Each of them is one of a kind: A corpus-based study on two type-noun morphemes in spoken Mandarin

This article presents a corpus-based analysis of how two near-synonymous type nouns (TNs) in Mandarin, zhǒng and lèi , diverge in their distributional patterns and interactional functions in spoken discourse. Using quantitative collocational profiling and qualitative analysis informed by Interactional Linguistics, the study examines 968 instances of zhǒng and 179 instances of lèi in the NCCU Corpus of Spoken Mandarin. The findings show that zhǒng forms a broad set of prefabricated expressions, each favoring particular lexico-grammatical constructions and serving evaluative, referential, and turn-projecting functions. In contrast, lèi , most prominently in the form zhīlèi , is more restricted in distribution, occurs predominantly in utterance-final position, and indexes uncertainty and turn completion. These results demonstrate that even near-synonymous TN morphemes differentiate in systematic ways shaped by linguistic form, sequential context, and interactional pressures. The study contributes to research on TNs, classifier systems, and pragmatic markers in Mandarin, while offering implications for cross-linguistic comparison and Chinese language pedagogy.