The present study focuses on the syntax of the sentence-final particle gaa3 in Cantonese from a cartographic perspective. Gaa3 can be observed in three distinct constructions (exclamative, declarative and interrogative) with respect to five different contexts (plain statement, simple information-seeking question, exclamation, warning, and quick-answer-demanding question). By examining the grammatical properties and interpretations of each gaa3, this paper offers a unified syntactic analysis for the different occurrences of gaa3. We argue that Tang’s (2020) framework can well account for the different occurrences of gaa3, which involve different underlying movements. This adds weight to the neo-perfomative approach which aims at syntacticizing properties of speech acts with a fine-grained hierarchy at the left periphery.

Based on a tri-domain perspective (content, epistemic, and speech-act), this paper examines the differentiation of “V qilai” in modern Chinese and the generative mechanism of its connective function. Corpus analysis reveals that “V qilai” exhibits a progressive differentiation across the three domains. In the content domain, it functions as a verb-complement structure, connecting actions syntactically through temporal sequences. In the epistemic domain, it serves as the initial component of the “V qilai AP” construction, establishing a perspectivized evaluative framework for supra-sentential inferential connection. In the speech-act domain, it acts as a discourse marker, organizing text through stance negotiation. The study shows that the three domains of “V qilai” form a continuum across three parametric systems: syntactic hierarchy (syntactic → supra-syntactic → textual), subjectivity (objective → subjective → intersubjective), and functional orientation (propositional → inferential → discursive). The connective function of “V qilai” results from the interaction of cognitive domain shifts, subjectivization, and syntactic regulation.

In a numeral-classifier phrase, the approximative duo has three possible syntactic positions: post-classifier, between the numeral and classifier, and within the numeral constituent. This paper terms this the floating phenomenon of approximative duo. Starting from this phenomenon and base on the cardinal-multiplier analysis and the phrase structure hypothesis for numerals, we proposes a novel analysis of the Chinese numeral-classifier structure: the Chinese “numeral” should be decomposed into a series of Cardinal Phrases (CardP), headed by a cardinal word (e.g., shi ‘ten’, bai ‘hundred’) and taking a multiplier word (e.g., yi ‘one’, er ‘two’) as its specifier. Within this analysis, the approximative duo is argued to be a special type of multiplier word, and classifiers are analyzed as special types of cardinal words. The essential nature of duo’s floating phenomenon is thus that duo functions as the multiplier for different cardinal words.

The de-type compound marker gezai in early Wu dialect literature has tense-aspect functions. Compared to zai, gezai cannot report hot news in on-site context. In narrative discourse, VP+gezai can only show non-main event line and cannot advance the event chain forward, whereas VP+zai has no such restriction. This distinction indicates different tense-aspect meanings encoded by gezai and zai. Gezai denotes the relative-past tense and expresses the completion of an event before the reference time, while zai denotes the meaning of “change of event” besides relative-past tense.

The disyllabic tentative marker kanqiao in the Nanjing dialect conveys meanings of both attempt and warning, each exhibiting distinct syntactic distributions. Kanqiao is derived from the contraction of kankanqiao and has undergone grammaticalization to become a tentative marker through syntactic reanalysis. Kanqiao mostly conveys the meaning of pseudo attempt (warning), showing functional differentiation from the monosyllabic tentative markers kan/qiao, which typically indicate genuine attempts. This phenomenon is prevalent in Chinese dialects, and its fundamental motivation lies in the different perspectives of human cognition and understanding of the attempted action.

This study employs sociolinguistic methods to examine the use of erhua words in the spoken language of 205 Beijing residents. It is found that erhua words are still active in the daily conversations of contemporary Beijing residents, although with significant individual differences in usage. Age and education are important social factors influencing the use of erhua words. Females, teenagers, and students use fewer erhua words, and their expressions are relatively standardized. Over the past 40 years, the use of erhua words has experience a noticible decline, with some vulgar erhua words nearly disappearing. Additionally, interpersonal differences in usage have also significantly decreased. However, the standardization of language use and the positive attitude of Beijing residents towards erhua words still ensure its continued use.

