本研究以自然发生的口语语料——电视剧台词转写材料为基础，以有声口语语料库和文献语料作为补充，在话语分析和语义地图理论的指导下，考察了广州话“ho”、台湾国语“hoNh”、北京话“ha”、东北方言“ha”和美国英语“huh”等五种语言/方言中的ha类交互主观性标记以及与之相关的跨语言现象。

第一章为绪论，交代了本文的研究对象和拟解决的问题，梳理评述了交互主观性和ha类标记的相关研究，并对本文所用的理论方法和语料来源进行了介绍。

第二章通过观察ha类标记所处的具体语境，提取了小句分布、话轮分布、序列分布等形式上的依据，以及宿主成分性质、言语行为和言者意图、言听双方信息掌握的权威性等语义语用依据作为跨语言功能分类的判别标准，概括出包括依附性和非依附性两大类的18个语义基元：事实确认、立场确认、知晓确认、核实确认、信息接收、引证说明、追问促应、疑惑困惑、立场对抗、呼叫回应、惊奇惊讶、醒悟认同、反馈信号、征询同意、建议要求、寒暄客套、句外叙述、句内叙述。

第三章将归纳的18个语义基元作为功能节点，逐步构建了ha类标记的概念空间。首先，遵循连续性要求和最小关联原则，计算比较了节点间的关联度，构建出粗颗粒的概念空间。然后，借助语义分析手段提高了概念空间的解析度，得到更为清晰的细颗粒版本，并在细颗粒底图上分割出了五种模式的ha类标记语义地图。

第四章分析了跨语言中ha类标记在功能、韵律、分布上的共性与个性，并对相关现象进行了解释。具体讨论了ha类标记概念空间和语义地图所反映的功能共性与差异，考察了ha前停顿与其辖域和语义指向的关系问题，还对汉语方言中ha类标记与其他语气词共现时的后置性作出了相应的解释，并借助立场三角探讨了交互主观性的分类问题。最后，从ha类标记在不同语言中的词类范畴归属和句法地位出发，提出语气词、叹词和附加问两两之间存在着功能共通性，或可将其视为一个功能连续统。

第五章对研究内容和主要结论进行了总结，同时指出本研究所存在的不足，并对今后的研究方向提出了展望。

The current study obtains naturally-occurring data from transcriptions of television series, as well as supplementary data of spoken language corpus and previous studies. Under the framework of discourse analysis and semantic map model, we investigate ha-type intersubjective markers and cross-linguistically relevant phenomena in five languages/dialects, which include “ho” in Cantonese, “hoNh” in Taiwanese Mandarin, “ha” in Beijing dialect, “ha” in Northeast dialect and “huh” in American English.

Chapter 1 serves as the introduction, presenting the research object and questions to be addressd in this study, reviewing related literature on intersubjectivity and ha-type markers and providing theoretical background and data sources.

Based on observation of specific contexts, Chapter 2 draws distributions of ha-type markers within the sentence, the turn and the sequence as formally diagnostic criteria, and properties of the host constituent, speech acts and the speaker’s intention and informational authority of interlocutors as the semantic and pragmatic criteria, categorising 18 semantic primitives.

In Chapter 3, a conceptual space of ha-type markers with 18 semantic primitives has been constructed step by step. First of all, following the connectivity hypothesis and minimum relevance principle, we calculate and compare the degree of association between different primitives and construct a low-resoluiton conceptual space, after which we get a higher-resolution version by virtue of semantic analysis. And then five modes of semantic maps are separated from the conceptual space according to collected data.

Chapter 4 discusses the commonalities and individualities of ha-type markers from the functional, distributional and prosodic perspective and provides explanations for cross-linguistically relevant phenomena. We shed light on similarities and differences reflected on the conceptual space and semantic maps, and explore the relationship between the pause of ha-type markers and their scope and semantic orientation. And we also provide rationales for the postposition of ha-type markers when they co-occur with other sentence final particles. With the help of the Stance Triangle, we explore the categorisation of intersubjectivity. Taking ha-type markers as the starting point, we propose that there may be a functional continuum of sentence-final paritcle, interjection and tag.

Chapter 5 is a brief summary of this paper, in which deficiencies of the present study and avenues of the future work are pointed out.