- Email: [email protected]
Familiarity with visual forms contributes to a left-lateralized and increased N170 response for Chinese characters
Neuropsychologia 134 (2019) 107194
Contents lists available at ScienceDirect
Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia
Familiarity with visual forms contributes to a left-lateralized and increased N170 response for Chinese characters
T
Licheng Xuea,b,c, Urs Maurerd,e, Xuchu Wenga,b,c,f, Jing Zhaob,c,∗ a
Center for Cognition and Brain Disorder, Hangzhou Normal University, China Institute of Psychological Sciences, Hangzhou Normal University, China c Zhejiang Key Laboratory for Research in Assessment of Cognitive Impairments, China d Department of Psychology, Chinese University of Hong Kong, China e Brain and Mind Institute, Chinese University of Hong Kong, China f Institute for Brain Research and Rehabilitation, South China Normal University, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: N170 Chinese character Visual expertise Visual form familiarity
While skilled readers produce an increased and left-lateralized event-related-potential (ERP) component, known as N170, for strings of letters compared to strings of less familiar units, it remains unclear whether perceptual familiarity plays an important role in driving the increased and left-lateralized N170 for print. The present study addressed this issue by examining N170 responses for regular Chinese characters and cursive Chinese characters which are visually less familiar regarding their form, yet with phonological and semantic properties. Stroke combinations, which are with unfamiliar visual form and without phonological or semantic properties, were used as low-level control stimuli. Twenty college students (22.6 ± 1.2 years) were examined. A content-irrelevant color matching task was used to control for differences of attention load across familiar and unfamiliar stimuli. A left-lateralized N170 was evoked only by regular characters, but not by cursive characters or stroke combinations. Moreover, cursive characters, which are principally readable but visually unfamiliar, produced a lower N170 than regular characters, and no N170 difference was found compared with stroke combinations. These results suggest that visual form familiarity serves as an important driver for the increased and left-lateralized N170 response.
1. Introduction Through extensive experience with learning to read, humans acquire the ability to quickly recognize visual words, and show specialized responses in their brain (Schlaggar and McCandliss, 2007). In skilled readers, visual words evoke an increased and left-lateralized event-related potential (ERP) component peaking within 200 ms after stimulus onset, usually termed N170 or N1 (Bentin et al., 1999; Maurer et al., 2006). Specifically, the N170 amplitude is larger and left-lateralized for familiar orthographic stimuli such as real words and pseudowords compared to less familiar stimuli such as consonant strings and symbol strings in alphabetic scripts (Brem et al., 2006; Maurer et al., 2005; Maurer et al., 2006,; Araújo et al., 2015; Fraga González et al., 2014; Fraga González et al., 2016, Varga et al., 2018). Similarly, real and pseudo Chinese characters produce an increased and left-lateralized N170 response relative to less familiar false characters and stroke combinations (Lin et al., 2011; Yang et al., 2017; Zhao et al., 2012; Cao et al., 2011). However, it remains unclear whether perceptual ∗
familiarity plays an important role in driving the increased and leftlateralized N170 for print. A straightforward approach to address this issue is to examine N170 responses for stimuli with less familiar visual forms while having phonological and semantic properties. Chinese cursive characters accord with these properties. In the modern Chinese writing system, cursive characters are one font of printed Chinese although less used than regular characters (Zhou, 2001). Cursive characters are generally employed for daily use in some situations (e.g., billboards and trademarks) as well as in calligraphic art (Wang, 1958), and thus skilled Chinese readers can recognize common cursive characters. In the framework of visual word reading, visual input is first transformed into abstract visual word forms (invariant under changes in location, size, case and font) in the visual word form system, and then activates semantic and phonological processing (at this stage, surface visual features would no longer play a role), although top-down modulation may play a role in early visual processing (Cohen and Dehaene, 2004; see Dehaene et al., 2005 for a review; Warrington and
Corresponding author. 2318 Yuhang Tang Rd., Yuhang District, Hangzhou, 311121, China. E-mail address: [email protected] (J. Zhao).
https://doi.org/10.1016/j.neuropsychologia.2019.107194 Received 22 March 2019; Received in revised form 5 August 2019; Accepted 12 September 2019 Available online 19 September 2019 0028-3932/ © 2019 Published by Elsevier Ltd.
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
frequency of occurrence of 1170 per million (the modern Chinese frequency dictionary, 1985). Cursive characters were the same characters typed in cursive font. They were comparable with regular characters in phonological and semantic properties. Stroke combinations were constructed by recombining the strokes of real characters. They were matched with regular characters in basic visual stimulus properties (i.e., stroke and visual complexity), but the internal structure was disrupted, and they had no pronunciation or meaning. For each type, 36 stimuli were created. To further confirm the visual form familiarity of cursive characters and regular characters, we asked another twenty participants to rate the visual form familiarity of the two types of stimuli in a seven-point scale (1 = never to 7 = frequent). The score of the visual form familiarity judgment was analyzed in a paired-sample t-test with Stimulus Type as a within-subject factor. Results showed that the rating score of cursive characters was significantly lower than that of regular characters (3.59 vs. 6.54, t(19) = 11.34, p < 0.001).
Table 1 Description and examples of different types of stimuli. Type of stimuli
Example
Familiarity
Pronunciation
Meaning
Regular character
high
high
high
Cursive character
low
high
high
Stroke combination
0
0
0
Shallice, 1980). Thus, in this framework, cursive characters, once transformed into abstract visual forms, are essentially the same as regular characters in terms of phonological and semantic properties (Marshall and Newcombe, 1973; Price and Devlin, 2003). In the present study, we aimed to examine whether perceptual familiarity plays an important role in driving the increased and left-lateralized N170 for print by comparing cursive characters (e.g., ) and regular characters (e.g., ) (see Table 1). In addition, stroke combinations (e.g., ) were formed by randomly recombining the strokes of the regular characters as a low-level visual control. Stroke combinations were matched with regular characters in terms of the low-level visual features, while the combined pattern of strokes was entirely disrupted, and had no phonological or semantic properties. To control for differences of attention load across familiar and unfamiliar stimuli, we used a content-irrelevant color matching task with a short presentation time (300 ms), in which participants were asked to make judgments about the color of the stimuli only and to ignore the other properties (see Zhao et al., 2012 for more discussion). Similar to results in previous studies (Cao et al., 2011; Lin et al., 2011; Zhao et al., 2012), we expected regular characters would evoke an increased and left-lateralized N170 response, while stroke combinations would evoke a lower and less left-lateralized N170 response relative to regular characters. Importantly, due to reduced visual form familiarity, we expected a reduced and less left-lateralized N170 response for cursive characters if perceptual familiarity plays an important role to drive this N170 effect. Alternatively, we expected a similar N170 response for the two types of characters that have the same linguistic properties regardless of fonts (regular vs. cursive) if the increased and left-lateralized N170 effect is not modulated by perceptual familiarity.
2.3. Procedure Each stimulus was presented in one of the three colors green, red or yellow on a gray background in the center of the screen extending 1.32° in visual angle. Each trial was constructed as follows: a mean interval of 1600 ms (1450 ms–1750 ms, with ± 75 ms random jitter) and fixed stimulus duration of 300 ms was used. Participants were instructed to press a button, whenever a color of a stimulus occurred consecutively two times in a row (i.e., the one-back color repetition detection) (as shown in Fig. 1b). All stimuli were presented four times as non-targets, and a subset was additionally presented as targets (17% of all stimuli). The three conditions were presented randomly within a block. The order of the left and right hand to respond was counterbalanced between participants. After the ERP experiment, participants were instructed to read aloud all the regular and cursive Chinese characters presented in the ERP experiment. Each character was presented for 300 ms, and participants pronounced the characters as accurately and fast as possible. Both accuracy and reaction time (the time interval between item onset and the pronunciation onset) were recorded. To evaluate the difficultly of visual form processing of the three stimulus types, we asked additional 27 participants to participate in a visual form matching task. The stimuli were the same as those used in the ERP experiment, but were presented in three pairs (i.e., regular character-regular character, cursive character-cursive character, stroke combination-stroke combination). Participants were asked to judge whether the two stimuli are same. The trial started with a fixation cross presented at the center of the screen for 800 ms. The stimulus pair was displayed for 300 ms. All stimuli were presented in black against a gray background, and subtended 1.3° of visual angle at a viewing distance of 100 cm. Horizontally, the two characters were 3.6° of visual angle apart. The test was composed of 180 trials (90 ‘same’ trials and 90 ‘different’ trials). For each stimulus type, there were 60 trials with 30 ‘same’ trials and 30 ‘different’ trials. The order of trials was randomized and the keys corresponding to same/different responses were counterbalanced between participants.
2. Methods The research protocol reported here was approved by a local institutional review board and written informed consent was obtained from the participants in the present study. 2.1. Participants There were twenty-two participants in the experiment (11 of them were male). One of the participants did not finish the experiment, and the hit rates of one participant in the one-back color repetition detection task was only at random level. Therefore, two participants (2 males) were excluded from the analyses. Mean age of the remaining participants was 22.6 years ( ± 1.2 years). Participants were all undergraduate students at the university. They were all right-handed according to the Edinburgh handedness inventory (Oldfield, 1971), with normal or corrected-to-normal vision and all reported themselves to be in good health.
2.4. Electrophysiological recording and analysis Electroencephalogram (EEG) signal was recorded from 64 Ag/AgCl electrodes secured in an elastic cap (NeuroScan Inc., EI Paso, TX, USA). And horizontal electrooculogram (EOG) was recorded with additional electrodes on the left and right external cantus, while the vertical EOG was recorded with two electrodes placed above and below the left eye. An electrode located between Cz and CPz served as a reference, but an average reference transformation was used for the data analysis. An electrode between Fpz and Fz on the cap served as ground. All channels were amplified with a band pass from AC 0.1 to 100 HZ and the sampling rate was 1000 Hz. Electrode impedances were kept below 5 kΩ.
2.2. Material Three types of stimuli were used: regular characters, cursive characters, and stroke combinations. Examples were illustrated in Fig. 1a. All regular characters had a single-structure, composed of only one logographeme and five strokes on average (range: four-seven) with a 2
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 1. a. Examples of stimuli: regular character (RC), cursive character (CC), and stroke combination (SC). b. Sketch map of the one-back color repetition detection task. The arrow implied the direction of presentation.
at each time point in three pairwise comparisons (regular character vs. cursive character, regular character vs. stroke combination, cursive character vs. stroke combination). To correct for multiple comparisons, RAGU has an in-built procedure (Global Duration Test), which uses a duration threshold that is detected using a randomization procedure. In other words, differences between two types of stimuli are detected that last longer than would be expected by chance (Koenig et al., 2011).
