteworthy that only the properties of the right ILF, which is associated with visual processes, and not the
right SLFp, which is more associated with auditory processes, had a signicant association with

24

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

orthography learning in Mandarin Chinese. In Mandarin, semantic analysis of Chinese characters is

difcult via phonology because there is a lack of transparent mapping between phonemes and
graphemes, and there are many homophones per single-syllable word. Therefore, direct mapping of
orthographic information to meaning may be efcient, especially in early learning. The correlation of
the right ILF, but not the right SLFp, with superior orthographic learning is consistent with the idea that
early learning of Chinese characters is more related to visual and less to phonological processes than
that occurs in alphabetic languages.
Neuropsychological evidence also supports the possibility that learning tonal and visuo-spatial
aspects of Mandarin may invoke right-hemisphere brain regions. Damage to the right auditory cortex leads to impairment in pitch-related calculation, including production and perception of speech
prosody (Baum & Pell, 1999; Loui et al., 2009; Pell, 1998; Warrier & Zatorre, 2004). Lesion at the right
posterior parietal regions and occipital regions leads to disruption of visuo-spatial attention network
(Halligan, Fink, Marshall, & Vallar, 2003; Mosidze, Mkheidze, & Makashvili, 1994; Pisella et al., 2011).
Both the right dorsal and ventral visual streams have been implicated to play important roles in the
cognitive control of visual search (Lane, Groisman, & Ferreira, 2006).
An alternative interpretation of our nding is that the microstructure of white matter in the right
hemisphere is involved not only in Mandarin learning, but also second-language learning in general. A
behavioral meta-analyses study of bilingual laterality found that second language prociency in late
bilinguals was positively correlated with right-hemisphere involvement (Hull & Vaid, 2007). The
extent of right-hemisphere involvement appears to be modulated by the similarity between L1 and L2.
A dichotic behavioral test with English words as L2 showed reduced leftward asymmetry in Italian
speakers, whose native language is less similar to English compared to German speakers (D'Anselmo,
Reiterer, Zuccarini, Tommasi, & Brancucci, 2013).
The contribution of right hemisphere in bilingualism is further supported by a number of neuroimaging studies. An fMRI study revealed more bilateral activation elicited by children who processed
English speech as a second language compared to those who processed English speech as a native
language (Archila-Suerte, Zevin, Ramos, & Hernandez, 2013). Similar ndings from a MEG study
suggested greater right visual word form area activation and therefore more bilateral recruitment in L2
(English) word processing compared to L1 (Spanish, Leonard et al., 2010). Structural MRI studies found
that increased exposure to second language is associated with greater grey matter density in the right
hippocampus (Mrtensson et al., 2012). Compared to monolingual English speakers, early bilingual
speakers of two alphabetical languages showed stronger anterior-posterior structural and functional
connectivity within both the left and right hemispheres (Luk, Bialystok, Craik, & Grady, 2011). Early
bilingual speakers also showed increased structural connectivity in a subnetwork involving the right
 n, Pe
rez
superior frontal gyrus and the left reading network measured by graph efciency (Garca-Pento
ndez, Iturria-Medina, Gillon-Dowens, & Carreiras, 2014). One longitudinal study revealed
Ferna
learning-induced change of right-hemisphere structure within individuals. Japanese speakers showed
enhancement of grey matter and white matter density in the right IFG after learning English (Hosoda,
Tanaka, Nariai, Honda, & Hanakawa, 2013).
Together, these studies suggest right-hemisphere involvement in second-language learning may
not be limited to Mandarin learning. The cognitive control demands of adapting to new linguistic rules,
as well as selecting, switching, and translating between native and foreign languages might lead to
recruiting areas outside of the network for one's native language, such as bilateral prefrontal and
parietal areas (Abutalebi & Green, 2007; Li, Legault & Litcofsky, 2014; Luk & Green, 2011). The present
study is not able to differentiate between these two alternative explanations. Only an experiment with
another control group of participants learning an alphabetic language could test the specic role of the
right hemisphere in Mandarin learning.
Finally, it is noteworthy that the association between white matter structure and Mandarin learning
success revealed in the current study is based on the prociency tests after only one month of training.
Language learning at a more advanced stage may involve distinct underlying processes from earlier
stages, and therefore may show a different relationship with brain structure. The importance of lefthemisphere engagement in uent Mandarin has been implicated by the ndings that native Mandarin speakers show left-lateralized functional representation of language (Gandour et al., 2004; Hsieh,
Gandour, Wong, & Hutchins, 2001; Klein, Zatorre, Milner, & Zhao, 2001; Tong et al., 2005; Xu et al.,

