Neural specialization for print in Chinese-English language learners

Highlights

• This study focused on print specialization in Chinese-English second graders.

• The second graders showed clear N1 specialization for both Chinese and English.

• The N1 specialization for Chinese words appeared to be bilateral.

• The N1 specialization for English words appeared to be right-lateralized.

Abstract

Visual specialization for print is an essential structural element of fluent reading. Previous research suggests that lateralization of N1 print tuning may depend on both the writing system and the skill level. Here, we tested which of these effects was more prominent in beginning readers of Chinese (L1) and English (L2). Using an ERP measure with an implicit one-back repetition detection task, we investigated the neural basis of visual specialization for print in Chinese and English among eleven 7-to-8-year-old Hong Kong children, who were learning to read Chinese and English in parallel, though showing lower skills in English. N1 print tuning was found for both Chinese and English words, but was bilateral for Chinese and right-lateralized for English. In addition, the specialization effect of print started as early as in the time window of the P1, with a reduced P1 for Chinese and English words as compared to the visually matched control stimuli. These findings suggest that Chinese readers in 2nd grade show clear visual specialization for both Chinese and English print and that right-lateralized N1 specialization may be driven by less reading experience in English rather than by writing-specific effects induced by Chinese.

Introduction

Word reading is an interactive process involving multiple cognitive processes associated with a complex brain network (e.g., McCandliss et al., 2003, Posner et al., 1988, Schlaggar and McCandliss, 2007, Seidenberg and McClelland, 1989). Among those cognitive processes, visual specialization for print, which could be assumed to represent a type of perceptual expertise (Brem, Lang-Dullenkopf, Maurer, Halder, Bucher,& Brandeis, 2005), is crucial for fluent reading across writing systems (e.g., Cao et al., 2011, Maurer et al., 2005; Maurer et al., 2006; Xue, Jiang, Chen, & Dong, 2008) and children's ability to understand the form and function of print for initial reading predicts their later literacy achievement (e.g., Badian, 2001, Pullen and Justice, 2003). While there is extensive research using different methodologies on the visual specialization for print in alphabetic language readers (e.g., using ERP, MEG and fMRI) (e.g., Brem et al., 2009, Maurer et al., 2010, Maurer,Brem et al., 2005, Maurer et al., 2006, Tarkiainen et al., 1999), little is known about how Chinese-English bilingual child learners, who learn L1 Chinese and L2 English in parallel, process both L1 and L2 print. Such L1-L2 differences include not only proficiency but also linguistic properties. Thus, Chinese-English language learners provide a unique opportunity to investigate this question because Chinese differs strongly from alphabetic languages in visual and orthographic features (Tan et al., 2001). From a theoretical point of view, such research could help further highlight the nature of visual specialization for print that supports rapid word recognition. Thus, the main purpose of the present study was to characterize the visual specialization in response to Chinese and English words in Chinese-speaking children who learn Chinese and English in parallel, although at a lower proficiency level in English.

Neuropsychological research in skilled adult readers reveals that the process of visual specialization for print occurs very early, typically within 200 ms after onset of a stimulus. The neural sources involved in this process are considered to be located in the occipitotemporal brain regions often showing a left hemisphere dominance (e.g., Brem et al., 2009, Maurer et al., 2010, Maurer et al., 2006, Xue et al., 2008), corresponding to the visual word-form system (VWFS) (e.g., Brem et al., 2006), which is consistently implicated in the perceptual expertise for visual word recognition (for a review see Schlaggar & McCandliss, 2007) in neuroimaging studies. In studies using event-related EEG and MEG methodologies, the visual specialization for print is believed to be indexed by an early ERP component: N1 or N170 (e.g., Brem et al., 2006, Brem et al., 2009, Maurer et al., 2010, Maurer et al., 2005, Maurer,Brem et al., 2005, Maurer et al., 2006, Tarkiainen et al., 1999, Xue et al., 2008). The visual N1 is a negative evoked potential peaking between 150 and 200 ms over posterior brain regions (for a review see Maurer & McCandliss, 2008). The N1 is elicited by most visual stimuli, but has been shown to be increased for certain stimulus types such as faces and familiar objects compared to the control stimuli (e.g., Gauthier et al., 2003, Tanaka and Curran, 2001). The N1 elicited by linguistic stimuli such as letter strings differs from the one elicited by other types of visual stimuli (e.g., faces). For example, the N1 is typically more left-lateralized for visual word processing in readers of alphabetic languages (Brem et al., 2006, Maurer et al., 2005); in contrast, the N1 elicited by faces is observed to be more right-lateralized or bilateral-located (Tanaka & Curran, 2001).