The EEG data was preprocessed using EEGLAB software (Delorme and Makeig, 2004). First, the data was digitally filtered by band-pass (0.5–30HZ). The Artifact Subspace Reconstruction (ASR) method was applied to remove artifacts and bad channels, which were reconstructed from neighboring channels (Mullen et al., 2013). Secondly, the data was referenced to a common average reference. Finally, EOG artifacts were corrected using ICA and SASICA tools (Chaumon et al., 2015), in which the ICA components which correlated with vEOG and hEOG channels were removed. The EEG was then segmented into epochs of 700 ms including a 100 ms pre-stimulus baseline, used for baseline correction. Epochs with extreme amplitude exceeding threshold ( ± 80 μV) within the interested time window were rejected. Trials of non-targets were averaged separately for the three conditions. To minimize selection bias of electrodes mainly based on visual inspection and allow for comprehensive multichannel ERP analyses (Maurer et al., 2005; 2006; Wang and Maurer, 2017), we identified EEG segments for further analyses by the maxima of global field power (GFP) curve after averaging all grand means across all stimulus types and subjects (Lehmann and Skrandies, 1980). Similar to the procedure used in previous studies (Bentin et al., 1999; Lin et al., 2011; Wong et al., 2005; Zhao et al., 2012), five pairs of channels in occipito-temporal areas were selected according to the topographic maxima in the negative field over both hemispheres, including PO7/8, P7/8, PO5/6, PO3/4, and CB1/2. Based on the GFP curve, the first ERP segment was from 57 ms to 105 ms (bilateral troughs of the first wave), and was identified as the P100 component based on the bifocal posterior positivity. The following segment (106198 ms) was rather long, with a sharp initial peak and a subsequent decrease, which was identified as the N170 component. And the following segment (199-274 ms for maxima) was identified as the P2 component. The peak value of amplitude on these channels for these three components was obtained by the peak detection, respectively. And after that, the statistics was based on the peak amplitudes. Specifically, the mean of the P100, N170 or P2 peak amplitude of these channels on the left hemisphere (i.e., PO7/P7/PO5/PO3/CB1) and the right hemisphere (i.e., PO8/P8/PO6/PO4/CB2) was computed, respectively. The latency of these components in one pair of channels (PO7/PO8, the topographic maxima in the negative/positive field on the occipito-temporal area over both hemispheres) was detected. Finally, peak amplitudes and latencies of P100, N170 or P2 were analyzed in Analyses of Variance (ANOVA, procedure GLM) with two within-subject factors Stimulus Type (regular characters, cursive characters, stroke combinations) and Hemisphere (left, right), respectively. And further, to examine whether the peak amplitudes at the occipito-temporal electrodes reflected the effect across ERP maps, the data was analyzed using topographic analyses of variance (TANOVA) implemented in the Randomization Graphical User interface (RAGU) software (Koenig et al., 2011). Specifically, the TANOVA was computed
3. Results 3.1. Behavioral results 3.1.1. Color matching Hit rates and reaction times were analyzed for target trials of the three stimulus types. Means and standard deviations are illustrated in Table 2. Due to a file error, the behavioral data of four participants were lost. Since the participants were asked to respond as accurately as possible, their overall hit rates were high for all stimulus types. Results in one-way ANOVA with Stimulus Type (regular character, cursive character, stroke combination) as a within subject factor revealed a significant main effect, F (2, 30) = 14.62, p < 0.001, η2G = 0.49. Results in the post-hoc test showed that the hit rate for color repetition detection in the stroke combination condition was lower than that in the regular character condition (p = 0.01) and cursive character condition (p < 0.001), while no difference was found between regular and cursive characters (p = 0.29). For reaction times, the one-way ANOVA did not show a significant main effect of Stimulus Type, F (2, 30) = 0.96, p = 0.39, η2G = 0.06. 3.1.2. Chinese character naming After the EEG experiment, participants were measured in a Chinese character naming test, which consisted of the regular and cursive characters presented in the EEG experiment. Participants could name regular and cursive characters at the same accuracy level (t18 < 1, p > 0.05), and the accuracies for both types of stimuli were close to the ceiling level in naming regular characters (0.98) and cursive characters (0.97). Importantly, the cursive characters that were not correctly named were not consistent across participants. This together with very high accuracy indicates that all cursive characters were clear to each participant. For reaction time, results in a paired-sample t-test showed Table 2 Hit rate and reaction time [Mean (SE)].
Hit rate (%) Reaction time (ms)
3
Regular
Cursive
Stroke combination
95.57 (1.60) 495.30 (32.03)
96.87 (1.80) 493.04 (32.04)
92.45 (1.70) 502.82 (31.25)
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
the left hemisphere, the amplitude of N170 evoked by regular characters was significantly larger than those evoked by cursive characters (p = 0.002), and stroke combinations (p = 0.001), respectively. No N170 difference was found between cursive characters and stroke combinations (p = 0.17). However, in the right hemisphere, there was no significant difference between any two stimulus categories (regular characters vs. cursive characters, p = 0.82; regular characters vs. stroke combinations, p = 0.17). Similarly, no significant difference was found between cursive characters and stroke combinations (p = 0.46).
Table 3 Accuracy and reaction time [Mean (SE)].
Accuracy (%) Reaction time (ms)
Regular
Cursive
Stroke combination
94.94 (0.70) 707.39 (31.64)
90.93 (0.70) 741.51 (38.48)
89.01 (1.20) 750.58 (28.95)
that participants took longer time to name cursive characters (731 ± 98 ms) than regular characters (618 ± 57 ms), t18 = 6.70, p < 0.001.
3.2.3. N170 peak latency Table 4 illustrates means and standard errors of N170 peak latency for the three types of stimuli. Results in the two-way ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed significant main effects of Stimulus Type, F (2, 38) = 9.41, p < 0.001, η2G = 0.33, and Hemisphere F (1, 19) = 4.64, p = 0.044, η2G = 0.20. The interaction of Stimulus Type by Hemisphere was not significant F (2, 2 38) = 2.0, p = 0.16, ηG = 0.09. Results in the post-hoc test showed that N170 latency for regular characters was shorter than that for cursive characters (p = 0.003), and marginally shorter than that for stroke combinations (p = 0.05), while no difference was found between cursive characters and stroke combinations (p = 0.45).
3.1.3. Visual form matching Means and standard deviations of accuracy and reaction time are illustrated in Table 3. Data was analyzed in a one-way ANOVA with Stimulus Type as a within-subject factor. For the accuracy, results showed the main effect of Stimulus Type was significant, F(2, 52) = 14.53, p < 0.001. Results in the post-hoc test (with Bonferroni adjustment) showed that the accuracy of regular character was higher than that of cursive character (p < 0.001) and stroke combination (p < 0.001), while no difference was found between cursive character and stroke combination (p = 0.46). For the reaction time, results showed the main effect of Stimulus Type was significant, F(2, 52) = 8.75, p = 0.003. Results in the post-hoc test (with Bonferroni adjustment) showed that the reaction time of regular character was shorter than that of cursive character (p = 0.021) and stroke combination (p < 0.001), while no difference was found between cursive character and stroke combination (p = 0.99). These results suggest that the difficulty of visual form processing was regular character < cursive character = stroke combination.
3.2.4. P2 peak amplitude Results in the two-way ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed a significant main effect of Stimulus Type, F (2, 38) = 7.90, p < 0.01, η2G = 0.29. However, the main effect of Hemisphere (F (1, 19) = 0.34, p = 0.57, η2G = 0.02) and the interaction of Stimulus Type and Hemisphere (F (2, 38) = 0.31, p = 0.74, η2G = 0.02) were not significant. Post hoc tests within Stimulus Type showed that the P2 amplitude evoked by regular characters was significantly larger than that evoked by cursive characters (p = 0.001), while marginally larger than that evoked by stroke combinations (p = 0.07). No difference in P2 amplitude was found between cursive characters and stroke combinations (p = 0.90).
3.2. ERP results Fig. 2a shows grand average waveforms for three stimulus categories at the occipito-temporal electrodes on the left and right hemispheres. A positive deflection peaking within 100 ms, and a negative deflection peaking within 200 ms were clearly visible for each stimulus type over bilateral occipito-temporal sites.
3.2.5. P2 peak latency Table 5 illustrates means and standard error of P2 peak latency for the three types of stimuli. Results in the two-way ANOVA showed a significant main effect of Stimulus Type, F (2, 38) = 14.38, p < 0.001, η2G = 0.43. The main effect of Hemisphere, F (1, 19) = 3.12, p = 0.10, η2G = 0.14 and the interaction of Stimulus Type by Hemisphere, F (2, 2 38) = 2.40, p = 0.11, ηG = 0.11 were not significant. Results in the posthoc test of the Stimulus Type showed that the P2 latency for regular characters was shorter than that for cursive characters (p < 0.001), and stroke combinations (p = 0.02), while no significant difference was found between cursive characters and stroke combinations (p = 0.10).