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

25

2006). Thus, left-hemisphere networks and tracts may become increasingly important as learners'
prociency improves. More studies are necessary to unveil the complete picture of the neurobiology of
second language learning.
5. Conclusion
It has long been noted that adult learners of a second language often achieve quite different levels of
prociency. The current study presented a right-lateralized association between white-matter
microstructure and successful Mandarin learning, with higher FA in the right SLFp and right ILF predicting superior learning in a real language-learning environment. Our results also indicated a greater
contribution of right hemisphere than left hemisphere in successful Mandarin learning in the initial
phase of acquisition.
Acknowledgment
We thank Allyson Mackey for help with data analysis and the manuscript; Satra Ghosh, Jack Murtagh, and David Osher for their intellectual input on data analysis; Amy Finn for her valuable advice on
the project design; and Carlo de los Angeles, Qi Liu, and Valkyrie Felso for their great contribution in
data collection. We thank the staff of the Athinoula A. Martinos Imaging Center at McGovern Institute
for Brain Research, Massachusetts Institute of Technology for their technical support. This work was
funded by DID DI-MISC 80508B, Contract No. H98230-07-D-0175.
References
Abutalebi, J., & Green, D. (2007). Bilingual language production: the neurocognition of language representation and control.
Journal of Neurolinguistics, 20, 242e275. http://dx.doi.org/10.1016/j.jneuroling.2006.10.003.
Alexander, A. L., Lee, J. E., Lazar, M., & Field, A. S. (2007). Diffusion tensor imaging of the brain. Neurotherapeutics, 4(3), 316e329.
sche, T. R. (2007). Connectivity-based Parcellation of
Anwander, A., Tittgemeyer, M., von Cramon, D. Y., Friederici, A. D., & Kno
Broca's area. Cerebral Cortex (New York, N.Y.: 1991), 17(4), 816e825. http://dx.doi.org/10.1093/cercor/bhk034.
Archila-Suerte, P., Zevin, J., Ramos, A. I., & Hernandez, A. E. (2013). The neural basis of non-native speech perception in bilingual
children. NeuroImage, 67, 51e63. http://dx.doi.org/10.1016/j.neuroimage.2012.10.023.
Basic, S., Hajnsek, S., Poljakovic, Z., Basic, M., Culic, V., & Zadro, I. (2004). Determination of cortical language dominance using
functional transcranial Doppler sonography in left-handers. Clinical Neurophysiology, 115(1), 154e160. http://dx.doi.org/10.
1016/S1388-2457(03)00281-5.
Baum, S. R., & Pell, M. D. (1999). The neural bases of prosody: insights from lesion studies and neuroimaging. Aphasiology, 13(8),
581e608. http://dx.doi.org/10.1080/026870399401957.