This paper discusses the meaning of the conjunction huòzhě ‘or’. Firstly, it analyzes the interpretation of the semantic truth value of huòzhě-sentences, and points out that the semantic truth value of some huòzhě-sentences shows a different way from that of the logical meaning of disjunction. Secondly, huòzhě-sentences are categorized from the perspective of the semantic relation of connecting items, indicating that there are different semantic possibilities of the items connected by the conjunction huòzhě. Finally, based on the conceptual dichotomy of “role-value”, the core meaning of huòzhě is summarized as: choosing among multiple alternative “value” to satisfy the requirement of “role”.

Based on spoken dialogue, this study discusses the stance-taking function of the “Lao A le” structure from an interactional perspective. The findings show that the adjectives used this structure are qualitative and comparative in nature, primarily consisting of monosyllabic and disyllabic words. “Lao A le” reflects the speaker’s distinct epistemic stance, including incremental expression, evaluative expression, and counter-expectation expression. It also conveys a clear affective stance, typically expressing positive emotions. Furthermore, it demonstrates the speaker’s attitudinal stance, that is, inviting the hearer to align with their stance, thereby achieving stance alignment. The analysis of epistemic, affective, and especially attitudinal stance in this article offers a new perspective on the study of stance-taking in Chinese discourse.

In Standard Mandarin, when the sentence-final particles de and le₂ co-occur, examples can be found where de precedes le, as well as where de follows le. After differentiating between various uses of de and le through independent syntactic evidence, these distributions can be explained. However, in Chaoyang Min, the order of the elements seemingly corresponding to Mandarin’s de and le does not at all align with predictions based on the syntactic-functional analysis in Mandarin. When the A-not-A questions with de’s dialectal counterparts in Wu and Cantonese are also taken into comparision, it is all the more obvious that the de counterpart in Chaoyang Min shows a strong tendency to position itself close to the predicate. Based on the conjunction analysis of sentence-final particles, these differences can be generalized as a single parameter: whether the initial internal conjunction always chooses de.

This paper describes the sentence-final particle keh in declarative and interrogative sentences in Jinhua Wu dialect. In declaratives, keh exclusively occurs in verbal predicate clauses but is barred from nominal predicate clauses, typically conveying confirmatory assertion. In interrogatives, keh optionally appears under pragmatic conditions to reinforce interrogative force, signalling high response expectations of the speaker. The study further compares functions of sentence-final de- type particles across seven Wu dialects, revealing their robust functional development in all cases, with Jinhua dialect being particularly prominent. Specifically, keh exhibits syntactic obligatoriness in assertions of confirmation, and its functionality has historically expanded to yes-no questions, alternative questions, and tag questions, serving to strenthen interrogative force and heighten interactive expectations.

This study focuses on native speakers of the Lahu language and investigates the language processing mechanisms of Lahu nouns and verbs using ERP experiments with vocabulary judgment tasks. The results reveal that there are semantic differences between the categories of nouns and verbs in Lahu language, especially in the early and late processing stages. In the early stage, a direct mapping from phonological form to conceptual representation is observed, with Lahu nouns eliciting a more positive P2. The middle stage involves deep semantic analysis, where Lahu verbs elicit a more negative N400 than nouns. In the late stage, the reconfirmation of semantic information occurs, and Lahu nouns elicit a more positive LPC compared to verbs.

The diachronic evolution of “V→S” of Chinese motion events has received great attention. However, there is still a controversy regarding whether “V→S” has been completed or not in Chinese history. This study advocates that Chinese motion event has finished the change of “V→S”, displaying a cyclical pattern of “V→E→S→(V)” (V being in parentheses indicates that Modern Chinese has not become a typical V-language). Adopting both qualitative and quantitative research methodology, this study follows the grammaticalization cline of the PATH verbs in Chinese and constructs a model of “The Cyclical Change of Chinese Motion Events”. To test the feasibility of the model, we conduct a quantitative study of the diachronic change of the PATH verb lai. This paper reveals that the transition of “V→S” had already been completed during the Song and Yuan dynasties, and since then Chinese tended to change from S-language to V-language. V-framed representations are tending to be predominant in Modern Chinese. The cyclical change in the lexicalization pattern of Chinese motion events is the result of the interaction between “clarity” and “economy”, and also the outcome of the influence of multiple factors, such as disyllabification, grammaticalization, and lexicalization. This study is a preliminary attempt to observe the evolution pattern of Chinese language from the perspective of “linguistic cycle”, which will stimulate deeper understandings of the Chinese typology and its diachronic changes.