3.2.1. P100 peak amplitude Results in the ANOVA showed neither the main effect of the Stimulus Type, F (2, 38) = 1.36, p = 0.27, η2G = 0.07, nor the main effect of Hemisphere, F (1, 19) = 3.21, p = 0.09, η2G = 0.14 was significant. The interaction of Stimulus Type and Hemisphere was not significant, F (2, 2 38) = 0.61, p = 0.54, ηG = 0.03. These results showed no significant differences in low-level visual features across the three types of stimuli. 3.2.2. N170 peak amplitude Fig. 2b shows the average of N170 peak amplitudes for three types of stimuli at the occipito-temporal electrodes on the left and right hemisphere. ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed a significant main effect of the Stimulus Type, F (2, 38) = 9.87, p < 0.01, η2G = 0.34, while the main effect of Hemisphere was not significant, F (1, 19) = 2.99, p = 0.10, η2G = 0.14. Importantly, the interaction of Stimulus Type and Hemisphere was significant, F (2, 38) = 5.54, p = 0.01, η2G = 0.23. For the lateralization, we observed a left-lateralization of N170 was specific to regular characters (i.e., stimuli with high familiarity with visual forms). Specifically, results in the analysis of N170 amplitude between the left and right hemispheres showed only a significant leftlateralization of N170 response for regular characters (p = 0.04), while neither cursive character (p = 0.14) nor stroke combinations (p = 0.22) evoked such a left-lateralized N170 response. The following results reflect subsequent analyses for differences between stimulus categories separately for each hemisphere. Regarding
3.2.6. Topographic map TANOVA results To investigate whether the difference showed at the occipito-temporal electrodes at N170 peak latency existed across the ERP maps, the TANOVAs were used for comparisons between stimulus categories. Similar to the N170 analysis, the trial length of data sample was 800 ms including 100 ms baseline. To examine our hypothesis, we focused on the first 400 ms after stimulus onset. For regular-cursive comparison, as shown in Fig. 3b and c, the time windows that survived the Global Duration Test are shown in green, and the topographies in the difference t-maps are separately illustrated for the time periods corresponding to different components (N1/N170, P2). We divided the segment into two time-window as counterparts of N170 and P2 components, based on the increased p-value that coincides with a change in the topography of the difference wave. Similar to the N170 and P2 time ranges (see Fig. 3a), results in TANOVA statistics showed significant difference between regular and cursive characters 4
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 2. a. Grand average waveforms with 95% confidence intervals (error bands) at occipito-temporal channels on both the left and the right hemispheres b. Box plots of average of N170 peak amplitude evoked by different stimulus categories on left and right hemispheres. Lines in the boxes denote means, **p < 0.01, *p < 0.05. Table 4 N170 peak latency [Mean (SE)].
Left hemisphere Right hemisphere
Table 5 P2 peak latency [Mean (SE)].
Regular
Cursive
Stroke combination
154.30 (3.41) 146.20 (4.14)
157.78 (3.67) 153.28 (4.60)
156.55 (3.60) 152.23 (4.25)
Left hemisphere Right hemisphere
5
Regular
Cursive
Stroke combination
232.20 (4.36) 234.73 (4.88)
239.75 (4.86) 250.54 (4.50)
238.60 (4.92) 244.56 (4.86)
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 3. a. N170 waveforms averaged with 95% confidence intervals (error bands) across five occipito-temporal channels on both the left and the right hemispheres, respectively. b., d., f. TANOVA results of paired comparisons of three stimulus categories (800 ms, including 100 ms baseline, RC-regular characters, CC-cursive characters, SC-stroke combinations). Green periods survived Global Duration Test (p < 0.05). c., e., g. t-maps during the time window for N170 and P2 components, respectively. ∗∗∗means p-value of TANOVA < 0.001.
semantic properties. With the goal of identifying the role of visual form familiarity and controlling for differences of attention load across familiar and unfamiliar stimuli, we used a content-irrelevant color matching task with a short duration of stimulus presentation (300 ms). In various commonly-used explicit language tasks with long stimulus durations, participants are often faster in response to target letters embedded in native than non-native words and more accurate at detecting repetitions of orthographic stimuli (such as words or letter strings) than at detecting repetitions of less familiar symbol strings (Maurer et al., 2005; Maurer et al., 2008). Therefore, the increased N170 in response for orthographic stimuli reported in these studies may be partly explained by attentional bias to familiar stimulus categories (see Zhao et al., 2012 for more discussion). In contrast, in the current study, we did not find any behavioral response difference between regular and cursive characters. Hence, the task we used here effectively controlled for possible confounding caused by different attentional loads between familiar and unfamiliar stimuli. An important finding of the present study is that the left-lateralized and increased N170 was sensitive to familiarity of visual form. We found that only regular characters evoked a significantly left-lateralized
between 112 ms and 157 ms for N170 and 157 to 252 ms for P2, respectively (Fig. 3b). The t-maps of the difference between these two stimulus categories are used for illustration (Fig. 3c). In addition, a similar pattern was found in the regular character-stroke combination comparison, 108-166 ms for N170 and 116-243 ms for P2 (t-maps seen in Fig. 3d and e). By contrast, as shown in Fig. 3f, there was no significant difference in the cursive character vs. stroke combination comparison in the similar time range as the regular character vs. stroke combination comparison (108-166 ms for N170 and 166-243 ms for P2, see Fig. 3f and g). 4. Discussion Although there is a consensus that the increased left-lateralized N170 response for print is depending on extensive experience with this visual type, it remains unclear whether perceptual familiarity plays an important role in driving increased and left-lateralized N170 for print. We addressed this issue by comparing N170 responses for three types of visual categories (regular Chinese character, cursive Chinese character, and stroke combination) with different degrees of visual form familiarity, yet only regular and cursive characters have phonological and 6
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
González et al., 2016), while in the present study we used an implicit task that does not require specific phonological and semantic processing. Results in some of previous studies suggest that N170 sensitivity to words is modulated by task demands (Strijkers et al., 2015; Wang and Maurer, 2017). It is possible that the task-demand may contribute to the effect of visual form familiarity on the N170 responses. With respect to script types, in addition to Chinese, the increased and left-lateralized N170 for print has been found in scripts with different alphabetic scripts (e.g., Dutch in Fraga González et al., 2014; Fraga González et al., 2016; Portuguese in Araújo et al., 2015). Some of the results in previous studies suggest that reading Chinese may not automatically engage strong phonological processing as reading alphabetic words (Chen and Shu, 2001; Yeh and Li, 2002). Thus, script-type may also contribute to the effect of visual form familiarity on the N170 responses. Finally, we would like to acknowledge a few limitations of the present study and note some directions for future studies. First, the present study only examined Chinese characters processing on Chinese participants. Future research needs to examine whether this pattern of N170 lateralization can be extended to other writing systems. Second, the current findings are not sufficient to directly address whether visual form familiarity a lone can drive the increased and left-lateralized N170. In future studies one can ask participants to learn new symbols and manipulate exposure. One can also investigate the effect of Chinese Characters processing in alphabetic readers who are learning to read Chinese, and vice versa. In conclusion, using an implicit color matching task, we found an increased and left-lateralized N170 response for the regular characters compared to the cursive characters that were less familiar in terms of visual form but had phonological and semantic properties. We also examined the N170 evoked by the stroke combinations that were with less familiar visual form and had no pronunciation and meaning, and found no N170 difference between cursive characters and stroke combinations. These findings provide empirical evidence that the perceptual familiarity plays an important role to drive the increased and leftlateralized N170 response for print.
N170, while either cursive characters or stroke combinations evoked a bilateral N170 response. Moreover, detailed analysis of N170 amplitude at channels over the left posterior region showed that regular characters evoked stronger N170 amplitude relative to both cursive characters and stroke combinations. As a font that is relatively rarely used in the modern Chinese writing system, cursive characters are visually less familiar than regular characters in standard style of handwriting (see Zhou et al., 2001 for detailed discussion). This conclusion was confirmed by our results in visual form familiarity rating, which showed that cursive characters were visually less familiar than regular characters for skilled readers. In the naming task, all participants correctly named virtually all cursive characters (the accuracy was 0.97). These results together suggest that the cursive characters used in this study were readable, while visually unfamiliar. ERP results show that cursive characters produced weaker and less left-lateralized N170 than regular characters and a similar N170 response to stroke combinations that have no pronunciation and meaning. These results together demonstrated that perceptual familiarity plays an important role to drive the left-lateralized and increased N170. The remaining issue is whether the increased and left-lateralized N170 response could be related to the mapping of visual form of a character to its phonological/semantic properties. Compared to regular characters, it may be more difficult to decode the cursive characters (access to phonological and semantic information). This interpretation seems to be consistent with results of the name task, which showed that participants took longer time to name cursive characters than regular character. However, it is important to note that participants also took longer time to process cursive characters than regular characters in visual form matching. Therefore, it is also possible that the longer reaction times of naming cursive characters may be due to the difficulty of visual form processing. Moreover, this interpretation is not supported by the N170 results of the current study. If a more effortful access of phonological and semantic information was reflected in the N170 component, an increased or prolonged N170 would be expected (similar to increased activation for pseudowords compared to words or for low-frequency compared to high-frequency words in fMRI studies; e.g., Kronbichler et al., 2004). However, we did not find this N170 effect. Rather our ERP results suggest that the more difficult lexical access is reflected in the subsequent P2 component, with not only larger amplitude but also shorter latency for regular characters compared with cursive characters and stroke combinations. This interpretation is compatible with previous studies suggesting that the P2 is associated with later attentional processes and the mapping of visual form of a word to its phonological properties (Barnea and Breznitz, 1998; Bentin et al., 1999; Sereno et al., 1998; Zhang et al., 2009). There is an ongoing discussion about whether visual expertise effects in reading are driven by bottom-up processes or by interactions between automatic top-down predictions and bottom-up input (Dehaene and Cohen, 2011; Price and Devlin, 2011). While our current results are unlikely to reflect effortful access of phonological and semantic representation, these results are not sufficient to resolve the debate regarding the nature of the visual expertise effects. Therefore, it is highly possible that the visual form familiarity effects in the N170 reflect the sensitivity of feature detectors to bottom-up visual input (Dehaene and Cohen, 2011). However, there is another possibility that the visual familiarity effects in the N170 reflect prediction errors arising from matching automatic top-down predictions from higher-level visual or language regions with visual input (Price and Devlin, 2011). Whether the visual expertise effects in the N170 is driven by purely bottomup processing or by interactive processing, and whether potential predictions are arising from visual regions or language regions remains to be elucidated in future studies. It is worth noting that earlier studies have examined the left-lateralized and increased N170 for print in different tasks (e.g., lexical decision in Simon et al., 2007; naming in Chauncey et al., 2008; Wang and Maurer, 2017; repetition detection in Fraga González et al., 2014; Fraga
Funding This work was supported by grants from the National Natural Science Foundation of China (NSFC) (grant No.: 31771229), a grant from the Zhejiang Province Natural Science Foundation (grant No.: LY17C090008), and Project of the Ministry of Science and Technology of the People's Republic China (2016YFE0130400). Declaration of interest The authors have no competing interests. CRediT authorship contribution statement Licheng Xue: Data curation, Formal analysis, Methodology, Resources, Software, Visualization, Writing - original draft. Urs Maurer: Formal analysis, Methodology, Resources, Software, Writing review & editing. Xuchu Weng: Conceptualization, Funding acquisition, Methodology, Resources, Software, Writing - review & editing. Jing Zhao: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Project administration, Supervision, Validation, Writing - review & editing. References Araújo, S., Faísca, L., Bramão, I., Reis, A., Petersson, K.M., 2015. Lexical and sublexical orthographic processing: an ERP study with skilled and dyslexic adult readers. Brain Lang. 141, 16–27. https://doi.org/10.1016/j.bandl.2014.11.007. Barnea, A., Breznitz, Z., 1998. Phonological and orthographic processing of Hebrew words: electrophysiological aspects. J. Genet. Psychol. 159 (4), 492–504. https://doi. org/10.1080/00221329809596166.