Beaulieu, C. (2002). The basis of anisotropic water diffusion in the nervous system - a technical review. NMR in Biomedicine,
15(7e8), 435e455. http://dx.doi.org/10.1002/nbm.782.
Bolger, D. J., Perfetti, C. A., & Schneider, W. (2005). Cross-cultural effect on the brain revisited: universal structures plus writing
system variation. Human Brain Mapping, 25(1), 92e104. http://dx.doi.org/10.1002/hbm.20124.
Catani, M., Jones, D. K., & Ffytche, D. H. (2005). Perisylvian language networks of the human brain. Annals of Neurology, 57(1),
8e16. http://dx.doi.org/10.1002/ana.20319.
Cohen, L., & Dehaene, S. (2004). Specialization within the ventral stream: the case for the visual word form area. NeuroImage,
22(1), 466e476. http://dx.doi.org/10.1016/j.neuroimage.2003.12.049.
Crinion, J. T., Green, D. W., Chung, R., Ali, N., Grogan, A., Price, G. R., et al. (2009). Neuroanatomical markers of speaking Chinese.
Human Brain Mapping, 30(12), 4108e4115. http://dx.doi.org/10.1002/hbm.20832.
Cummine, J., & Boliek, C. A. (2013). Understanding white matter integrity stability for bilinguals on language status and reading
performance. Brain Structure & Function, 218(2), 595e601. http://dx.doi.org/10.1007/s00429-012-0466-6.
D'Anselmo, A., Reiterer, S., Zuccarini, F., Tommasi, L., & Brancucci, A. (2013). Hemispheric asymmetries in bilinguals: tongue
similarity affects lateralization of second language. Neuropsychologia, 51(7), 1187e1194. http://dx.doi.org/10.1016/j.neuropsychologia.2013.03.016.
DiCiccio, T. J., & Efron, B. (1996). Bootstrap condence intervals. Statistical Science, 11(3), 189e228.
Dick, A. S., & Tremblay, P. (2012). Beyond the arcuate fasciculus: consensus and controversy in the connectional anatomy of
language. Brain: A Journal of Neurology, 135(Pt 12), 3529e3550. http://dx.doi.org/10.1093/brain/aws222.
Dronkers, N. F., Plaisant, O., Iba-Zizen, M. T., & Cabanis, E. A. (2007). Paul Broca's historic cases: high resolution MR imaging of
the brains of Leborgne and Lelong. Brain: A Journal of Neurology, 130(Pt 5), 1432e1441. http://dx.doi.org/10.1093/brain/
awm042.
Ellmore, T. M., Beauchamp, M. S., Breier, J. I., Slater, J. D., Kalamangalam, G. P., O'Neill, T. J., et al. (2010). Temporal lobe white
matter asymmetry and language laterality in epilepsy patients. NeuroImage, 49(3), 2033e2044. http://dx.doi.org/10.1016/j.
neuroimage.2009.10.055.
Epelbaum, S., Pinel, P., Gaillard, R., Delmaire, C., Perrin, M., Dupont, S., et al. (2008). Pure alexia as a disconnection syndrome:
new diffusion imaging evidence for an old concept. Cortex; a Journal Devoted to the Study of the Nervous System and Behavior,
44(8), 962e974. http://dx.doi.org/10.1016/j.cortex.2008.05.003.

