The processing of multiword expressions in children and adults: An eye-tracking study of Chinese

Abstract

The processing advantage for multiword expressions over novel language has long been attested in the literature. However, the evidence pertains almost exclusively to multiword expression processing in adults. Whether or not other populations are sensitive to phrase frequency effects is largely unknown. Here, we sought to address this gap by recording the eye movements of third and fourth graders, as well as adults (first-language Mandarin) as they read phrases varying in frequency embedded in sentence context. We were interested in how phrase frequency, operationalized as phrase type (collocation vs. control) or (continuous) phrase frequency, and age might influence participants’ reading. Adults read collocations and higher frequency phrases consistently faster than control and lower frequency phrases, respectively. Critically, fourth, but not third, graders read collocations and higher frequency phrases faster than control and lower frequency sequences, respectively, although this effect was largely confined to a late measure. Our results reaffirm phrase frequency effects in adults and point to emerging phrase frequency effects in primary school children. The use of eye tracking has further allowed us to tap into early versus late stages of phrasal processing, to explore different areas of interest, and to probe possible differences between phrase frequency conceptualized as a dichotomy versus a continuum.

How Does Modification Affect the Processing of Formulaic Language? Evidence from L1 and L2 Speakers of Chinese

“…Although children showed a similar downward trend of FL advantage across original and modified forms as adults did, the magnitude of the processing advantage was found to be larger in adults than in children. This finding is in line with existing research (e.g., Hallin & Van Lancker, 2017;Jiang et al, 2020). Compared with L1 adults, children had relatively limited exposure to target FSs, which decreased their sensitivity to FL in its original and modified forms.…”

“…For L1 children, as their age increased, the magnitude of the processing advantage increased. This finding is not surprising (e.g., Bannard & Matthews, 2008;Jiang et al, 2020). However, this was not the case in L2 adults.…”

How Does Modification Affect the Processing of Formulaic Language? Evidence from L1 and L2 Speakers of Chinese

“…Besides, although studies on formulaic language processing have proliferated recently, most of them have focused on adults, with children being largely disregarded. Given that formulaic language plays an important role in first language (L1) acquisition and development (e.g., Clark, 1974;Lieven, Behrens, Speares, & Tomasello, 2003;Nelson, 1973;Peters, 1977), a better understanding of how it is processed by children is urgently needed (e.g., Jiang, Jiang, & Siyanova-Chanturia, 2020;Wray, 2002).…”

“…Other researchers examined phrase frequency both as a categorical and a continuous variable in the data analysis (e.g., Arnon & Snider, 2010;Jiang, Jiang, & Siyanova-Chanturia, 2020;Northbrook & Conklin, 2019;Shantz, 2017;Siyanova-Chanturia et al, 2011b;Siyanova-Chanturia & Janssen, 2018;Wolter & Gyllstad, 2013). Some studies found the effect of phrase frequency was significant when it was treated as a continuous, not categorical, variable (e.g., Arnon & Snider, 2010;Shantz, 2017;Siyanova-Chanturia & Janssen, 2018).…”

“…Some studies found the effect of phrase frequency was significant when it was treated as a continuous, not categorical, variable (e.g., Arnon & Snider, 2010;Shantz, 2017;Siyanova-Chanturia & Janssen, 2018). Others found the phrase frequency effect was significant both as a categorical and a continuous variable (e.g., Jiang et al, 2020;Northbrook & Conklin, 2019;Siyanova-Chanturia et al, 2011b;Wolter & Gyllstad, 2013); however, the magnitude of the continuous measure was always found to be greater than the categorical measure (e.g., Jiang et al, 2020;Northbrook & Conklin, 2019;Siyanova-Chanturia et al, 2011b;Wolter & Gyllstad, 2013). This finding suggests that, compared with phrase frequency as a categorical predictor, the continuous phrase frequency might be a more precise and sensitive measure for FL processing (see Jiang et al, 2020 for more details).…”