7
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
https://doi.org/10.1016/j.neuroimage.2006.06.025. Maurer, U., Zevin, J.D., McCandliss, B.D., 2008. Left-lateralized N170 effects of visual expertise in reading: evidence from Japanese syllabic and logographic scripts. J. Cogn. Neurosci. 20 (10), 1878–1891. https://doi.org/10.1162/jocn.2008.20125. Modern Chinese Frequency Dictionary. Beijing Language Institute Press, Beijing (In Chinese). Mullen, T., Kothe, C., Chi, Y.M., Ojeda, A., Kerth, T., Makeig, S., et al., 2013. Real-time modeling and 3D visualization of source dynamics and connectivity using wearable EEG. In: Conference Proceedings : Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2013, pp. 2184–2187. https://doi.org/10.1109/ EMBC.2013.6609968. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9 (1), 97–113. https://doi.org/10.1016/0028-3932(71) 90067-4. Price, C.J., Devlin, J.T., 2003. The myth of the visual word form area. Neuroimage 19 (3), 473–481. https://doi.org/10.1016/S1053-8119(03)00084-3. Price, C.J., Devlin, J.T., 2011. The Interactive Account of ventral occipitotemporal contributions to reading. Trends Cogn. Sci. 15 (6), 246–253. https://doi.org/10.1016/j. tics.2011.04.001. Schlaggar, B.L., McCandliss, B.D., 2007. Development of neural systems for reading. Annu. Rev. Neurosci. 30 (1), 475–503. https://doi.org/10.1146/annurev.neuro.28. 061604.135645. Sereno, S.C., Rayner, K., Posner, M.I., 1998. Establishing a time-line of word recognition. Neuroreport 9 (10), 2195–2200. https://doi.org/10.1097/00001756-19980713000009. Simon, G., Petit, L., Bernard, C., Rebaï, M., 2007. N170 ERPs could represent a logographic processing strategy in visual word recognition. Behav. Brain Funct. 3, 1–11. https://doi.org/10.1186/1744-9081-3-21. Strijkers, K., Bertrand, D., Grainger, J., 2015. Seeing the same words differently: the time course of automaticity and top-down intention in reading. J. Cogn. Neurosci. 27 (8), 1542–1551. https://doi.org/10.1162/jocn_a_00797. Varga, V., Tóth, D., Csépe, V., 2018. Orthographic‐Phonological Mapping and the Emergence of Visual Expertise for Print: A Developmental Event‐Related Potential Study. Child Dev 1–13. https://doi.org/10.1111/cdev.13159. Wang, F.Y., 1958. Introduction to Chinese Cursive Script. Institute of Far Eastern Languages, Yale University Retrieved from. https://yalebooks.yale.edu/book/ 9780887100338/chinese-cursive-script. Wang, F., Maurer, U., 2017. Top-down modulation of early print-tuned neural activity in reading. Neuropsychologia 102, 29–38. https://doi.org/10.1016/j.neuropsychologia. 2017.05.028. Warrington, E.K., Shallice, T., 1980. Word-form dyslexia. Brain 103 (1), 99–112. https:// doi.org/10.1093/brain/103.1.99. Wong, A.C.N., Gauthier, I., Woroch, B., Debuse, C., Curran, T., 2005. An early electrophysiological response associated with expertise in letter perception. Cognit. Affect Behav. Neurosci. 5 (3), 306–318. https://doi.org/10.3758/CABN.5.3.306. Yang, H., Zhao, J., Gaspar, C.M., Chen, W., Tan, Y., Weng, X., 2017. Selectivity of N170 for visual words in the right hemisphere: evidence from single-trial analysis. Psychophysiology 54 (8), 1128–1137. https://doi.org/10.1111/psyp.12867. Yeh, S.-L., Li, J.-L., 2002. Role of structure and component in judgments of visual similarity of Chinese characters. J. Exp. Psychol. Hum. Percept. Perform. 28 (4), 933–947. Retrieved from. http://www.ncbi.nlm.nih.gov/pubmed/12190259. Zhang, Q., Zhang, J.X., Kong, L., 2009. An ERP study on the time course of phonological and semantic activation in Chinese word recognition. Int. J. Psychophysiol. 73 (3), 235–245. https://doi.org/10.1016/J.IJPSYCHO.2009.04.001. Zhao, J., Li, S., Lin, S.E., Cao, X.H., He, S., Weng, X.C., 2012. Selectivity of N170 in the left hemisphere as an electrophysiological marker for expertise in reading Chinese. Neurosci. Bull. 28 (5), 577–584. https://doi.org/10.1007/s12264-012-1274-y. Zhou, Y.G., 2001. History of World Character Development. Shanghai Educational Publishing House (in Chinese).
Bentin, S., Mouchetant-Rostaing, Y., Giard, M.H., Echallier, J.F., Pernier, J., 1999. ERP manifestations of processing printed words at different psycholinguistic levels: time course and scalp distribution. J. Cogn. Neurosci. 11 (3), 235–260. https://doi.org/10. 1162/089892999563373. Brem, S., Bucher, K., Halder, P., Summers, P., Dietrich, T., Martin, E., Brandeis, D., 2006. Evidence for developmental changes in the visual word processing network beyond adolescence. Neuroimage 29 (3), 822–837. https://doi.org/10.1016/j.neuroimage. 2005.09.023. Cao, X., Li, S., Zhao, J., Lin, S., Weng, X., 2011. Left-lateralized early neurophysiological response for Chinese characters in young primary school children. Neurosci. Lett. 492 (3), 165–169. https://doi.org/10.1016/j.neulet.2011.02.002. Chaumon, M., Bishop, D.V.M., Busch, N.A., 2015. A practical guide to the selection of independent components of the electroencephalogram for artifact correction. J. Neurosci. Methods 250, 47–63. https://doi.org/10.1016/j.jneumeth.2015.02.025. Chauncey, K., Holcomb, P.J., Grainger, J., 2008. Effects of stimulus font and size on masked repetition priming: an event-related potentials (ERP) investigation. Lang. Cogn. Process. 23 (1), 183–200. https://doi.org/10.1080/01690960701579839. Chen, H.C., Shu, H., 2001. Lexical activation during the recognition of Chinese characters: evidence against early phonological activation. Psychon. Bull. Rev. 8 (3), 511–518. Retrieved from. http://www.ncbi.nlm.nih.gov/pubmed/11700902. Cohen, L., Dehaene, S., 2004. Specialization within the ventral stream: the case for the visual word form area. Neuroimage 22 (1), 466–476. https://doi.org/10.1016/j. neuroimage.2003.12.049. Dehaene, S., Cohen, L., 2011. The unique role of the visual word form area in reading. Trends Cogn. Sci. 15 (6), 254–262. https://doi.org/10.1016/J.TICS.2011.04.003. Dehaene, S., Cohen, L., Sigman, M., Vinckier, F., 2005. The neural code for written words: a proposal. Trends Cogn. Sci. 9 (7), 335–341. https://doi.org/10.1016/j.tics.2005.05. 004. Delorme, A., Makeig, S., 2004. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J. Neurosci. Methods 134 (1), 9–21. https://doi.org/10.1016/j.jneumeth.2003.10.009. Fraga González, G., Žarić, G., Tijms, J., Bonte, M., Blomert, L., van der Molen, M.W., 2014. Brain-potential analysis of visual word recognition in dyslexics and typically reading children. Front. Hum. Neurosci. 8, 474. https://doi.org/10.3389/fnhum. 2014.00474. Fraga González, G., Žarić, G., Tijms, J., Bonte, M., Blomert, L., Leppänen, P., van der Molen, M.W., 2016. Responsivity to dyslexia training indexed by the N170 amplitude of the brain potential elicited by word reading. Brain Cogn. 106, 42–54. https://doi. org/10.1016/j.bandc.2016.05.001. Koenig, T., Kottlow, M., Stein, M., Melie-García, L., 2011. Ragu: a free tool for the analysis of EEG and MEG event-related scalp field data using global randomization statistics. Comput. Intell. Neurosci. 1–14. 2011. https://doi.org/10.1155/2011/938925. Kronbichler, M., Hutzler, F., Wimmer, H., Mair, A., Staffen, W., Ladurner, G., 2004. The visual word form area and the frequency with which words are encountered: evidence from a parametric fMRI study. Neuroimage 21 (3), 946–953. https://doi.org/ 10.1016/j.neuroimage.2003.10.021. Lehmann, D., Skrandies, W., 1980. Reference-free identification of components of checkerboard-evoked multichannel potential fields. Electroencephalogr. Clin. Neurophysiol. 48 (6), 609–621. https://doi.org/10.1016/0013-4694(80)90419-8. Lin, S.E., Chen, H.C., Zhao, J., Li, S., He, S., Weng, X.C., 2011. Left-lateralized N170 response to unpronounceable pseudo but not false Chinese characters-the key role of orthography. Neuroscience 190, 200–206. https://doi.org/10.1016/j.neuroscience. 2011.05.071. Marshall, J.C., Newcombe, F., 1973. Patterns of paralexia: a psycholinguistic approach. J. Psycholinguist. Res. 2 (3), 175–199. https://doi.org/10.1007/BF01067101. Maurer, U., Brem, S., Bucher, K., Brandeis, D., 2005. Emerging neurophysiological specialization for letter strings. J. Cogn. Neurosci. 17 (10), 1532–1552. https://doi.org/ 10.1162/089892905774597218. Maurer, U., Brem, S., Kranz, F., Bucher, K., Benz, R., Halder, P., et al., 2006. Coarse neural tuning for print peaks when children learn to read. Neuroimage 33 (2), 749–758.