26

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

el, A., de Vries, M. H., Scholz, J., Breitenstein, C., & Johansen-Berg, H. (2009). White matter integrity in the vicinity of Broca's
Flo
area predicts grammar learning success. NeuroImage, 47(4), 1974e1981. http://dx.doi.org/10.1016/j.neuroimage.2009.05.
046.
Friederici, A. D. (2009). Pathways to language: ber tracts in the human brain. Trends in Cognitive Sciences, 13(4), 175e181.
http://dx.doi.org/10.1016/j.tics.2009.01.001.
Gandour, J., Tong, Y., Wong, D., Talavage, T., Dzemidzic, M., Xu, Y., et al. (2004). Hemispheric roles in the perception of speech
prosody. NeuroImage, 23(1), 344e357. http://dx.doi.org/10.1016/j.neuroimage.2004.06.004.
n, L., Pe
rez Ferna
ndez, A., Iturria-Medina, Y., Gillon-Dowens, M., & Carreiras, M. (2014). Anatomical connectivity
Garca-Pento
changes in the bilingual brain. NeuroImage, 84, 495e504. http://dx.doi.org/10.1016/j.neuroimage.2013.08.064.
Glasser, M. F., & Rilling, J. K. (2008). DTI tractography of the human brain's language pathways. Cerebral Cortex (New York, N.Y.:
1991), 18(11), 2471e2482. http://dx.doi.org/10.1093/cercor/bhn011.
Gu, F., Zhang, C., Hu, A., & Zhao, G. (2013). Left hemisphere lateralization for lexical and acoustic pitch processing in Cantonese
speakers as revealed by mismatch negativity. NeuroImage, 83, 637e645.
Halligan, P. W., Fink, G. R., Marshall, J. C., & Vallar, G. (2003). Spatial cognition: evidence from visual neglect. Trends in Cognitive
Sciences, 7(3), 125e133. http://dx.doi.org/10.1016/S1364-6613(03)00032-9.
Hickok, G., & Poeppel, D. (2004). Dorsal and ventral streams: a framework for understanding aspects of the functional anatomy
of language. Cognition, 92(1-2), 67e99. http://dx.doi.org/10.1016/j.cognition.2003.10.011.
Hickok, G., & Poeppel, D. (2007). The cortical organization of speech processing. Nature Reviews Neuroscience, 8(May), 393e402.
Hosoda, C., Tanaka, K., Nariai, T., Honda, M., & Hanakawa, T. (2013). Dynamic neural network reorganization associated with
second language vocabulary acquisition: a multimodal imaging study. The Journal of Neuroscience: The Ofcial Journal of the
Society for Neuroscience, 33(34), 13663e13672. http://dx.doi.org/10.1523/JNEUROSCI.0410-13.2013.
Hsieh, L., Gandour, J., Wong, D., & Hutchins, G. D. (2001). Functional heterogeneity of inferior frontal gyrus is shaped by linguistic experience. Brain and Language, 76(3), 227e252. http://dx.doi.org/10.1006/brln.2000.2382.
Hull, R., & Vaid, J. (2007). Bilingual language lateralization: a meta-analytic tale of two hemispheres. Neuropsychologia, 45(9),
1987e2008. http://dx.doi.org/10.1016/j.neuropsychologia.2007.03.002.
Johnson, J. S., & Newport, E. L. (1989). Critical period effects in second language learning: the inuence of maturational state on
the acquisition of English as a second language. Cognitive Psychology, 21(1), 60e99.
sche, T. R., & Turner, R. (2012). White matter integrity, ber count, and other fallacies: the do's and don'ts of
Jones, D. K., Kno
diffusion MRI. NeuroImage. http://dx.doi.org/10.1016/j.neuroimage.2012.06.081.
Kaufman, A. S., & Kaufman, N. L. (2004). Kaufman brief intelligence test (2nd ed.). Pearson.
Klein, D., Zatorre, R. J., Milner, B., & Zhao, V. (2001). A cross-linguistic PET study of tone perception in Mandarin Chinese and
English speakers. NeuroImage, 13(4), 646e653. http://dx.doi.org/10.1006/nimg.2000.0738.
Kuhl, P. K. (2004). Early language acquisition: cracking the speech code. Nature Reviews Neuroscience, 5(11), 831e843. http://dx.
doi.org/10.1038/nrn1533.
Kunimatsu, A., Aoki, S., Masutani, Y., Abe, O., Hayashi, N., Mori, H., et al. (2004). The optimal trackability threshold of fractional
