Lecture | CHiLL series
From Motion to Future to Speech Act: The Functional Elevation of luai-khə ‘come-go’ in Ji’an Gan Chinese
- Meixiang Chang / Zhaole Yang
- Date
- Wednesday 10 September 2025
- Time
- Serie
- Chinese Linguistics in Leiden (ChiLL)
- Address
-
Lipsius
Cleveringaplaats 1
2311 BD Leiden - Room
- 1.31
Abstract
This study contributes to the ongoing discussion of the syntacticization of discourse, often referred to as the “Tree Tops” (Miyagawa 2022), by examining the lexicalized motion verb combination luai-khə (来去, lit. ‘come-go’) in Ji’an Gan Chinese, a Sinitic variety. We present evidence for its grammaticalization and demonstrate three distinct layers of meaning: (i) a bleached motion sense, (ii) a bouletic modal sense expressing intention or planning, and (iii) a future temporal sense. Ultimately, we show that sentences with luai-khə most often encode two types of speech acts: commissives and directives. Strikingly, the multifunctionality of luai-khə appears to align with a broader crosslinguistic grammaticalization pathway for motion verbs.
A central finding is a form–function mismatch: although luai-khə is realized in a low structural position at the edge of vP, it encodes interpretive content typically associated with the highest clausal domains, namely the speech act structure. More specifically, luai-khə simultaneously activates both discourse participants. In hearer-oriented directives, it highlights the speaker’s role and shifts the second-person subject toward an inclusive reading. In speaker-oriented commissives, it foregrounds the hearer’s role, pragmatically functioning as an attention-getter or as a marker of negotiation. This functional elevation is further supported by distributional parallels with the Japanese politeness marker -mas, which is also restricted to root environments (Emonds 1970) and has been analyzed by Miyagawa as evidence for a syntactic projection dedicated to speech acts. To account for this discrepancy between structural position and interpretive scope, and following Miyagawa’s analysis of -mas, we develop a feature-based movement analysis. We propose that luai-khə bears a complex bundle of two uninterpretable person-agreement features: one targeting the speaker and one targeting the hearer. It undergoes successive movement from its base position to S(peaker)A(ddressee)P, the locus where the roles of speaker and hearer are syntactically encoded.