It is widely assumed that yi ‘one’-CL-NP phrases in Mandarin are consistently interpreted as indefinite, whereas numeral phrases with numbers greater than one can receive definite interpretations. This paper challenges that assumption, arguing instead that Mandarin numeral phrases do not intrinsically encode distinct semantic properties. The observed contrast in definiteness between yi-CL-NP and num(>1)-CL-NP phrases, as well as the apparent restrictions on the anaphoric use of yi-CL-NP, are best explained in terms of their competition with bare nouns. Adopting the principle Minimize DP! (Patel-Grosz & Grosz 2017), this paper proposes that when a yi-CL-NP phrase and a bare noun denote the same entity, and no independent factor licenses the former, the syntactically and semantically simpler bare noun is obligatorily preferred in definite contexts.

The adverb guozhen carries an independent syntactic function. It deals with the relationship between the preceding information and the current sentence, whose core meaning is “the aforementioned information has been verified”. The semantic structure of guozhen includes three parts: the information to be verified, the act of verification, and the result of verification, which reflect respectively the speaker’s epistemic stance of uncertainty, intention to seek verification, and acceptance of the verified result. The information state therefore changes from unverified to verified. The difference between guozhen and guoran lies mainly in the speaker’s attitude and the semantic functions they perform. With guozhen, speakers tend to take a skeptical stance, intending to verify the preceding information. With guoran, speakers often assume a believing stance, and the subsequent propositions typically support the preceding information. In some discourses, the speaker’s attitude may be omitted and their stance is unknown, making guozhen and guoran interchangeable. However, the semantic structures they activate remain different.

There is a sentence-final particle [kəʔ] that expresses realis existence in Haimen dialect. The nature of the particle differs from the focal marker [kəʔ] and the modal particle [kəʔ]. Depending on the semantic type of the predicate, [kəʔ] can be divided into three forms: [kəʔ]_past, which expresses the existence of past events, [kəʔ]_static, which indicates the existence of static states, and [kəʔ]_habitual, which denotes the existence of habitual events. All three forms are used obligatorily. Notably, [kəʔ]_habitual can appear in negative sentences, while [kəʔ]_past and [kəʔ]_static cannot. Functionally, all three forms encode the “existence” aspect, with [kəʔ]_past also indicating past tense and [kəʔ]_static conveying an affirmative tone. The sentence-final particle [kəʔ] in Haimen dialect demonstrates that the sentence-final particle ge in Wu dialect has a tense-aspect function.

The article examines the corresponding particle of de in Penang Hokkien, a Southern Min variety used in Malaysia, which is [e¹³]. We first distinguish three usages of the sentence-final particle [e¹³] in Penang Hokkien: as a modifier marker, used in clefts, and in sentence-final position. We then examine the syntactic level of the sentence-final [e¹³] in Penang Hokkien while also discussing the system of sentence-final particles in Mandarin. We find that the sentence-final [e¹³] is within the CP layer, specifically in AssertionP. This projection is higher than OnlyP, but lower than Degree SFPs, including focus particles. Despite being in the same projection, Mandarin’s sentence-final de and Penang Hokkien’s sentence-final [e¹³] differ both grammatically and semantically. The article also demonstrates the periphery structure of Penang Hokkien through the order and co- occurrence of sentence-final particles.

The particle [kɤ] (including its phonetic variants) is widely used as structural particles in southeastern Chinese dialects, showing a remarkable areal feature. [kɔ³³] in Yangchun Cantonese in western Guangdong can be used as classifier, demonstrative, structural particle, and sentence-final particle synchronically. In the construction “subject+copula+VP+[kɔ³³]+NP”, [kɔ³³] can be interpreted either as a structural particle or a sentence-final particle with the omission of the head NP. It has been conventionalized as a dedicated sentence-final particle eventually. Comparison with the other Chinese dialects and several non-Chinese languages in southeastern China indicates that allowing the omission of the head NP in the construction “subject+copula+VP+[kɤ]+NP” is the necessary but not sufficient condition for the formation of the sentence-final usage of [kɤ] as structural particles.