anisotropy for diffusion tensor tractography of the corticospinal tract. Magnetic Resonance in Medical Sciences, 3, 11e17.
Lane, L., Groisman, M., & Ferreira, V. (2006). Don't talk about Pink Elephants!: speakers' control over leaking private information
during language production. Psychological Science, 17(4), 273e277.
Leonard, M. K., Brown, T. T., Travis, K. E., Gharapetian, L., Hagler, D. J., Dale, A. M., et al. (2010). Spatiotemporal dynamics of
bilingual word processing. NeuroImage, 49(4), 3286e3294. http://dx.doi.org/10.1016/j.neuroimage.2009.12.009.
Li, P., Legault, J., & Litcofsky, K. A. (2014). Neuroplasticity as a function of second language learning: anatomical changes in the
human brain. Cortex, 58, 301e324. http://dx.doi.org/10.1016/j.cortex.2014.05.001.
Linck, J. a., Hughes, M. M., Campbell, S. G., Silbert, N. H., Tare, M., Jackson, S. R., et al. (2013). Hi-LAB: a new measure of aptitude
for high-level language prociency. Language Learning, 63(3), 530e566. http://dx.doi.org/10.1111/lang.12011.
Liu, L., Peng, D., Ding, G., Jin, Z., Zhang, L., Li, K., et al. (2006). Dissociation in the neural basis underlying Chinese tone and vowel
production. NeuroImage, 29(2), 515e523. http://dx.doi.org/10.1016/j.neuroimage.2005.07.046.
Liu, X. (2010). In (2nd ed.)New practical Chinese reader (2nd ed.), (Vol. 1). Beijing Language Culture University Press.
pez-Barroso, D., Catani, M., Ripolle
s, P., Dell'acqua, F., Rodrguez-Fornells, A., & de Diego-Balaguer, R. (2013). Word learning is
Lo
mediated by the left arcuate fasciculus. Proceedings of the National Academy of Sciences of the United States of America. http://
dx.doi.org/10.1073/pnas.1301696110.
Lorenz, M. W., Thoelen, N., Loesel, N., Lienerth, C., Gonzalez, M., Humpich, M., et al. (2008). Assessment of cerebral autoregulation with transcranial Doppler sonography in poor bone windows using constant infusion of an ultrasound contrast
agent. Ultrasound in Medicine & Biology, 34(3), 345e353. http://dx.doi.org/10.1016/j.ultrasmedbio.2007.09.001.
Loui, P., Alsop, D., & Schlaug, G. (2009). Tone deafness: a new disconnection syndrome? The Journal of Neuroscience : The Ofcial
Journal of the Society for Neuroscience, 29(33), 10215e10220. http://dx.doi.org/10.1523/JNEUROSCI.1701-09.2009.
Loui, P., Li, H. C., & Schlaug, G. (2011). White matter integrity in right hemisphere predicts pitch-related grammar learning.
NeuroImage, 55(2), 500e507. http://dx.doi.org/10.1016/j.neuroimage.2010.12.022.
Luk, G., Bialystok, E., Craik, F. I. M., & Grady, C. L. (2011). Lifelong bilingualism maintains white matter integrity in older adults.
The Journal of Neuroscience : The Ofcial Journal of the Society for Neuroscience, 31(46), 16808e16813. http://dx.doi.org/10.
1523/JNEUROSCI.4563-11.2011.
Luk, G., & Green, D. (2011). Cognitive control for language switching in bilinguals: a quantitative meta-analysis of functional
neuroimaging studies. Language and Cognitive Processes, 27(10), 1479e1488. http://dx.doi.org/10.1080/01690965.2011.
613209.Cognitive.
Makris, N., Kennedy, D. N., McInerney, S., Sorensen, A. G., Wang, R., Caviness, V. S., et al. (2005). Segmentation of subcomponents
within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cerebral Cortex (New York, N.Y. :
1991), 15(6), 854e869. http://dx.doi.org/10.1093/cercor/bhh186.
Mrtensson, J., Eriksson, J., Bodammer, N. C., Lindgren, M., Johansson, M., Nyberg, L., et al. (2012). Growth of language-related brain
areas after foreign language learning. NeuroImage, 63(1), 240e244. http://dx.doi.org/10.1016/j.neuroimage.2012.06.043.