8
Contents lists available at ScienceDirect
Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia
Familiarity with visual forms contributes to a left-lateralized and increased N170 response for Chinese characters
T
Licheng Xuea,b,c, Urs Maurerd,e, Xuchu Wenga,b,c,f, Jing Zhaob,c,∗ a
Center for Cognition and Brain Disorder, Hangzhou Normal University, China Institute of Psychological Sciences, Hangzhou Normal University, China c Zhejiang Key Laboratory for Research in Assessment of Cognitive Impairments, China d Department of Psychology, Chinese University of Hong Kong, China e Brain and Mind Institute, Chinese University of Hong Kong, China f Institute for Brain Research and Rehabilitation, South China Normal University, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: N170 Chinese character Visual expertise Visual form familiarity
While skilled readers produce an increased and left-lateralized event-related-potential (ERP) component, known as N170, for strings of letters compared to strings of less familiar units, it remains unclear whether perceptual familiarity plays an important role in driving the increased and left-lateralized N170 for print. The present study addressed this issue by examining N170 responses for regular Chinese characters and cursive Chinese characters which are visually less familiar regarding their form, yet with phonological and semantic properties. Stroke combinations, which are with unfamiliar visual form and without phonological or semantic properties, were used as low-level control stimuli. Twenty college students (22.6 ± 1.2 years) were examined. A content-irrelevant color matching task was used to control for differences of attention load across familiar and unfamiliar stimuli. A left-lateralized N170 was evoked only by regular characters, but not by cursive characters or stroke combinations. Moreover, cursive characters, which are principally readable but visually unfamiliar, produced a lower N170 than regular characters, and no N170 difference was found compared with stroke combinations. These results suggest that visual form familiarity serves as an important driver for the increased and left-lateralized N170 response.
1. Introduction Through extensive experience with learning to read, humans acquire the ability to quickly recognize visual words, and show specialized responses in their brain (Schlaggar and McCandliss, 2007). In skilled readers, visual words evoke an increased and left-lateralized event-related potential (ERP) component peaking within 200 ms after stimulus onset, usually termed N170 or N1 (Bentin et al., 1999; Maurer et al., 2006). Specifically, the N170 amplitude is larger and left-lateralized for familiar orthographic stimuli such as real words and pseudowords compared to less familiar stimuli such as consonant strings and symbol strings in alphabetic scripts (Brem et al., 2006; Maurer et al., 2005; Maurer et al., 2006,; Araújo et al., 2015; Fraga González et al., 2014; Fraga González et al., 2016, Varga et al., 2018). Similarly, real and pseudo Chinese characters produce an increased and left-lateralized N170 response relative to less familiar false characters and stroke combinations (Lin et al., 2011; Yang et al., 2017; Zhao et al., 2012; Cao et al., 2011). However, it remains unclear whether perceptual ∗
familiarity plays an important role in driving the increased and leftlateralized N170 for print. A straightforward approach to address this issue is to examine N170 responses for stimuli with less familiar visual forms while having phonological and semantic properties. Chinese cursive characters accord with these properties. In the modern Chinese writing system, cursive characters are one font of printed Chinese although less used than regular characters (Zhou, 2001). Cursive characters are generally employed for daily use in some situations (e.g., billboards and trademarks) as well as in calligraphic art (Wang, 1958), and thus skilled Chinese readers can recognize common cursive characters. In the framework of visual word reading, visual input is first transformed into abstract visual word forms (invariant under changes in location, size, case and font) in the visual word form system, and then activates semantic and phonological processing (at this stage, surface visual features would no longer play a role), although top-down modulation may play a role in early visual processing (Cohen and Dehaene, 2004; see Dehaene et al., 2005 for a review; Warrington and
Corresponding author. 2318 Yuhang Tang Rd., Yuhang District, Hangzhou, 311121, China. E-mail address: [email protected] (J. Zhao).
https://doi.org/10.1016/j.neuropsychologia.2019.107194 Received 22 March 2019; Received in revised form 5 August 2019; Accepted 12 September 2019 Available online 19 September 2019 0028-3932/ © 2019 Published by Elsevier Ltd.
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
frequency of occurrence of 1170 per million (the modern Chinese frequency dictionary, 1985). Cursive characters were the same characters typed in cursive font. They were comparable with regular characters in phonological and semantic properties. Stroke combinations were constructed by recombining the strokes of real characters. They were matched with regular characters in basic visual stimulus properties (i.e., stroke and visual complexity), but the internal structure was disrupted, and they had no pronunciation or meaning. For each type, 36 stimuli were created. To further confirm the visual form familiarity of cursive characters and regular characters, we asked another twenty participants to rate the visual form familiarity of the two types of stimuli in a seven-point scale (1 = never to 7 = frequent). The score of the visual form familiarity judgment was analyzed in a paired-sample t-test with Stimulus Type as a within-subject factor. Results showed that the rating score of cursive characters was significantly lower than that of regular characters (3.59 vs. 6.54, t(19) = 11.34, p < 0.001).
Table 1 Description and examples of different types of stimuli. Type of stimuli
Example
Familiarity
Pronunciation
Meaning
Regular character
high
high
high
Cursive character
low
high
high
Stroke combination
0
0
0
Shallice, 1980). Thus, in this framework, cursive characters, once transformed into abstract visual forms, are essentially the same as regular characters in terms of phonological and semantic properties (Marshall and Newcombe, 1973; Price and Devlin, 2003). In the present study, we aimed to examine whether perceptual familiarity plays an important role in driving the increased and left-lateralized N170 for print by comparing cursive characters (e.g., ) and regular characters (e.g., ) (see Table 1). In addition, stroke combinations (e.g., ) were formed by randomly recombining the strokes of the regular characters as a low-level visual control. Stroke combinations were matched with regular characters in terms of the low-level visual features, while the combined pattern of strokes was entirely disrupted, and had no phonological or semantic properties. To control for differences of attention load across familiar and unfamiliar stimuli, we used a content-irrelevant color matching task with a short presentation time (300 ms), in which participants were asked to make judgments about the color of the stimuli only and to ignore the other properties (see Zhao et al., 2012 for more discussion). Similar to results in previous studies (Cao et al., 2011; Lin et al., 2011; Zhao et al., 2012), we expected regular characters would evoke an increased and left-lateralized N170 response, while stroke combinations would evoke a lower and less left-lateralized N170 response relative to regular characters. Importantly, due to reduced visual form familiarity, we expected a reduced and less left-lateralized N170 response for cursive characters if perceptual familiarity plays an important role to drive this N170 effect. Alternatively, we expected a similar N170 response for the two types of characters that have the same linguistic properties regardless of fonts (regular vs. cursive) if the increased and left-lateralized N170 effect is not modulated by perceptual familiarity.
2.3. Procedure Each stimulus was presented in one of the three colors green, red or yellow on a gray background in the center of the screen extending 1.32° in visual angle. Each trial was constructed as follows: a mean interval of 1600 ms (1450 ms–1750 ms, with ± 75 ms random jitter) and fixed stimulus duration of 300 ms was used. Participants were instructed to press a button, whenever a color of a stimulus occurred consecutively two times in a row (i.e., the one-back color repetition detection) (as shown in Fig. 1b). All stimuli were presented four times as non-targets, and a subset was additionally presented as targets (17% of all stimuli). The three conditions were presented randomly within a block. The order of the left and right hand to respond was counterbalanced between participants. After the ERP experiment, participants were instructed to read aloud all the regular and cursive Chinese characters presented in the ERP experiment. Each character was presented for 300 ms, and participants pronounced the characters as accurately and fast as possible. Both accuracy and reaction time (the time interval between item onset and the pronunciation onset) were recorded. To evaluate the difficultly of visual form processing of the three stimulus types, we asked additional 27 participants to participate in a visual form matching task. The stimuli were the same as those used in the ERP experiment, but were presented in three pairs (i.e., regular character-regular character, cursive character-cursive character, stroke combination-stroke combination). Participants were asked to judge whether the two stimuli are same. The trial started with a fixation cross presented at the center of the screen for 800 ms. The stimulus pair was displayed for 300 ms. All stimuli were presented in black against a gray background, and subtended 1.3° of visual angle at a viewing distance of 100 cm. Horizontally, the two characters were 3.6° of visual angle apart. The test was composed of 180 trials (90 ‘same’ trials and 90 ‘different’ trials). For each stimulus type, there were 60 trials with 30 ‘same’ trials and 30 ‘different’ trials. The order of trials was randomized and the keys corresponding to same/different responses were counterbalanced between participants.
2. Methods The research protocol reported here was approved by a local institutional review board and written informed consent was obtained from the participants in the present study. 2.1. Participants There were twenty-two participants in the experiment (11 of them were male). One of the participants did not finish the experiment, and the hit rates of one participant in the one-back color repetition detection task was only at random level. Therefore, two participants (2 males) were excluded from the analyses. Mean age of the remaining participants was 22.6 years ( ± 1.2 years). Participants were all undergraduate students at the university. They were all right-handed according to the Edinburgh handedness inventory (Oldfield, 1971), with normal or corrected-to-normal vision and all reported themselves to be in good health.
2.4. Electrophysiological recording and analysis Electroencephalogram (EEG) signal was recorded from 64 Ag/AgCl electrodes secured in an elastic cap (NeuroScan Inc., EI Paso, TX, USA). And horizontal electrooculogram (EOG) was recorded with additional electrodes on the left and right external cantus, while the vertical EOG was recorded with two electrodes placed above and below the left eye. An electrode located between Cz and CPz served as a reference, but an average reference transformation was used for the data analysis. An electrode between Fpz and Fz on the cap served as ground. All channels were amplified with a band pass from AC 0.1 to 100 HZ and the sampling rate was 1000 Hz. Electrode impedances were kept below 5 kΩ.
2.2. Material Three types of stimuli were used: regular characters, cursive characters, and stroke combinations. Examples were illustrated in Fig. 1a. All regular characters had a single-structure, composed of only one logographeme and five strokes on average (range: four-seven) with a 2
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 1. a. Examples of stimuli: regular character (RC), cursive character (CC), and stroke combination (SC). b. Sketch map of the one-back color repetition detection task. The arrow implied the direction of presentation.
at each time point in three pairwise comparisons (regular character vs. cursive character, regular character vs. stroke combination, cursive character vs. stroke combination). To correct for multiple comparisons, RAGU has an in-built procedure (Global Duration Test), which uses a duration threshold that is detected using a randomization procedure. In other words, differences between two types of stimuli are detected that last longer than would be expected by chance (Koenig et al., 2011).