In the Poyang Gan dialect, the sentence final particle [kɯ] has two pronunciations: an unstressed [kɯ⁰] and a stressed [kɯ¹⁴]. The unstressed [kɯ⁰] corresponds to the Mandarin sentence-final particle de though there are differences in specific usage. Based on its function, [kɯ⁰] can be divided into [kɯ⁰]_Confirmation and [kɯ⁰]_Focus. The stressed [kɯ¹⁴] is mainly used in the “V+ zhi (+O)+[kɯ¹⁴]” structure, emphasizing that the verbal event has been completed. In this context, it serves a partial perfect aspect function. Through a detailed discussion of the two forms of [kɯ], this paper argues that the functional evolution of the two forms of [kɯ] follows the progression: [kɯ⁰]_Focus > [kɯ⁰]_Confirmation > [kɯ¹⁴].

On the basis of an extensive investigation of historical literature in Cantonese, it is proposed that the adverbs referring to cai in the Guangzhou dialect has undergone two diachronic substitutions of the dominant word since the early Qing dynasty. Zheng was the primary expression form, which was replaced by zhi by the early 19th century, and then by xian in the mid-20th century. The disyllabic word xianzhi has seen an increase in usage frequency since its emergence, but it has always been in a secondary position. The replacement and diffusion of the adverbs like cai in Cantonese have created a geographical distribution pattern of such adverbs in modern Yue dialects. This study has enlightening significance in examining the kinship, evolution speed, evolution mechanism, and literature nature of Chinese dialects.

The Chinese measure word jiān demonstrates exceptional collocational ability in the regions of Hong Kong and Macao, where it can be used not only to measure individual rooms like “classrooms”, but also to quantify nouns with holistic architectural entities or institutional organizations such as “companies”, “schools”, “factories”, and “centres”. This usage differs significantly from the word’s usage in modern standard Chinese. Therefore, this research adopts a corpus-based, data-driven approach to analyze the emergence and evolutionary path of jiān across time and regions. The findings show that, as language continuously evolves, jiān has undergone semantic generalization. In Hong Kong and Macao, the broadened usage has stabilized and persisted, whereas it later experienced semantic narrowing in mainland China, resulting in a reduced collocational capacity.

The modern Chinese sentence-final particle na has two origins: first, it derives from a phonetic change of a when the preceding word has an “-n” coda, denoted as na₁; second, it originates from the syllabic fusion of the sentence-final particles ne and a, denoted as na₂. Drawing on a series of Ho-Kian Fu dialect documents compiled by late Qing missionaries, and supplemented by Beijing dialect documents and contemporary Hebei dialect data, this study demonstrates from multiple perspectives the existence of the consecutive use of the sentence-final particles ne and a, thereby substantiating the plausibility of the syllabic fusion theory for na₂. After ne a fused into na₂, the mood expressed by na₂ has also developed from superimposed mood to composite mood. Specifically, in non-interrogative sentences, na₂ conveys a strong reminding mood; in interrogative sentences, it conveys a mitigated pragmatic force of reminder and inquiry (in information-seeking questions), or the speaker’s critical or reproachful stance (in rhetorical questions). Owing to its multifunctionality and high frequency, na₂ likely exerted an assimilative effect on na₁, which is homonymous and occupies the same syntactic position. This ultimately led to the merger of this portion of na₁ with na₂, together constituting the sentence-final particle na in modern Chinese.

Interrogatives can be categorized into polar questions and constituent questions based on the dichotomy of question types. The former seeks confirmation of the proposition put forth, while the latter seeks information specifically targeted by the interrogative constituent. In the Hanzhong dialect of Shaanxi, “…ba” interrogatives can also be divided into two types: the confirmation-seeking polar question with ba and the information-seeking constituent question “VP ba”. By analyzing the nature of ba and the usage, syntactic, and semantic characteristics of “VP ba”, we argue that “VP ba” essentially belongs to the category of A-not-A questions. Since the fact that the Hanzhong dialect is spoken in the transitional zone between Zhongyuan Mandarin and Southwest Mandarin, “VP ba” is inevitably influenced by multiple internal and external factors of the language system. There are four stages in the evolution from A-not-A questions to polar questions. “VP ba” is at the third stage of the evolution chain.

The plural markers in the Guanzhong dialect primarily fall into four categories: the de type, the xie type, the men type, and the huo type. These plural markers exhibit certain differences in terms of geographical distribution, scope of application, and semantic expression. Among them, the plural marker men has the widest geographical distribution, while the plural marker xie has the narrowest. The plural markers men, de, huo, huoli, and xiege can only be attached to human-referring nominal elements to denote plurality, whereas the plural marker xie is not subject to this restriction and can also be attached to non-human-referring nominal elements to indicate plurality. The plural markers huo, huoli, and xiege can only express ordinary plural, while the plural markers men and de can express both ordinary plural and associative plural. The plural marker xie, on the other hand, can express ordinary plural, associative plural, and similative plural. The divergent origins of these plural markers are the primary reason for these observed differences.

