Recent research has highlighted the value of investigating the online language processing of heritage speakers
(HS) as a means of accessing their implicit language knowledge (Montrul, 2023). While
studies have shown that Spanish and Polish HS use grammatical gender cues to predict upcoming nouns (Fuchs, 2022a, 2022b), less is known about Vietnamese HS’ use
of prenominal classifiers (e.g., con for animate objects, cái for inanimate objects) to
facilitate their processing (but see Ito et al., 2024). This study examined if and how
home-country raised and heritage speakers of Vietnamese in the U.S. use classifiers to facilitate the processing of upcoming
nouns, and whether heritage language proficiency is a modulating factor. Forty-one adult native speakers of Vietnamese (18
home-country raised, 23 HS) completed a visual-world eye-tracking experiment, an offline cloze test to assess knowledge of
classifier–noun pairings, and a Vietnamese listening proficiency test. The results indicate that despite more variable knowledge
of classifier–noun pairings and generally slower lexical access, HS use classifiers as a semantically informative cue during
real-time comprehension, albeit to a somewhat lesser extent than their home-country raised peers. Increased proficiency in the
heritage language, whether measured objectively or self-rated, was not found to enhance engagement in prediction.