Abstract
<jats:title>Abstract</jats:title> <jats:p>This chapter provides an overview of three key properties of classifiers in Asian languages: (1) semantic and syntactic congruency in classifier–noun phrases, (2) the degree of specificity, and (3) the count/mass distinction. It then reviews empirical studies on the real-time processing of classifiers and their use as tools to investigate human sentence processing. The chapter synthesizes five strands of research on classifier processing: predictive processing, the nature of processing classifier–noun combinations, classifier animacy and count–mass distinctions, relative clauses headed by classifiers, and second language sentence processing. Existing research has shown that classifiers, despite being single words, can generate predictions (Qian and Garnsey 2016), rapidly adjust those predictions in response to new information (Chow and Chen 2020), and pre-activate the semantics of predicted words (Kwon et al. 2017). Eye-tracking studies on L2 sentence processing demonstrate that both native and L2 speakers can use classifiers predictively (Lau and Grüter 2015; Mitsugi 2020). Native speakers tend to prioritize grammatical form cues to eliminate incompatible nouns, whereas L2 learners rely more on semantic congruity (Grüter et al. 2020). Regarding classifier–noun combination processing, most studies report N400 effects in response to mismatches, suggesting that processing is primarily semantic when meanings clash (Chou et al. 2014; Huang and Schiller 2021; Wang et al. 2019). A few studies report P600 or anterior negativities when semantic conflicts are removed by omitting classifiers or substituting the semantically empty classifier gè, indicating that morphosyntactic processes are also involved (Chan 2019). Overall, the processing of classifier–noun agreement differs from morphosyntactic number or gender agreement in Indo-European languages.</jats:p>