The EEG data was preprocessed using EEGLAB software (Delorme and Makeig, 2004). First, the data was digitally filtered by band-pass (0.5–30HZ). The Artifact Subspace Reconstruction (ASR) method was applied to remove artifacts and bad channels, which were reconstructed from neighboring channels (Mullen et al., 2013). Secondly, the data was referenced to a common average reference. Finally, EOG artifacts were corrected using ICA and SASICA tools (Chaumon et al., 2015), in which the ICA components which correlated with vEOG and hEOG channels were removed. The EEG was then segmented into epochs of 700 ms including a 100 ms pre-stimulus baseline, used for baseline correction. Epochs with extreme amplitude exceeding threshold ( ± 80 μV) within the interested time window were rejected. Trials of non-targets were averaged separately for the three conditions. To minimize selection bias of electrodes mainly based on visual inspection and allow for comprehensive multichannel ERP analyses (Maurer et al., 2005; 2006; Wang and Maurer, 2017), we identified EEG segments for further analyses by the maxima of global field power (GFP) curve after averaging all grand means across all stimulus types and subjects (Lehmann and Skrandies, 1980). Similar to the procedure used in previous studies (Bentin et al., 1999; Lin et al., 2011; Wong et al., 2005; Zhao et al., 2012), five pairs of channels in occipito-temporal areas were selected according to the topographic maxima in the negative field over both hemispheres, including PO7/8, P7/8, PO5/6, PO3/4, and CB1/2. Based on the GFP curve, the first ERP segment was from 57 ms to 105 ms (bilateral troughs of the first wave), and was identified as the P100 component based on the bifocal posterior positivity. The following segment (106198 ms) was rather long, with a sharp initial peak and a subsequent decrease, which was identified as the N170 component. And the following segment (199-274 ms for maxima) was identified as the P2 component. The peak value of amplitude on these channels for these three components was obtained by the peak detection, respectively. And after that, the statistics was based on the peak amplitudes. Specifically, the mean of the P100, N170 or P2 peak amplitude of these channels on the left hemisphere (i.e., PO7/P7/PO5/PO3/CB1) and the right hemisphere (i.e., PO8/P8/PO6/PO4/CB2) was computed, respectively. The latency of these components in one pair of channels (PO7/PO8, the topographic maxima in the negative/positive field on the occipito-temporal area over both hemispheres) was detected. Finally, peak amplitudes and latencies of P100, N170 or P2 were analyzed in Analyses of Variance (ANOVA, procedure GLM) with two within-subject factors Stimulus Type (regular characters, cursive characters, stroke combinations) and Hemisphere (left, right), respectively. And further, to examine whether the peak amplitudes at the occipito-temporal electrodes reflected the effect across ERP maps, the data was analyzed using topographic analyses of variance (TANOVA) implemented in the Randomization Graphical User interface (RAGU) software (Koenig et al., 2011). Specifically, the TANOVA was computed
3. Results 3.1. Behavioral results 3.1.1. Color matching Hit rates and reaction times were analyzed for target trials of the three stimulus types. Means and standard deviations are illustrated in Table 2. Due to a file error, the behavioral data of four participants were lost. Since the participants were asked to respond as accurately as possible, their overall hit rates were high for all stimulus types. Results in one-way ANOVA with Stimulus Type (regular character, cursive character, stroke combination) as a within subject factor revealed a significant main effect, F (2, 30) = 14.62, p < 0.001, η2G = 0.49. Results in the post-hoc test showed that the hit rate for color repetition detection in the stroke combination condition was lower than that in the regular character condition (p = 0.01) and cursive character condition (p < 0.001), while no difference was found between regular and cursive characters (p = 0.29). For reaction times, the one-way ANOVA did not show a significant main effect of Stimulus Type, F (2, 30) = 0.96, p = 0.39, η2G = 0.06. 3.1.2. Chinese character naming After the EEG experiment, participants were measured in a Chinese character naming test, which consisted of the regular and cursive characters presented in the EEG experiment. Participants could name regular and cursive characters at the same accuracy level (t18 < 1, p > 0.05), and the accuracies for both types of stimuli were close to the ceiling level in naming regular characters (0.98) and cursive characters (0.97). Importantly, the cursive characters that were not correctly named were not consistent across participants. This together with very high accuracy indicates that all cursive characters were clear to each participant. For reaction time, results in a paired-sample t-test showed Table 2 Hit rate and reaction time [Mean (SE)].
Hit rate (%) Reaction time (ms)
3
Regular
Cursive
Stroke combination
95.57 (1.60) 495.30 (32.03)
96.87 (1.80) 493.04 (32.04)
92.45 (1.70) 502.82 (31.25)
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
the left hemisphere, the amplitude of N170 evoked by regular characters was significantly larger than those evoked by cursive characters (p = 0.002), and stroke combinations (p = 0.001), respectively. No N170 difference was found between cursive characters and stroke combinations (p = 0.17). However, in the right hemisphere, there was no significant difference between any two stimulus categories (regular characters vs. cursive characters, p = 0.82; regular characters vs. stroke combinations, p = 0.17). Similarly, no significant difference was found between cursive characters and stroke combinations (p = 0.46).
Table 3 Accuracy and reaction time [Mean (SE)].
Accuracy (%) Reaction time (ms)
Regular
Cursive
Stroke combination
94.94 (0.70) 707.39 (31.64)
90.93 (0.70) 741.51 (38.48)
89.01 (1.20) 750.58 (28.95)
that participants took longer time to name cursive characters (731 ± 98 ms) than regular characters (618 ± 57 ms), t18 = 6.70, p < 0.001.
3.2.3. N170 peak latency Table 4 illustrates means and standard errors of N170 peak latency for the three types of stimuli. Results in the two-way ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed significant main effects of Stimulus Type, F (2, 38) = 9.41, p < 0.001, η2G = 0.33, and Hemisphere F (1, 19) = 4.64, p = 0.044, η2G = 0.20. The interaction of Stimulus Type by Hemisphere was not significant F (2, 2 38) = 2.0, p = 0.16, ηG = 0.09. Results in the post-hoc test showed that N170 latency for regular characters was shorter than that for cursive characters (p = 0.003), and marginally shorter than that for stroke combinations (p = 0.05), while no difference was found between cursive characters and stroke combinations (p = 0.45).
3.1.3. Visual form matching Means and standard deviations of accuracy and reaction time are illustrated in Table 3. Data was analyzed in a one-way ANOVA with Stimulus Type as a within-subject factor. For the accuracy, results showed the main effect of Stimulus Type was significant, F(2, 52) = 14.53, p < 0.001. Results in the post-hoc test (with Bonferroni adjustment) showed that the accuracy of regular character was higher than that of cursive character (p < 0.001) and stroke combination (p < 0.001), while no difference was found between cursive character and stroke combination (p = 0.46). For the reaction time, results showed the main effect of Stimulus Type was significant, F(2, 52) = 8.75, p = 0.003. Results in the post-hoc test (with Bonferroni adjustment) showed that the reaction time of regular character was shorter than that of cursive character (p = 0.021) and stroke combination (p < 0.001), while no difference was found between cursive character and stroke combination (p = 0.99). These results suggest that the difficulty of visual form processing was regular character < cursive character = stroke combination.
3.2.4. P2 peak amplitude Results in the two-way ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed a significant main effect of Stimulus Type, F (2, 38) = 7.90, p < 0.01, η2G = 0.29. However, the main effect of Hemisphere (F (1, 19) = 0.34, p = 0.57, η2G = 0.02) and the interaction of Stimulus Type and Hemisphere (F (2, 38) = 0.31, p = 0.74, η2G = 0.02) were not significant. Post hoc tests within Stimulus Type showed that the P2 amplitude evoked by regular characters was significantly larger than that evoked by cursive characters (p = 0.001), while marginally larger than that evoked by stroke combinations (p = 0.07). No difference in P2 amplitude was found between cursive characters and stroke combinations (p = 0.90).
3.2. ERP results Fig. 2a shows grand average waveforms for three stimulus categories at the occipito-temporal electrodes on the left and right hemispheres. A positive deflection peaking within 100 ms, and a negative deflection peaking within 200 ms were clearly visible for each stimulus type over bilateral occipito-temporal sites.
3.2.5. P2 peak latency Table 5 illustrates means and standard error of P2 peak latency for the three types of stimuli. Results in the two-way ANOVA showed a significant main effect of Stimulus Type, F (2, 38) = 14.38, p < 0.001, η2G = 0.43. The main effect of Hemisphere, F (1, 19) = 3.12, p = 0.10, η2G = 0.14 and the interaction of Stimulus Type by Hemisphere, F (2, 2 38) = 2.40, p = 0.11, ηG = 0.11 were not significant. Results in the posthoc test of the Stimulus Type showed that the P2 latency for regular characters was shorter than that for cursive characters (p < 0.001), and stroke combinations (p = 0.02), while no significant difference was found between cursive characters and stroke combinations (p = 0.10).
