摘要 本文以ABB、A不BB、A里AB、A不CD等汉语形容词重叠格式为对象，提出“感叹语气观”以揭示其核心功能。研究发现，此类重叠格式多用作谓语，常带语气词“的”，主要表达说话人的主观情绪态度，尤以消极情感为主。文中指出，中缀“不”“里”等非单纯句法成分，而是具有感叹功能的语气语素。基于普通话及多个南北方言的语料，本文构建“重叠构形×语气功能×方言分布”的三维分析框架，分析重叠格式的语义极性、构式变异及地域分布，并进一步发现ABA类结构中语气词“呔”“吧”等的广泛使用。研究显示，这些格式不仅构成富有情感表达力的语气化构式，也具备“雪克隆”式的构式扩展机制，对形容词重叠式的功能理解与方言类型学研究具有重要启发意义。
摘 要 約量詞常用來表達不確定的數量，不同約量詞在實際使用中呈現出語義與構式自然度的差異。本文以漢語「來」與「多」為研究對象，探究臺灣地區的母語者如何判斷「數詞X＋來」所表示的數值範圍（即取值區間），檢視「多」置於位數詞之前（如「三百多萬人」，位數為「萬」）或之後（如「三百萬多人」）的語感自然度，進而兼論量詞對此句式自然度的影響。本研究發現，「來」表示的數值範圍隨前接數詞而不同，如「十來」多被理解為11至19，顯示「來」的取值區間受到數詞限制，本研究稱之為框定效應。另，當數詞增大，作答數值趨於分散，並偏好使用10、100等尾數為零的圓整數作答。關於「多」的使用，當「多」置於位數詞之前時（「三百多萬人」），自然度評分顯著高過其於位數詞之後，且後者的自然度會因前接係數單位詞（「三百萬多」中的「百」）的縮小而提升，亦會受到有無量詞嵌入的影響。
本文以“直接”和“反手”二词为例，探讨了网络环境中副词由表方式向表语气演变的现象。在自媒体和社交媒体话语中，方式副词如“直接”和“反手”经常用于超常性事件的因果链叙事，凸显言者对前后事件关系的主观认识继而表达言者态度和立场。我们认为这类副词虚化的直接机制是主观化——主观认识中事件和方式的超常性成为说话人传递情态立场的落点，同时伴随交互主观化——不仅表达说话人态度，还隐含对受众反应的预期。本文也在微观层面勾画了数字时代算法推动语言用于强化、共建情感和立场从而导致语言形式的功能演变。
This study examines lai and duo, focusing on how native speakers from Taiwan interpret the value range of "numeral + lai," and the naturalness of "duo X" (e.g., sanbai duo wan) versus "X duo" (e.g., sanbai wan duo), with X representing the base numeral unit. Results show the lai's value range varies with numerals (e.g., shi lai as 11-19), indicating a "delimit effect." Larger numerals yield dispersed and round-number responses, such as 10. "Duo X" is rated more natural, while the naturalness of "X duo" improves when the coefficient unit preceding the base (e.g., bai in sanbai wan duo) is smaller.
This study focuses on Chinese adjective reduplication patterns such as ABB, A-bu-BB, A-li-AB, and A-bu-CD, and proposes an "exclamatory modal perspective" to reveal their core function. These patterns are typically used as predicates, often followed by the modal particle de, and primarily convey the speaker's subjective emotional stance, particularly negative affect. The paper argues that medial elements like bu ("not") and li are not merely syntactic fillers but function as exclamatory modal particles. Drawing on data from both Mandarin and a range of northern and southern dialects, the study constructs a three-dimensional analytical framework-"reduplication structure & times;modal function & times;dialectal distribution"-to examine the semantic polarity, formal variation, and areal features of these patterns. It further identifies the productive use of modal particles such as tai and ba in the ABA pattern. The findings demonstrate that these forms not only serve as emotion-laden modal constructions but also exhibit snowclone-like expandability, offering new insights into the functional interpretation of adjective reduplication and its relevance to dialectal typology.
Chinese in the Malay Archipelago has evolved into distinct regional written varieties due to adaptation to local languages and cultures, requiring exploration of their lexical and syntactic differences from standard written Chinese. In this paper, we conduct a quantitative analysis on Chinese news texts collected from five Malay Archipelago regions, trying to figure out their differences with the texts in standard written Chinese from a lexical and syntactic perspective. The statistical results show that the regional written Chinese varieties used in these five regions are quite different, diverging from their modern Chinese mainland counterpart. All these discrepancies reflect how Chinese evolves overseas, and demonstrate the profound impact of local societies and cultures on the development of Chinese.