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

27

Meyer, M., Alter, K., Friederici, A. D., Lohmann, G., & von Cramon, D. Y. (2002). FMRI reveals brain regions mediating
slow prosodic modulations in spoken sentences. Human Brain Mapping, 17(2), 73e88. http://dx.doi.org/10.1002/hbm.
10042.
Mohades, S. G., Struys, E., Van Schuerbeek, P., Mondt, K., Van Den Craen, P., & Luypaert, R. (2012). DTI reveals structural differences in white matter tracts between bilingual and monolingual children. Brain Research, 1435, 72e80.
Mosidze, V. M., Mkheidze, R. A., & Makashvili, M. A. (1994). Disorders of visuo-spatial attention in patients with unilateral brain
damage. Behavioural Brain Research, 65(1), 121e122. http://dx.doi.org/10.1016/0166-4328(94)90081-7.
Nelson, J. R., Liu, Y., Fiez, J., & Perfetti, C. A. (2009). Assimilation and accommodation patterns in ventral occipitotemporal cortex
in learning a second writing system. Human Brain Mapping, 30(3), 810e820. http://dx.doi.org/10.1002/hbm.20551.
Oguz, I., Farzinfar, M., Matsui, J., Budin, F., Liu, Z., Gerig, G., et al. (2014). DTIPrep: quality control of diffusion-weighted images.
Frontiers in Neuroinformatics, 8, 4. http://dx.doi.org/10.3389/fninf.2014.00004.
Pell, M. D. (1998). Recognition of prosody following unilateral brain lesion: inuence of functional and structural attributes of
prosodic contours. Neuropsychologia, 36(8), 701e715. http://dx.doi.org/10.1016/S0028-3932(98)00008-6.
Perani, D., Paulesu, E., Galles, N. S., Dupoux, E., Dehaene, S., Bettinardi, V., et al. (1998). The bilingual brain prociency and age of
acquisition of the second language (pp. 1841e1852).
Perfetti, C. A., Liu, Y., & Tan, L. H. (2005). The lexical constituency model: some implications of research on Chinese for general
theories of reading. Psychological Review, 112(1), 43e59.
Pisella, L., Alahyane, N., Blangero, A., Thery, F., Blanc, S., & Pelisson, D. (2011). Right-hemispheric dominance for visual
remapping in humans. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 366(1564),
572e585. http://dx.doi.org/10.1098/rstb.2010.0258.
Pugh, K. R., Mencl, W. E., Jenner, A. R., Katz, L., Frost, S. J., Lee, J. R., et al. (2001). Neurobiological studies of reading and reading
disability. Journal of Communication Disorders, 34(6), 479e492.
Rauschecker, A. M., Deutsch, G. K., Ben-Shachar, M., Schwartzman, A., Perry, L. M., & Dougherty, R. F. (2009). Reading impairment in a patient with missing arcuate fasciculus. Neuropsychologia, 47(1), 180e194. http://dx.doi.org/10.1016/j.neuropsychologia.2008.08.011.
Sakai, K. L. (2005). Language acquisition and brain development. Science (New York, N.Y.), 310(5749), 815e819. http://dx.doi.org/
10.1126/science.1113530.
Saur, D., Kreher, B. W., Schnell, S., Kmmerer, D., Kellmeyer, P., Vry, M.-S., et al. (2008). Ventral and dorsal pathways for language.
Proceedings of the National Academy of Sciences of the United States of America, 105(46), 18035e18040. http://dx.doi.org/10.
1073/pnas.0805234105.
Saur, D., Schelter, B., Schnell, S., Kratochvil, D., Kpper, H., Kellmeyer, P., et al. (2010). Combining functional and anatomical
connectivity reveals brain networks for auditory language comprehension. NeuroImage, 49(4), 3187e3197. http://dx.doi.org/
10.1016/j.neuroimage.2009.11.009.
Saygin, Z. M., Norton, E. S., Osher, D. E., Beach, S. D., Cyr, A. B., Ozernov-Palchik, O., et al. (2013). Tracking the roots of reading
ability: white matter volume and integrity correlate with phonological awareness in prereading and early-reading
kindergarten children. The Journal of Neuroscience : The Ofcial Journal of the Society for Neuroscience, 33(33),