3.2.1. P100 peak amplitude Results in the ANOVA showed neither the main effect of the Stimulus Type, F (2, 38) = 1.36, p = 0.27, η2G = 0.07, nor the main effect of Hemisphere, F (1, 19) = 3.21, p = 0.09, η2G = 0.14 was significant. The interaction of Stimulus Type and Hemisphere was not significant, F (2, 2 38) = 0.61, p = 0.54, ηG = 0.03. These results showed no significant differences in low-level visual features across the three types of stimuli. 3.2.2. N170 peak amplitude Fig. 2b shows the average of N170 peak amplitudes for three types of stimuli at the occipito-temporal electrodes on the left and right hemisphere. ANOVA with the two within-subject factors Stimulus Type (regular character, cursive character, stroke combination) and Hemisphere (left, right) showed a significant main effect of the Stimulus Type, F (2, 38) = 9.87, p < 0.01, η2G = 0.34, while the main effect of Hemisphere was not significant, F (1, 19) = 2.99, p = 0.10, η2G = 0.14. Importantly, the interaction of Stimulus Type and Hemisphere was significant, F (2, 38) = 5.54, p = 0.01, η2G = 0.23. For the lateralization, we observed a left-lateralization of N170 was specific to regular characters (i.e., stimuli with high familiarity with visual forms). Specifically, results in the analysis of N170 amplitude between the left and right hemispheres showed only a significant leftlateralization of N170 response for regular characters (p = 0.04), while neither cursive character (p = 0.14) nor stroke combinations (p = 0.22) evoked such a left-lateralized N170 response. The following results reflect subsequent analyses for differences between stimulus categories separately for each hemisphere. Regarding
3.2.6. Topographic map TANOVA results To investigate whether the difference showed at the occipito-temporal electrodes at N170 peak latency existed across the ERP maps, the TANOVAs were used for comparisons between stimulus categories. Similar to the N170 analysis, the trial length of data sample was 800 ms including 100 ms baseline. To examine our hypothesis, we focused on the first 400 ms after stimulus onset. For regular-cursive comparison, as shown in Fig. 3b and c, the time windows that survived the Global Duration Test are shown in green, and the topographies in the difference t-maps are separately illustrated for the time periods corresponding to different components (N1/N170, P2). We divided the segment into two time-window as counterparts of N170 and P2 components, based on the increased p-value that coincides with a change in the topography of the difference wave. Similar to the N170 and P2 time ranges (see Fig. 3a), results in TANOVA statistics showed significant difference between regular and cursive characters 4
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 2. a. Grand average waveforms with 95% confidence intervals (error bands) at occipito-temporal channels on both the left and the right hemispheres b. Box plots of average of N170 peak amplitude evoked by different stimulus categories on left and right hemispheres. Lines in the boxes denote means, **p < 0.01, *p < 0.05. Table 4 N170 peak latency [Mean (SE)].
Left hemisphere Right hemisphere
Table 5 P2 peak latency [Mean (SE)].
Regular
Cursive
Stroke combination
154.30 (3.41) 146.20 (4.14)
157.78 (3.67) 153.28 (4.60)
156.55 (3.60) 152.23 (4.25)
Left hemisphere Right hemisphere
5
Regular
Cursive
Stroke combination
232.20 (4.36) 234.73 (4.88)
239.75 (4.86) 250.54 (4.50)
238.60 (4.92) 244.56 (4.86)
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
Fig. 3. a. N170 waveforms averaged with 95% confidence intervals (error bands) across five occipito-temporal channels on both the left and the right hemispheres, respectively. b., d., f. TANOVA results of paired comparisons of three stimulus categories (800 ms, including 100 ms baseline, RC-regular characters, CC-cursive characters, SC-stroke combinations). Green periods survived Global Duration Test (p < 0.05). c., e., g. t-maps during the time window for N170 and P2 components, respectively. ∗∗∗means p-value of TANOVA < 0.001.
semantic properties. With the goal of identifying the role of visual form familiarity and controlling for differences of attention load across familiar and unfamiliar stimuli, we used a content-irrelevant color matching task with a short duration of stimulus presentation (300 ms). In various commonly-used explicit language tasks with long stimulus durations, participants are often faster in response to target letters embedded in native than non-native words and more accurate at detecting repetitions of orthographic stimuli (such as words or letter strings) than at detecting repetitions of less familiar symbol strings (Maurer et al., 2005; Maurer et al., 2008). Therefore, the increased N170 in response for orthographic stimuli reported in these studies may be partly explained by attentional bias to familiar stimulus categories (see Zhao et al., 2012 for more discussion). In contrast, in the current study, we did not find any behavioral response difference between regular and cursive characters. Hence, the task we used here effectively controlled for possible confounding caused by different attentional loads between familiar and unfamiliar stimuli. An important finding of the present study is that the left-lateralized and increased N170 was sensitive to familiarity of visual form. We found that only regular characters evoked a significantly left-lateralized
between 112 ms and 157 ms for N170 and 157 to 252 ms for P2, respectively (Fig. 3b). The t-maps of the difference between these two stimulus categories are used for illustration (Fig. 3c). In addition, a similar pattern was found in the regular character-stroke combination comparison, 108-166 ms for N170 and 116-243 ms for P2 (t-maps seen in Fig. 3d and e). By contrast, as shown in Fig. 3f, there was no significant difference in the cursive character vs. stroke combination comparison in the similar time range as the regular character vs. stroke combination comparison (108-166 ms for N170 and 166-243 ms for P2, see Fig. 3f and g). 4. Discussion Although there is a consensus that the increased left-lateralized N170 response for print is depending on extensive experience with this visual type, it remains unclear whether perceptual familiarity plays an important role in driving increased and left-lateralized N170 for print. We addressed this issue by comparing N170 responses for three types of visual categories (regular Chinese character, cursive Chinese character, and stroke combination) with different degrees of visual form familiarity, yet only regular and cursive characters have phonological and 6
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
González et al., 2016), while in the present study we used an implicit task that does not require specific phonological and semantic processing. Results in some of previous studies suggest that N170 sensitivity to words is modulated by task demands (Strijkers et al., 2015; Wang and Maurer, 2017). It is possible that the task-demand may contribute to the effect of visual form familiarity on the N170 responses. With respect to script types, in addition to Chinese, the increased and left-lateralized N170 for print has been found in scripts with different alphabetic scripts (e.g., Dutch in Fraga González et al., 2014; Fraga González et al., 2016; Portuguese in Araújo et al., 2015). Some of the results in previous studies suggest that reading Chinese may not automatically engage strong phonological processing as reading alphabetic words (Chen and Shu, 2001; Yeh and Li, 2002). Thus, script-type may also contribute to the effect of visual form familiarity on the N170 responses. Finally, we would like to acknowledge a few limitations of the present study and note some directions for future studies. First, the present study only examined Chinese characters processing on Chinese participants. Future research needs to examine whether this pattern of N170 lateralization can be extended to other writing systems. Second, the current findings are not sufficient to directly address whether visual form familiarity a lone can drive the increased and left-lateralized N170. In future studies one can ask participants to learn new symbols and manipulate exposure. One can also investigate the effect of Chinese Characters processing in alphabetic readers who are learning to read Chinese, and vice versa. In conclusion, using an implicit color matching task, we found an increased and left-lateralized N170 response for the regular characters compared to the cursive characters that were less familiar in terms of visual form but had phonological and semantic properties. We also examined the N170 evoked by the stroke combinations that were with less familiar visual form and had no pronunciation and meaning, and found no N170 difference between cursive characters and stroke combinations. These findings provide empirical evidence that the perceptual familiarity plays an important role to drive the increased and leftlateralized N170 response for print.
N170, while either cursive characters or stroke combinations evoked a bilateral N170 response. Moreover, detailed analysis of N170 amplitude at channels over the left posterior region showed that regular characters evoked stronger N170 amplitude relative to both cursive characters and stroke combinations. As a font that is relatively rarely used in the modern Chinese writing system, cursive characters are visually less familiar than regular characters in standard style of handwriting (see Zhou et al., 2001 for detailed discussion). This conclusion was confirmed by our results in visual form familiarity rating, which showed that cursive characters were visually less familiar than regular characters for skilled readers. In the naming task, all participants correctly named virtually all cursive characters (the accuracy was 0.97). These results together suggest that the cursive characters used in this study were readable, while visually unfamiliar. ERP results show that cursive characters produced weaker and less left-lateralized N170 than regular characters and a similar N170 response to stroke combinations that have no pronunciation and meaning. These results together demonstrated that perceptual familiarity plays an important role to drive the left-lateralized and increased N170. The remaining issue is whether the increased and left-lateralized N170 response could be related to the mapping of visual form of a character to its phonological/semantic properties. Compared to regular characters, it may be more difficult to decode the cursive characters (access to phonological and semantic information). This interpretation seems to be consistent with results of the name task, which showed that participants took longer time to name cursive characters than regular character. However, it is important to note that participants also took longer time to process cursive characters than regular characters in visual form matching. Therefore, it is also possible that the longer reaction times of naming cursive characters may be due to the difficulty of visual form processing. Moreover, this interpretation is not supported by the N170 results of the current study. If a more effortful access of phonological and semantic information was reflected in the N170 component, an increased or prolonged N170 would be expected (similar to increased activation for pseudowords compared to words or for low-frequency compared to high-frequency words in fMRI studies; e.g., Kronbichler et al., 2004). However, we did not find this N170 effect. Rather our ERP results suggest that the more difficult lexical access is reflected in the subsequent P2 component, with not only larger amplitude but also shorter latency for regular characters compared with cursive characters and stroke combinations. This interpretation is compatible with previous studies suggesting that the P2 is associated with later attentional processes and the mapping of visual form of a word to its phonological properties (Barnea and Breznitz, 1998; Bentin et al., 1999; Sereno et al., 1998; Zhang et al., 2009). There is an ongoing discussion about whether visual expertise effects in reading are driven by bottom-up processes or by interactions between automatic top-down predictions and bottom-up input (Dehaene and Cohen, 2011; Price and Devlin, 2011). While our current results are unlikely to reflect effortful access of phonological and semantic representation, these results are not sufficient to resolve the debate regarding the nature of the visual expertise effects. Therefore, it is highly possible that the visual form familiarity effects in the N170 reflect the sensitivity of feature detectors to bottom-up visual input (Dehaene and Cohen, 2011). However, there is another possibility that the visual familiarity effects in the N170 reflect prediction errors arising from matching automatic top-down predictions from higher-level visual or language regions with visual input (Price and Devlin, 2011). Whether the visual expertise effects in the N170 is driven by purely bottomup processing or by interactive processing, and whether potential predictions are arising from visual regions or language regions remains to be elucidated in future studies. It is worth noting that earlier studies have examined the left-lateralized and increased N170 for print in different tasks (e.g., lexical decision in Simon et al., 2007; naming in Chauncey et al., 2008; Wang and Maurer, 2017; repetition detection in Fraga González et al., 2014; Fraga
Funding This work was supported by grants from the National Natural Science Foundation of China (NSFC) (grant No.: 31771229), a grant from the Zhejiang Province Natural Science Foundation (grant No.: LY17C090008), and Project of the Ministry of Science and Technology of the People's Republic China (2016YFE0130400). Declaration of interest The authors have no competing interests. CRediT authorship contribution statement Licheng Xue: Data curation, Formal analysis, Methodology, Resources, Software, Visualization, Writing - original draft. Urs Maurer: Formal analysis, Methodology, Resources, Software, Writing review & editing. Xuchu Weng: Conceptualization, Funding acquisition, Methodology, Resources, Software, Writing - review & editing. Jing Zhao: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Project administration, Supervision, Validation, Writing - review & editing. References Araújo, S., Faísca, L., Bramão, I., Reis, A., Petersson, K.M., 2015. Lexical and sublexical orthographic processing: an ERP study with skilled and dyslexic adult readers. Brain Lang. 141, 16–27. https://doi.org/10.1016/j.bandl.2014.11.007. Barnea, A., Breznitz, Z., 1998. Phonological and orthographic processing of Hebrew words: electrophysiological aspects. J. Genet. Psychol. 159 (4), 492–504. https://doi. org/10.1080/00221329809596166.