In addition, converging evidence demonstrates that a larger N1 is commonly elicited by certain types of visual stimuli in participants who have perceptual expertise with those stimuli over visually matched stimuli (Gauthier et al., 2003, Tanaka and Curran, 2001). In other words, a larger N1 appears to be elicited when participants respond to those visual stimuli with which they are familiar. For example, participants who specialize in the recognition of birds or cars show an increased N1 when categorizing objects in their domain of expertise relative to the N1 elicited when they categorize objects beyond their domain of expertise (Tanaka & Curran, 2001). Nevertheless, the visual N1 is known to reflect a rapid visual categorization process and recognition of a familiar object (e.g., Bentin and Allison, 1996, Brem et al., 2009, Maurer et al., 2006, Rossion et al., 2000; Schendan, Ganis, & Kutas, 1998; Vogel and Luck, 2000, Xue et al., 2008).

A number of studies in native-speaking skilled adult readers of alphabetic languages have revealed that the N1 is typically larger for word and/or word-like stimuli (e.g., words, consonant strings, pseudowords) in comparison to visually matched stimuli (e.g., symbols, false font characters or pictures), and the scalp topography of the N1 specialization for print is particularly left hemisphere dominated, but the degree of the left-lateralization varies across studies (e.g., Bentin et al., 1999, Brem et al., 2005, Praverbio et al., 2004). For example, Bentin et al. (1999) reported a larger left occipitotemporal N1 for words, pseudowords, and nonwords compared to symbols and forms in a size judgment task in French native adult readers. Maurer et al. (2005) also observed that English native adult readers showed a larger N1 in response to words than visually matched symbol strings over the inferior occipitotemporal area in a one-back repetition detection task in which participants were required to indicate whenever a stimulus was presented continuously twice in a row. It was, thus, suggested that the N1 specialization for print appears to be automatic, indexing participants’ perceptual expertise for visual word form or print (for a review see Maurer & McCandliss, 2008).

Developmental ERP studies in child readers of alphabetic scripts have demonstrated that N1 print tuning emerges soon after reading training has begun in first graders showing a bilateral distribution (Eberhard-Moscicka et al., 2015, Zhao et al., 2014). N1 print tuning becomes more left-lateralized with further reading training (Eberhard-Moscicka et al., 2015; Maurer et al., 2006), and was found to be left-lateralized in children after grade 2 (Maurer et al., 2011; Eberhard-Moscicka et al., 2015). In some studies, however, N1 print tuning was rather bilateral also in older children (Araujo et al., 2012).

In a longitudinal project in German-speaking children, Maurer and colleagues found that the entire group of preschool children, who had no reading experience, did not yet show robust N1 specialization, but those children with high letter knowledge showed an atypically right-lateralized form of N1 specialization, even if they could not read (Maurer et al., 2005). The notion of a right-lateralized precursor state of print tuning during reading acquisition is supported by a study showing that the size of right hemisphere print tuning in preschool predicts later reading achievement in school (Brem et al., 2013). Moreover, initial familiarization effects in the N1 also showed a right-lateralized topography in adults who learned an artificial orthography (Maurer et al., 2010), suggesting that print tuning not only is reflected by the size, but also by a change from initial right-lateralization to a more bilateral or left-lateralized topography.

Right-lateralized effects of visual expertise for print have not only been associated with an early stage of learning to read, but also with logographic writing systems that might encourage more holistic processing (Wang, Kuo, & Cheng, 2011). In contrast to alphabetic writing systems, where there are coherent links between phonemes (spoken units) and graphemes (units of print), there are no regular or quasi-regular grapheme-phoneme conversions in Chinese, although most of the characters have a phonetic radical that provides a clue about the pronunciation of characters (et al., Tan et al., 2001). In addition, the visual information contained within a Chinese word is much more complex than in an English word in terms of visual-spatial structure (e.g., Hoosian, 1991). The letters comprising an English word are linearly constructed. In contrast, characters, the basic writing units in Chinese, are made of strokes or compact stroke patterns, displayed within a square box. There are great variations of shapes and forms of strokes and stroke patterns. Previous studies have suggested that brain responses underlying the reading of Chinese may differ from those underlying the reading of English words (e.g., Chee et al., 1999, Tan et al., 2001). Particularly, an increased involvement of right occipitotemporal regions in orthographic processing is suggested by meta-analyses of fMRI studies for Chinese compared to alphabetic languages (Bolger et al., 2005; Tan, Laird, Li, & Fox, 2005). In contrast to these fMRI results, however, the results regarding right-lateralization of the N1 response to Chinese characters have been rather inconsistent, with some studies reporting a bilateral or right-lateralized topography (Kim et al., 2004, Wang et al., 2011, Zhang et al., 2011) and some studies reporting a left-lateralized topography (Zhao Li, Lin, Cao, He, & Weng, 2012).

