Language and Linguistics - Volume 22, Issue 4, 2021

This study of Chinese adverbial ziji investigates why cross-linguistically adverbial intensifiers often develop two different uses, namely the exclusive use and the inclusive use. Arguing against the polysemous account proposed in previous works like Siemund (2000), and assuming the mechanism suggested in Liao (2018) for exclusive ziji, the paper presents a new analysis revised from Gast’s (2006) account for intensifiers. In the analysis, there is only one ziji for all its adverbial uses. By adjoining to different X’ positions in the structure, adverbial ziji may get different surface meanings. Despite the surface differences, adverbial ziji always has the following semantics: it works as an identity function, evokes alternatives for consideration, and receives an exclusive meaning after the application of the covert exhaustivity operator O. Based on the evidence presented, the analysis crucially assumes that adverbial ziji may adjoin to Topic’, and this adjunction leads to the effect that the subsequent exhaustification is done over a set of alternative propositions that vary in topics. In such a case, alternative individuals evoked by ziji do not have to be excluded from having the property described by the VP in question. This makes the assertion of a ziji-sentence in inclusive context possible, and accounts for why intensifier ziji has a disguised inclusive function. By proposing such a unified account of ziji, the paper explains why cross-linguistically intensifiers often develop the various uses observed.

Chinese Quadrisyllabic Idiomatic Expressions (henceforth QIEs) are highly productive in the modern language. They can be used to understand the cognitive processing of structure and meaning during reading comprehension, as in the patterning of [qian-A-wan-B] ‘1k-A-10k-B’ (e.g. one-thousand army ten-thousand horse). However, little is known about the underlying mechanisms of QIEs during reading comprehension. Adopting the framework of Construction Grammar, in the present study, we aimed to study the convergence and divergence between native speakers and L2 learners in the processing of Chinese idiomatic constructions.

In the present study, twenty-three native university-level Mandarin speakers and twenty-three L2 learners of intermediate and advanced levels of Mandarin, all speakers of the non Sinosphere, participated in the experiment, and were instructed to make a semantic congruency judgment during the presentation of a QIE. Our results showed that, for both native speakers and L2 learners, semantically transparent idiomatic constructions elicited much shorter RTs than semantically opaque idiomatic constructions. Our behavioral results also showed that native speakers processed low frequency QIEs faster than high frequency ones, implying semantic satiation to impede the interpretation of high frequency idioms. For L2 learners, it was semantic transparency, rather than frequency, that played a more prominent role in idiom processing.

Choswateng Tibetan, spoken in the south-eastern corner of the Khams region, has three negative prefixes: /ȵi-/, /ma-/, and /ka-/. The first two are derived from two morphemes which are ubiquitous across Tibetic languages, whereas the third is a newly generated negative prefix found in Choswateng Tibetan as well as its surrounding dialects belonging to the rGyalthang subgroup of Khams and its neighbours. This article describes the morphological feature and use of the prefix /ka-/ in Choswateng Tibetan. Morphologically, the prefix /ka-/ can co-occur with most verbs except for the copulative verb /ˊreʔ/. Pragmatically, the prefix /ka-/ occurs and is restricted in the following ways: (1) expresses ‘definitely not’ for statements regarding the self, and ‘possibly not, judging from the speaker’s knowledge’ for statements regarding others; (2) co-occurs with egophoric and sensory evidentials; (3) is not used for a negation of accomplished aspect; and (4) does not deprive the function of the other two negative prefixes. These two analyzes are mutually related; it is suggested that the reason why /ka-/ cannot co-occur with the copulative verb /ˊreʔ/ is triggered by a contradiction of implied evidentials: /ka-/ is related to egophoric and sensory, whereas /ˊreʔ/ is statemental. Following the description of its use, we discuss the origin of /ka-/, claiming a possible grammaticalization from an interrogative word gar (‘where’ in Literary Tibetan and common throughout the rGyalthang area) in a rhetorical question to a prefix. Referring to several morphological features of /ka-/, we consider its grammaticalization as ongoing, but most advanced in Choswateng Tibetan.

The Mandarin renhe is similar to the English any in terms of polarity sensitivity (Wang 1993; Wang & Hsieh 1996; Kuo 2003; Cheng & Giannakidou 2013; Shyu 2016). However, the following phenomena regarding any in relative clause environments have not been surveyed with respect to renhe: (a) the NPI illusion effect reported in studies like Parker & Phillips (2011; 2016); (b) the subtrigging effect discussed in LeGrand (1975) and Dayal (1998; 2004). We conducted two untimed, offline acceptability judgment experiments and the results suggest that: (i) NPI illusion does not appear in Mandarin in untimed offline processing, (ii) the subtrigging effect of renhe holds, and (iii) renhe can be licensed by certain types of declarative verbs like tongyi ‘agree’ and zancheng ‘approve’. The results confirm the strict structural requirement of the c-commanding relation between a negation licensor and renhe (Wang 1993) and the licensing of renhe in non-veridical contexts (Cheng & Giannakidou 2013), and further suggest additional licensing environments for renhe: relative clauses and declarative verbs. This requires reconsideration of positing non-veridicality as a necessary licensing condition for renhe and calls for future research on how renhe is licensed under these two licensing environments.

Korean causal markers ‑ese, ‑nikka and Mandarin Chinese yinwei are used most widely in their respective languages to denote causality. Due to the syntactic structure of the adjunct-preceding-main order in both languages, ‑ese and ‑nikka are considered to be clause-initial connectives. However, in both languages, because-clauses also occur frequently in final position. By examining the behavioral patterns of ‑ese and ‑nikka in clause-initial and -final positions in relation to the sequence of yinwei through the original Korean texts translated to Mandarin Chinese and the Korean texts translated from the original Mandarin Chinese texts, this paper aims to show: (1) ese-tokens correspond more to the preposed yinwei and other Mandarin Chinese result causal markers that strengthen the “cause-consequence” construction, as ‑ese mainly expresses forward logical reasoning, whereas nikka-tokens correspond more to the postposed yinwei, as ‑nikka expresses epistemic inference and elaboration; (2) these logical relations are more explicitly expressed by ‑ese and ‑nikka, whereas Mandarin Chinese relies more on zero-marking when expressing a strong sense of sequential logic or an utterance of elaboration.