This paper provides a systematic analysis of the syntactic structure of the “NP₁ + V + you ‘have’ + NP₂” construction in the Jiyuan dialect, examining the grammatical functions of you ‘have’ and exploring its grammaticalization process and conditions. The research findings are as follows: (1) You ‘have’ serves multiple functions, including possession, existence, completion-persistence, and focus marking, forming a continuum that ranges from “spatial existence” to “temporal completion”. (2) It has undergone an evolution characterized by a progression from “possession > existence > completion-persistence > focus marking”. (3) In comparison to Mandarin, high transitive verbs such as duo ‘chop’, chi ‘eat’, and you ‘swim’ can freely enter this construction in the Jiyuan dialect. Additionally, the role of “NP₁” has expanded from locative expressions to include agent subjects. (4) The grammaticalization of you ‘have’ is influenced by several factors: the reduction of verb reduplication prompts a reanalysis of the structure (V^D + you ‘have’ →V⁰ + you ‘have’), semantic appropriateness facilitates its transition from spatial contexts to temporal ones, and frequent usage accelerates its grammaticalization process. The multifunctionality of you ‘have’ in the Jiyuan dialect illustrates its evolutionary trajectory from being spatially driven to temporally driven.

Xiang serves as an incipient proximative marker in the Xuzhou dialect. Compared with fully grammaticalized proximative markers, it exhibits a wider range of syntactic and semantic properties characteristic of the early stages of grammaticalization. Drawing on the synchronic syntactic differences of xiang across its diverse semantic functions, a more comprehensive grammaticalization pathway from desiderative verbs to proximative markers can be constructed, enabling a deeper exploration of the evolutionary mechanism underlying this grammaticalization. Previous studies on proximative markers have rarely addressed the following syntactic features of xiang: (1) When the subject is a human and xiang is followed by a dynamic non-volitional situation, a distinction arises between subject-oriented and speaker-oriented readings, with the former encoding the subject’s perception; (2) When the subject is a human and xiang is followed by a dynamic volitional situation, xiang can convey the proximative reading only in highly constrained contexts, and in most cases retains its core desiderative meaning. According to feature (1), it is clear that the grammaticalization of desiderative verbs commences with the functional expansion of xiang to subsequent verbs. Taken together, features (1) and (2) demonstrate that the polysemous stage - where desiderative and proximative readings co-occur - emerges after the proximative has developed, rather than immediately following the desiderative verb stage.

This paper uses Teochew loanwords from Thailand, Vietnam, and Malaysia (recorded in Teochew dialect documents) as research material. By comparing syllable structure differences among overseas Teochew, Thai, Vietnamese, and Malay, it analyzes the selection mechanism adopted by consonants and vowels in different syllable positions during language contact, and summarizes the transformation rules and transliteration strategies of loanwords. The study finds that: (1) In Southeast Asian Teochew, loanwords are incorporated through syllable matching based on relative phonological similarity. (2) The phonological inventory and inventorial syllables of Teochew play a key role in the transliteration of loanwords. (3) Over the past century, the direction of phonetic evolution in Teochew has remained consistent both at home and abroad, reflecting frequent communication among Teochew communities worldwide.

The jiaomo-type words, functioning as verbs, adverbs, and conjunctions, occur in various forms. The most common form is the verb “jiao (pro) mo”, which derives from dissuasive-interrogative constructions of the pattern “jiao pro mo VP”, originally meaning “to allow to proceed unhindered”. Historically, these forms were carried into central regions by migrants from Jiangxi who settled in present-day Hunan and Hubei. Through grammaticalization, the verbal construction underwent syntactic reanalysis in certain contexts, evolving into a concessive conjunction equivalent to jinguan ‘despite’. These migrants outnumbered the local population in areas such as Huangzhou and De’an, leading to a contiguous distribution of jiaomo-type words in northeastern Hubei, while other regions exhibit a more scattered distribution. With further grammatical development in the Changyi dialect of Xiang, jiaomo-type verbs have been reanalyzed via inferential contexts into adverbs meaning “why not”. Geographically, jiaomo-type words are primarily distributed in the Huangxiao dialect and can be regarded as its dialect characteristic words.