7
Neuropsychologia 134 (2019) 107194
L. Xue, et al.
https://doi.org/10.1016/j.neuroimage.2006.06.025. Maurer, U., Zevin, J.D., McCandliss, B.D., 2008. Left-lateralized N170 effects of visual expertise in reading: evidence from Japanese syllabic and logographic scripts. J. Cogn. Neurosci. 20 (10), 1878–1891. https://doi.org/10.1162/jocn.2008.20125. Modern Chinese Frequency Dictionary. Beijing Language Institute Press, Beijing (In Chinese). Mullen, T., Kothe, C., Chi, Y.M., Ojeda, A., Kerth, T., Makeig, S., et al., 2013. Real-time modeling and 3D visualization of source dynamics and connectivity using wearable EEG. In: Conference Proceedings : Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2013, pp. 2184–2187. https://doi.org/10.1109/ EMBC.2013.6609968. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9 (1), 97–113. https://doi.org/10.1016/0028-3932(71) 90067-4. Price, C.J., Devlin, J.T., 2003. The myth of the visual word form area. Neuroimage 19 (3), 473–481. https://doi.org/10.1016/S1053-8119(03)00084-3. Price, C.J., Devlin, J.T., 2011. The Interactive Account of ventral occipitotemporal contributions to reading. Trends Cogn. Sci. 15 (6), 246–253. https://doi.org/10.1016/j. tics.2011.04.001. Schlaggar, B.L., McCandliss, B.D., 2007. Development of neural systems for reading. Annu. Rev. Neurosci. 30 (1), 475–503. https://doi.org/10.1146/annurev.neuro.28. 061604.135645. Sereno, S.C., Rayner, K., Posner, M.I., 1998. Establishing a time-line of word recognition. Neuroreport 9 (10), 2195–2200. https://doi.org/10.1097/00001756-19980713000009. Simon, G., Petit, L., Bernard, C., Rebaï, M., 2007. N170 ERPs could represent a logographic processing strategy in visual word recognition. Behav. Brain Funct. 3, 1–11. https://doi.org/10.1186/1744-9081-3-21. Strijkers, K., Bertrand, D., Grainger, J., 2015. Seeing the same words differently: the time course of automaticity and top-down intention in reading. J. Cogn. Neurosci. 27 (8), 1542–1551. https://doi.org/10.1162/jocn_a_00797. Varga, V., Tóth, D., Csépe, V., 2018. Orthographic‐Phonological Mapping and the Emergence of Visual Expertise for Print: A Developmental Event‐Related Potential Study. Child Dev 1–13. https://doi.org/10.1111/cdev.13159. Wang, F.Y., 1958. Introduction to Chinese Cursive Script. Institute of Far Eastern Languages, Yale University Retrieved from. https://yalebooks.yale.edu/book/ 9780887100338/chinese-cursive-script. Wang, F., Maurer, U., 2017. Top-down modulation of early print-tuned neural activity in reading. Neuropsychologia 102, 29–38. https://doi.org/10.1016/j.neuropsychologia. 2017.05.028. Warrington, E.K., Shallice, T., 1980. Word-form dyslexia. Brain 103 (1), 99–112. https:// doi.org/10.1093/brain/103.1.99. Wong, A.C.N., Gauthier, I., Woroch, B., Debuse, C., Curran, T., 2005. An early electrophysiological response associated with expertise in letter perception. Cognit. Affect Behav. Neurosci. 5 (3), 306–318. https://doi.org/10.3758/CABN.5.3.306. Yang, H., Zhao, J., Gaspar, C.M., Chen, W., Tan, Y., Weng, X., 2017. Selectivity of N170 for visual words in the right hemisphere: evidence from single-trial analysis. Psychophysiology 54 (8), 1128–1137. https://doi.org/10.1111/psyp.12867. Yeh, S.-L., Li, J.-L., 2002. Role of structure and component in judgments of visual similarity of Chinese characters. J. Exp. Psychol. Hum. Percept. Perform. 28 (4), 933–947. Retrieved from. http://www.ncbi.nlm.nih.gov/pubmed/12190259. Zhang, Q., Zhang, J.X., Kong, L., 2009. An ERP study on the time course of phonological and semantic activation in Chinese word recognition. Int. J. Psychophysiol. 73 (3), 235–245. https://doi.org/10.1016/J.IJPSYCHO.2009.04.001. Zhao, J., Li, S., Lin, S.E., Cao, X.H., He, S., Weng, X.C., 2012. Selectivity of N170 in the left hemisphere as an electrophysiological marker for expertise in reading Chinese. Neurosci. Bull. 28 (5), 577–584. https://doi.org/10.1007/s12264-012-1274-y. Zhou, Y.G., 2001. History of World Character Development. Shanghai Educational Publishing House (in Chinese).
Bentin, S., Mouchetant-Rostaing, Y., Giard, M.H., Echallier, J.F., Pernier, J., 1999. ERP manifestations of processing printed words at different psycholinguistic levels: time course and scalp distribution. J. Cogn. Neurosci. 11 (3), 235–260. https://doi.org/10. 1162/089892999563373. Brem, S., Bucher, K., Halder, P., Summers, P., Dietrich, T., Martin, E., Brandeis, D., 2006. Evidence for developmental changes in the visual word processing network beyond adolescence. Neuroimage 29 (3), 822–837. https://doi.org/10.1016/j.neuroimage. 2005.09.023. Cao, X., Li, S., Zhao, J., Lin, S., Weng, X., 2011. Left-lateralized early neurophysiological response for Chinese characters in young primary school children. Neurosci. Lett. 492 (3), 165–169. https://doi.org/10.1016/j.neulet.2011.02.002. Chaumon, M., Bishop, D.V.M., Busch, N.A., 2015. A practical guide to the selection of independent components of the electroencephalogram for artifact correction. J. Neurosci. Methods 250, 47–63. https://doi.org/10.1016/j.jneumeth.2015.02.025. Chauncey, K., Holcomb, P.J., Grainger, J., 2008. Effects of stimulus font and size on masked repetition priming: an event-related potentials (ERP) investigation. Lang. Cogn. Process. 23 (1), 183–200. https://doi.org/10.1080/01690960701579839. Chen, H.C., Shu, H., 2001. Lexical activation during the recognition of Chinese characters: evidence against early phonological activation. Psychon. Bull. Rev. 8 (3), 511–518. Retrieved from. http://www.ncbi.nlm.nih.gov/pubmed/11700902. Cohen, L., Dehaene, S., 2004. Specialization within the ventral stream: the case for the visual word form area. Neuroimage 22 (1), 466–476. https://doi.org/10.1016/j. neuroimage.2003.12.049. Dehaene, S., Cohen, L., 2011. The unique role of the visual word form area in reading. Trends Cogn. Sci. 15 (6), 254–262. https://doi.org/10.1016/J.TICS.2011.04.003. Dehaene, S., Cohen, L., Sigman, M., Vinckier, F., 2005. The neural code for written words: a proposal. Trends Cogn. Sci. 9 (7), 335–341. https://doi.org/10.1016/j.tics.2005.05. 004. Delorme, A., Makeig, S., 2004. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J. Neurosci. Methods 134 (1), 9–21. https://doi.org/10.1016/j.jneumeth.2003.10.009. Fraga González, G., Žarić, G., Tijms, J., Bonte, M., Blomert, L., van der Molen, M.W., 2014. Brain-potential analysis of visual word recognition in dyslexics and typically reading children. Front. Hum. Neurosci. 8, 474. https://doi.org/10.3389/fnhum. 2014.00474. Fraga González, G., Žarić, G., Tijms, J., Bonte, M., Blomert, L., Leppänen, P., van der Molen, M.W., 2016. Responsivity to dyslexia training indexed by the N170 amplitude of the brain potential elicited by word reading. Brain Cogn. 106, 42–54. https://doi. org/10.1016/j.bandc.2016.05.001. Koenig, T., Kottlow, M., Stein, M., Melie-García, L., 2011. Ragu: a free tool for the analysis of EEG and MEG event-related scalp field data using global randomization statistics. Comput. Intell. Neurosci. 1–14. 2011. https://doi.org/10.1155/2011/938925. Kronbichler, M., Hutzler, F., Wimmer, H., Mair, A., Staffen, W., Ladurner, G., 2004. The visual word form area and the frequency with which words are encountered: evidence from a parametric fMRI study. Neuroimage 21 (3), 946–953. https://doi.org/ 10.1016/j.neuroimage.2003.10.021. Lehmann, D., Skrandies, W., 1980. Reference-free identification of components of checkerboard-evoked multichannel potential fields. Electroencephalogr. Clin. Neurophysiol. 48 (6), 609–621. https://doi.org/10.1016/0013-4694(80)90419-8. Lin, S.E., Chen, H.C., Zhao, J., Li, S., He, S., Weng, X.C., 2011. Left-lateralized N170 response to unpronounceable pseudo but not false Chinese characters-the key role of orthography. Neuroscience 190, 200–206. https://doi.org/10.1016/j.neuroscience. 2011.05.071. Marshall, J.C., Newcombe, F., 1973. Patterns of paralexia: a psycholinguistic approach. J. Psycholinguist. Res. 2 (3), 175–199. https://doi.org/10.1007/BF01067101. Maurer, U., Brem, S., Bucher, K., Brandeis, D., 2005. Emerging neurophysiological specialization for letter strings. J. Cogn. Neurosci. 17 (10), 1532–1552. https://doi.org/ 10.1162/089892905774597218. Maurer, U., Brem, S., Kranz, F., Bucher, K., Benz, R., Halder, P., et al., 2006. Coarse neural tuning for print peaks when children learn to read. Neuroimage 33 (2), 749–758.
8