Drawing on findings from ten interviews conducted in 2021, this study examines how members of the British Chinese community in London accessed and interpreted COVID-19 pandemic information, the barriers they encountered in institutional health communication, and the shifting meanings of health artefacts. The findings reveal that transnational information flows created both protective advantages and confusion, that communicative inequalities were frequent during the pandemic, and that the semiotic meanings of key health artefacts became politicised and racialised. The analysis makes use of what we propose as an integrative discourse analysis, which examines data using concepts from several theoretical frameworks.
基于大规模语料库 (共3.2万句，86万词规模)，本文系统比较了中韩网络仇恨言论在性别、种族/国籍、地域三个维度的语言特征。同时，不仅描述语言现象，更深入分析其作为社会行为的语言策略与功能，揭示了两国仇恨言论背后的社会文化差异及动因。研究发现，韩国的性别仇恨言论更直接，包含针对身体特征的表演性暴力、基于派生构词法的语域化(Enregisterment)，用以激化性别对立；中国则更依赖高阶指示性(Higher-order indexicality)，通过隐喻或谐音间接构建性别议题。种族相关表达方面，两国的立场表达(Stance-taking)策略有所不同。韩国主要针对国内移民和外籍群体，反映内部社会紧张；中国则关注国外情景，强调集体认同。地域相关表达方面，韩国侧重于地方对立和首都中心主义；中国则受经济差距、民族多样性和社会认知差异的影响，呈现更复杂的表达模式。结论指出，网络仇恨言论并非单一现象，而是由特定语言意识形态与媒介环境共同塑造的，是反应文化特殊性的复杂言语行为。本文结合基于统计检验的关键词分析及其上下文(KWIC)考察，从定量和定性视角探讨中韩网络负面言论的语言特征及表达机制，为跨文化仇恨言论研究提供重要参考。
As a linking element, zongzhi 'in sum, to sum up' displays archaic qualities and is most commonly associated with formal written Chinese discourse. However, this paper demonstrates that zongzhi can also be deployed in dialogic interactions in contemporary spoken Chinese to accomplish distinct social actions. Drawing on a variety of dialogic data, we identify a major distinction between tokens produced by the primary speaker and those produced by the respondent. Tokens from the primary speaker fall into three subcategories, corresponding to a person-based grouping: (1) first-person tokens are used to display the speaker's affective stance toward the stated situation; (2) second-person tokens can address the interlocutor's concerns; and (3) third-person tokens function primarily for informing and/or explaining. Although zongzhi-prefaced turns by respondents are relatively rare, they are noteworthy for their role in transforming the participation framework with distinct interactional functions. Additionally, zongzhi tokens are often associated with prosodic prominence and occur at points of conversation sequential transition. We offer a unified account of these diverse uses by linking them to some of the requirements of sequence closings in conversation. We conclude by discussing the broader implications of these findings for understanding lexical expressions of summarization, totality, and/or inclusivity. We emphasize that viewing grammar and social interaction as interdependent provides a productive framework for analyzing lexical tokens cutting across both written and spoken genres, as well as their variation within speech registers.
Amid increasing global environmental concerns, Environmental, Social, and Governance (ESG) reporting has become standard practice for listed companies; few studies, however, analyzed Chinese ESG reports from an ecolinguistic perspective. This study employs quantitative corpus-assisted ecolinguistic methods to systematically compare vocabulary choices and discourse patterns in a corpus of 136 sustainability reports, further divided into two subcorpora - those issued by leading Chinese (public) companies by revenue (List A) and those aligned with Sustainable Development Goals criteria (List B). Results from combining keyword analysis and thematic clustering reveal diverging sustainability and political discourses across the two subcorpora, suggesting that commitment to SDGs and ownership influence language use. These variations center around two main poles, one driven by economy and State ideology (List A) and the other by a techno-oriented perspective (List B), reflecting the diverse approaches currently present in China's sustainability arena.
