Cross-script phonological activation in Chinese–English bilinguals: The effect of SOA from masked priming.

Abstract

The issue of bilingual phonological access remains unclear for bilinguals with cross-script language systems, which is especially true when the time course of phonological activation is involved. To investigate the time course of cross-script phonological activation, the present study asked Chinese-English bilinguals to complete a word naming task that was conducted in a forward-masked phonological priming paradigm in three stimulus onset asynchrony (SOA) conditions. By comparing the interlingual and intralingual phonological priming effects in a within-subjects design, we found that (a) target naming in Chinese and English was facilitated by a phonologically similar English or Chinese prime in the three SOA conditions (43 ms, 75 ms, and 150 ms) and the facilitation effect of the prime reached the peak when the pronunciation of the prime-target pair most resembled each other and (b) manipulation of the SOAs affected both the naming latencies of target words and the sizes of the phonological priming effect. In particular, naming latencies in each prime-target type displayed an increasing tendency as the SOA prolonged. Moreover, despite the varied sizes of the priming effect in the three SOA conditions, we found a consistent pattern that the priming effects in two interlingual conditions resembled their respective intralingual conditions along the time course. Taken together, these findings provide strong support for an integrated phonological representation of bilinguals and further extend the language nonselective access hypothesis to language pairs with very different orthographic systems. Implications for the manipulation of the SOAs in the masked priming paradigm are also discussed. Public Significance StatementThe present study found that even for Chinese-English cross-script bilinguals (with Chinese quite opaque in pronunciation), one language could prime another in visual word naming within a very short time (especially when primed by English). This shows that for people learning another language, their different languages are connected and integrated in the mind.

Automatic phonological access among bilinguals with cross-script languages

“…In this figure, the Chinese prime activates the phonological units shared with the target English words at the sublexical level, and this facilitates the recognition of the target through feedback from these sublexical phonological units to the lexical phonological units, and then to orthographic representations at both sublexical and lexical levels (Jiang, 2023). However, inconsistent with previous studies (e.g., Xu et al, 2021;Zhou et al, 2010), we failed to find a significant priming effect from L2 to L1 in both LDT and word naming tasks. A possible interpretation of the asymmetric priming effect could be the different roles of phonological information in reading Chinese and English.…”

“…A growing line of research examined the interlingual homophone priming effect in languages using different scripts, and supported the automatic cross-language phonological activation independent of script similarity (e.g., Choi et al, 2010;Dimitropoulou et al, 2011;Kim & Davis, 2003;Lim & Christianson, 2023;Voga & Grainger, 2007;Xu et al, 2021;Zhou et al, 2010). Most of these studies focused on two alphabetic writing systems with different scripts.…”

“…A more recent study, however, did not find a bidirectional cross-script activation. Xu et al (2021) used the same masks as Zhou et al (2010) and varied the stimulus onset asynchrony (SOA) length at 43, 75, and 150 ms (the SOA in Zhou et al (2010) was 85 ms). In a masked priming word naming task, the English prime significantly accelerated Chinese homophone recognition when the SOA was 43 or 150 ms, but the Chinese prime did not significantly facilitate its English homophone target recognition in any of the SOA conditions.…”

“…For instance, few of them reported the visibility of the primes (see also Lee et al, 2018, for discussion). Zhou et al (2010) and Xu et al (2021) used a four-field mask procedure (Forster et al, 2003), with a forward mask #### for 500 ms, the prime for 50 ms, and a backward mask “&&&&” for 35 ms displayed before the target. Some researchers suggested that a pattern mask such as a string of ampersands (&&&&) or hashtags (####) may not be sufficient to reduce the visibility of Chinese characters (Jiang, 1999; Wang, 2013; Wang & Forster, 2010).…”

“…Previous studies also paid attention to phonological activation among bilingual speakers. A great amount of evidence suggests that cross-language phonological activation occurs during bilingual visual word recognition, involving various orthographic pairs, such as English and French (e.g., Haigh & Jared, 2007; Jared & Kroll, 2001), Dutch and English (e.g., Dijkstra et al, 1999), Dutch and French (e.g., Brysbaert et al, 1999; Duyck et al, 2004), Greek and Spanish (e.g., Dimitropoulou et al, 2011), Chinese and English (Xu et al, 2021; Zhou et al, 2010), Japanese and English (Ando et al, 2014a, 2015; Nakayama et al, 2012), and Korean and English (Kim & Davis, 2003; Lee et al, 2005; Lim & Christianson, 2023).…”