13251e13258. http://dx.doi.org/10.1523/JNEUROSCI.4383-12.2013.
Schlegel, A. a., Rudelson, J. J., & Tse, P. U. (2012). White matter structure changes as adults learn a second language. Journal of
Cognitive Neuroscience, 24(8), 1664e1670. http://dx.doi.org/10.1162/jocn_a_00240.
Song, S. K., Sun, S. W., Ramsbottom, M. J., Chang, C., Russell, J., & Cross, A. H. (2002). Dysmyelination revealed through MRI as
increased radial (but unchanged axial) diffusion of water. NeuroImage, 17, 1429e1436.
Sun, Y., Yang, Y., Desroches, A. S., Liu, L., & Peng, D. (2011). The role of the ventral and dorsal pathways in reading Chinese
characters and English words. Brain and Language, 119(2), 80e88. http://dx.doi.org/10.1016/j.bandl.2011.03.012.
Takahashi, M., Hackney, D. B., Zhang, G. X., Wehrli, S. L., Wright, A. C., & O'Brien, W. T. (2002). Magnetic resonance microimaging
of intraaxonal water diffusion in live excised lamprey spinal cord. Proceedings of the National Academy of Sciences of the
United States of America, 99, 16192e16196.
Tan, L. H., Laird, A. R., Li, K., & Fox, P. T. (2005). Neuroanatomical correlates of phonological processing of Chinese characters and
alphabetic words: a meta-analysis. Human Brain Mapping, 25(1), 83e91. http://dx.doi.org/10.1002/hbm.20134.
Tan, L. H., Hoosain, R., & Peng, D. (1995). Role of early Presemantic phonological code in Chinese character identication. Journal
of Experimental Psychology: Learning, Memory, and Cognition, 21(1), 43e54.
Tan, L. H., Chen, L., Yip, V., Chan, A. H. D., Yang, J., Gao, J. H., et al. (2011). Activity levels in the left hemisphere caudate-fusiform
circuit predict how well a second language will be learned. Proceedings of the National Academy of Sciences of the United
States of America, 108(6), 2540e2544. http://dx.doi.org/10.1073/pnas.0909623108.
Tong, Y., Gandour, J., Talavage, T., Wong, D., Dzemidzic, M., Xu, Y., et al. (2005). Neural circuitry underlying sentence-level
linguistic prosody. NeuroImage, 28(2), 417e428. http://dx.doi.org/10.1016/j.neuroimage.2005.06.002.
Tracy, D. K., Ho, D. K., O'Daly, O., Michalopoulou, P., Lloyd, L. C., Dimond, E., et al. (2011). It's not what you say but the way that
you say it: an fMRI study of differential lexical and non-lexical prosodic pitch processing. BMC Neuroscience, 12, 128.
re, P. (2012). A tractography study in dyslexia:
Vandermosten, M., Boets, B., Poelmans, H., Sunaert, S., Wouters, J., & Ghesquie
neuroanatomic correlates of orthographic, phonological and speech processing. Brain : A Journal of Neurology, 135(Pt 3),
935e948. http://dx.doi.org/10.1093/brain/awr363.
 Rodrguez-Pujadas, A., Sebastia
n, A., Gonza
lez, J., Palomar-Garca, M.-A.,
n-Galle
s, N., et al. (2013).
Ventura-Campos, N., Sanjua
Spontaneous brain activity predicts learning ability of foreign sounds. The Journal of Neuroscience, 33(22), 9295e9305.
http://dx.doi.org/10.1523/JNEUROSCI.4655-12.2013.
Vernooij, M. W., Smits, M., Wielopolski, P. A., Houston, G. C., Krestin, G. P., & van der Lugt, A. (2007). Fiber density asymmetry of
the arcuate fasciculus in relation to functional hemispheric language lateralization in both right- and left-handed healthy
subjects: a combined fMRI and DTI study. Neuroimage, 35(3), 1064e1076.
Wahl, M., Li, Y.-O., Ng, J., Lahue, S. C., Cooper, S. R., Sherr, E. H., et al. (2010). Microstructural correlations of white matter tracts in
the human brain. NeuroImage, 51(2), 531e541. http://dx.doi.org/10.1016/j.neuroimage.2010.02.072.
Wakana, S., Caprihan, A., Panzenboeck, M. M., Fallon, J. H., Perry, M., Gol- lub, R. L., et al. (2007). Reproducibility of quantitative
tractography meth- ods applied to cerebralwhite matter. Neuroimage, 36, 630e644.