The N1 specialization for print is not only observed in children of alphabetic languages; children with logographic Chinese as a native language also show the N1 specialization for Chinese characters. For example, Cao et al. (2011) investigated the developmental trajectory of the N1 specialization for Chinese characters in Chinese 7-, 9- and 11-year-old children as well as college students when responding to Chinese characters and line drawings. The authors found that there was a larger left-lateralized N1 in response to Chinese characters as compared to visually matched line drawings in all four age groups, and the N1 elicited by Chinese characters was also larger for the 7-year-old group than the 11-year-old group and college students in comparison to line drawings. The similar pattern of N1 specialization for print across writing systems may indicate that a similar neural process underlies visual specialization for print (for a review see Maurer, Zevin, & McCandliss, 2008).

Although there have been many studies on the neural specialization for print in both child and adult native readers in alphabetic languages, it remains an open question as to what is the neural basis is for print in Hong Kong Chinese-speaking children, who are learning to read Chinese and English in parallel. Understanding the nature of the neural specialization for both Chinese and English print in Chinese-English language child learners might be of both theoretical and practical importance. Theoretically, studying visual specialization for print in Chinese-English speaking children provides evidence about the process of how one brain system is able to cope with contrastive writing systems. Specifically, it will address the question of whether the visual specialization for print in Chinese-English speaking children is similar for both Chinese and English print. To address this question we assessed event-related potentials (ERPs) at the scalp, allowing us to examine the neural activity for both L1 and L2 print, during an implicit reading task in Chinese-English speaking second graders. Practically, the findings on a neural basis for print specialization in bilinguals might be informative in understanding the causes of the comorbidity of reading difficulties in L1 Chinese and L2 English among Hong Kong Chinese-English speaking readers, possibly helping to establish a theoretical basis for building remediation strategies (Tong, Tong, & McBride-Chang, 2013).

Extending previous studies to Chinese-English language child learners, the present study examined the N1 specialization for print in Chinese-speaking second graders who learn English as a second language. In this study, we were interested in whether Chinese-second graders would show N1 specialization for both Chinese and English print and whether the less regular grapheme-phoneme mapping and the complex visual-spatial structure of Chinese script might lead to different brain activations when processing Chinese print in comparison to reading English print in terms of temporal and topographic distribution. We adopted the task of one-back repetition detection, which has been successfully used in prior studies on N1 specialization for print (e.g., Brem et al., 2009, Maurer et al., 2005, Maurer et al., 2006, Maurer et al., 2010), to examine the visual specialization for print in Chinese-speaking second graders who had started learning English as early as 3 years old. The one-back detection task is an implicit reading task in which children were required not to directly read the words, and they were asked to respond whenever a stimulus was presented continuously twice in a row. Four types of stimuli were involved in the present study consisting of Chinese two-character words, Korean two-character Hangul, and three letter English words, as well as false font characters. The visual specialization for Chinese print was defined by contrasting neural responses to Chinese words versus unfamiliar Korean Hangul, and the visual specialization for English was defined by contrasting neural responses to English words versus false font characters in the present study. We hypothesized that Chinese-English language learners would show a larger N1 in response to their familiar scripts, i.e., Chinese and English words, in comparison with unfamiliar controlled stimuli, i.e., Korean and false font characters.

We further expected that lateralization differences between visual specialization for Chinese and English would be informative about the underlying processes involved in early orthographic processing: If the properties of the writing systems (logographic vs. alphabetic) played an important role, then we would expect a more right-lateralized visual specialization for Chinese compared to English. However, if reading skill were particularly important, then we would expect a more right-lateralized visual specialization for English (L2) compared to Chinese (L1).

Although primarily focusing on the N1 component of the ERP, we also analyzed the preceding P1 component, to characterize starting points of visual specialization for print in these children. The visual P1 is a component peaking in adults at around 100 ms after stimuli onset with localization at posterior areas. This component is found to be sensitive to physical characteristics of visual stimuli. The P1 reflects a low-level perceptual analysis (e.g., Brem et al., 2006; Dien, 2009; Hauk & Pulvermüller, 2004). In print specialization research, the P1 is larger for symbol strings than for words with laterality in the occipital areas, and the amplitude and latency of the P1 decreases with age (e.g., Brem et al., 2006). These effects may reflect an early onset of specialized orthographic processing before such processing is maximal in the N1 time range (Maurer et al., 2005). It might be of interest to examine whether this early effect would be elicited in Chinese children in processing Chinese and English words. Additionally, we measured children's reading abilities and vocabulary knowledge in both Chinese and English in order to examine whether children in our study had mastery of word recognition in both Chinese and English. We also tested their nonverbal reasoning skills to confirm that all participants had normal ability in nonverbal reasoning.