Based on a large-scale corpus of over 32,000 sentences (860,000 tokens), this study compares the linguistic features of online hate speech in China and South Korea across the dimensions of gender, race/nationality, and region. In addition to describing linguistic phenomena, this study investigates the linguistic strategies and social functions of hate speech, highlighting the cultural and societal factors that shape its expression and difference between both countries. The findings indicate that gender-related hate speech in Korea is more direct, involving performative violence targeting physical characteristics and enregisterment through derivational morphology to intensify gender antagonism. In contrast, its Chinese counterpart relies more on higher-order indexicality, indirectly constructing gender issues through metaphors and homophones. Regarding race-related expressions, the stance-taking strategies differ. Korean hate speech primarily targets domestic migrants and foreign groups, reflecting internal social tensions, whereas Chinese discourse focuses on situations abroad to reinforce collective identity. In terms of regional expressions, Korean hate speech highlights interregional hostility and capital-centrism. Chinese expressions, however, exhibit more complex patterns influenced by economic disparities, ethnic tensions, and ideological differences in social cognition. The conclusion asserts that online hate speech is not an isolated phenomenon, but rather a culturally embedded discourse practice shaped by specific language ideologies and media environments. Combining a statistical keyword analysis with Keyword-in-Context (KWIC) methods, this study investigates the linguistic features and expressive patterns of hate speech in China and Korea from both quantitative and qualitative perspectives, offering a key reference for cross-cultural research on hate speech.
This study examines how teachers employ the Chinese manner demonstrative, zheyang(zi) 'like this; this way,' to direct students' attention toward their embodied actions during instructional demonstrations. Data consists of 84 cases from 12.5 hours of videotaped interaction in a cooking class in Taiwan. Drawing on interactional linguistics, the study explicates the distinct interactional imports of zheyang(zi) when used (a) before, (b) after, (c) concurrently with, or (d) during a continuous demonstrated action, analyzing its linguistic structure, multimodal configuration, and sequential positioning. The multimodal analysis reveals how the design of zheyang(zi)-marking instruction reflects teachers' orientation toward the progressivity and materiality of a demonstrative activity and how gaze shows participants' understanding of zheyang(zi). By examining the interplay between grammar, embodiment, and material transformations, the study contributes to our understanding of the reflexive relationship between grammar choices and activity context.
This paper examines the pragmatic functions of the pragmatic marker (pm, cf. Fraser 2009) jiushi in Chinese and further categorizes its functions as a discourse marker (dm) indicating elaboration and contrast, and a discourse management marker (dmm) signaling topic establishment, hesitation and self-repair. Previous research (cf. Zhang 2002) suggests that the development of jiushi involves a progression of meaning changes that were first subjectified, then textualized and eventually inter-subjectified. This paper, however, taking a constructional point of view argues that when jiushi is constructionalized into a conjunct, it automatically inherits from the pre-existing conjunct schema the subjective, textual, and intersubjective properties. (Inter)ubjective properties are increased when jiushi is further constructionalized into a pm. Such an analysis argues that the textual, subjective, and intersubjective qualities are independent, and each may in principle shift (or not) at any moment/order and characterizes the development of procedural structures without placing the diachronic concepts on a cline.
The present study aims to examine which cognitive processing model, namely the Default Model, Context-driven Model, or Standardization Model, offers the most effective explanation for how native Mandarin speakers process scalar implicatures. The experiment in the current study employs a forced-choice task to investigate whether native Mandarin speakers interpret scalar terms literally or pragmatically by having participants complete dialogues containing these terms, which were then analyzed for their scalar implicature interpretation rates and response times. The experimental materials contain gradable adjective pairs re/tang 'hot/scalding' and hao/youxiu 'good/excellent' as well as closed-class pairs youxie/suoyou 'some/all' and huo/he 'or/and' placed in three distinct types of contexts, specifically upper-bounded context, lower-bounded context and neutral context. The preliminary results show that the processing of both gradable adjective pairs is best explained by the Standardization Model while that of the two closed-class pairs best fits the predictions of the Default Model.
Adopting the framework of interactional linguistics and employing the methodology of conversation analysis, this study investigates the discourse-interactional functions of the causal conjunction yinwei (“because”) in naturally occurring Mandarin conversation. Moving beyond its conventional use as a marker of reason or cause, the analysis reveals two prominent interactional functions of yinwei in everyday talk: (1) continuing a pre-prior course of action to advance the speaker’s interactional agenda, and (2) reopening a previously completed telling to pursue affiliative or responsive uptake from the hearer. These findings demonstrate that yinwei functions not only as a semantic connector but also as a resource for managing conversational structure and speaker-hearer alignment. By showing how speakers strategically exploit yinwei to organize discourse and negotiate participation, this study contributes to a more nuanced and interactionally grounded understanding of causal conjunctions in Mandarin and highlights the importance of sequential positioning in shaping their discourse functions.