28

Z. Qi et al. / Journal of Neurolinguistics 33 (2015) 14e28

Wang, X. D., Wang, M., & Chen, L. (2013). Hemispheric lateralization for early auditory processing of lexical tones: dependence
on pitch level and pitch contour. Neuropsychologia, 51(11), 2238e2244.
Warrier, C. M., & Zatorre, R. J. (2004). Right temporal cortex is critical for utilization of melodic contextual cues in a pitch
constancy task. Brain : A Journal of Neurology, 127(Pt 7), 1616e1625. http://dx.doi.org/10.1093/brain/awh183.
Wildgruber, D., Pihan, H., Ackermann, H., Erb, M., & Grodd, W. (2002). Dynamic brain activation during processing of emotional
intonation: inuence of acoustic parameters, emotional valence, and sex. NeuroImage, 15(4), 856e869. http://dx.doi.org/10.
1006/nimg.2001.0998.
Wong, F. C. K., Chandrasekaran, B., Garibaldi, K., & Wong, P. C. M. (2011). White matter anisotropy in the ventral language
pathway predicts sound-to-word learning success. The Journal of Neuroscience, 31(24), 8780e8785. http://dx.doi.org/10.
1523/JNEUROSCI.0999-11.2011.
Wong, P. C. M. (2002). Hemispheric specialization of linguistic pitch patterns. Brain Research Bulletin, 59(2), 83e95. http://dx.
doi.org/10.1016/S0361-9230(02)00860-2.
Wong, P. C. M., Perrachione, T. K., & Parrish, T. B. (2007). Neural characteristics of successful and less successful speech and word
learning in adults. Human Brain Mapping, 28(10), 995e1006. http://dx.doi.org/10.1002/hbm.20330.
Xu, Y., Gandour, J., Talavage, T., Wong, D., Dzemidzic, M., Tong, Y., et al. (2006). Activation of the left planum temporale in pitch
processing is shaped by language experience. Human Brain Mapping, 27(2), 173e183. http://dx.doi.org/10.1002/hbm.20176.
Yeatman, J. D., Dougherty, R. F., Ben-Shachar, M., & Wandell, B. A. (2012). Development of white matter and reading skills.
Proceedings of the National Academy of Sciences of the United States of America, 109(44), E3045eE3053. http://dx.doi.org/10.
1073/pnas.1206792109.
Yeatman, J. D., Dougherty, R. F., Rykhlevskaia, E., Sherbondy, A. J., Deutsch, G. K., Wandell, B. A., et al. (2011). Anatomical
properties of the arcuate fasciculus predict phonological and reading skills in children. Journal of Cognitive Neuroscience,
23(11), 3304e3317. http://dx.doi.org/10.1162/jocn_a_00061.
llei, L., Augustinack, J., et al. (2011). Automated probabilistic reconstruction
Yendiki, A., Panneck, P., Srinivasan, P., Stevens, A., Zo
of white-matter pathways in health and disease using an atlas of the underlying anatomy. Frontiers in Neuroinformatics, 5,
23. http://dx.doi.org/10.3389/fninf.2011.00023.
Zatorre, R. J., & Gandour, J. T. (2008). Neural specializations for speech and pitch: moving beyond the dichotomies. Philosophical
Transactions of the Royal Society of London. Series B, Biological Sciences, 363(1493), 1087e1104. http://dx.doi.org/10.1098/rstb.
2007.2161.