The sentence-final particle ge in the Anyuan Dialect shares functional similarities with Mandarin Chinese de, with its core discourse function being to mark the speaker’s assertion towards realistic events and epistemic commitment towards unrealistic events. Other functions of ge/ de, such as explanation, refutation, correction, and reminder, emerge from this core function through interactive contextualization. Notably, the sentence-final ge in the Anyuan Dialect further emerges mirative mood, as well as pursuit questioning, confrontational challenging, admonition, suggestion, and directive functions, with certain functions exhibiting a higher degree of conventionalization.

The sentence-final [kai⁵⁵] in the Chao’an dialect can be classified as [kai^55-21]_focus and [kai^55-21]_confirm, based on the “[kai^-42]...[kai^55-21]” construction, which serves functions of past-focus and affirmation. However, [kai^55-21]_focus cannot independently serve as a focus marker, nor can [kai^55-21]_confirm function alone as a modal particle. They both require co-occurrence with [kai⁴²] in the “[kai⁴²]...[kai^55-21]” construction. The evolutionary paths of sentence-final [kai⁵⁵] in the Chao’an dialect include: (1) quantifier → indefinite pronoun → definite pronoun → copula → contrastive focus marker; (2) quantifier → indefinite pronoun → definite pronoun → structure auxiliary → contrastive focus marker. The functional development of “[kai^-42]...[kai^55-21]” construction follows the path: Expressing judgment → Marking contrastive focus → Expressing confirmatory mood.

This paper comprehensively describes the functions of the conjectural particle [mɐt⁵⁵ tsei²¹] in the Yangjiang Yue dialect, along with its pragmatic functions. It contrasts [mɐt⁵⁵ tsei²¹] with [kwa³³] in the Guangfu Yue dialect and [pa⁰] in Mandarin. Based on a synchronic analysis, the study reconstructs the grammaticalization pathway and pragmatic extension trajectory of [mɐt⁵⁵ tsei²¹], while exploring the origin of its conjectural meaning. The findings demonstrate that [mɐt⁵⁵ tsei²¹] has undergone a grammaticalization process from conjectural adverb to conjectural particle. Both of its pragmatic functions emerging in discourse derive directly from the pragmatic extensions of its core conjectural semantics. Furthermore, the conjectural semantics likely originate from the earlier function of [mɐt⁵⁵ tsei²¹] as a degree adverb. This functional expansion not only represents a new case of the adverb > modal particle grammaticalization pattern in the Yue dialect but also corroborates the cross-linguistically attested extension model: quantity > epistemic modality > pragmatic function, which is observed in Chinese and even the world languages.

As a common linguistic phenomenon in ancient Chinese, preposition stranding nearly dies out in Mandarin Chinese and only makes appearance in modern Chinese dialects. In the process of diachronic evolution, prepositional phrases degrade from pP phases into PP projections, which gives rise to different behaviours concerning preposition stranding between ancient times and modern times. This paper makes an analysis of the mechanism of preposition stranding in ancient Chinese and the restrictions concerning the same phenomenon in Mandarin Chinese under the theoretical framework of generative grammar and offers an explanation for preposition stranding in modern Chinese dialect. This paper holds that phase degradation and proper government serve as the intrinsic causes for the difference in preposition stranding in ancient and modern times, and the imbalance of diachronic development offers conditions for preposition stranding in modern Chinese dialects.

This paper analyzes the restriction on the distributivity of Chinese adverbial ge. We propose that ge needs to satisfy the condition of differentiation, which requires that both its sorting key and distributed share contain multiple distinct members, and further delineate the syntactic, lexical, and contextual factors which facilitate differentiation. The paper also addresses ge’s meaning of “difference”, arguing that this meaning of “difference” targets the level of pragmatics and conveys the speaker’s expectation which derives from the condition of differentiation. The notion of differentiation proposed in this paper can give the licensing conditions and semantic properties of ge discussed in the literature a more systemic explanation and uncover more factors which affect the acceptability of ge-sentences.

The grammaticalization of content words is an important aspect in the study of Chinese grammar. This paper examines the multifunctional usage of hou in the Hanshan dialect and finds that hou can be used as both a verb and a conjunction. As a conjunction, it introduces time clauses and conditional clauses; however, its ability to introduce counterfactual conditionals is less developed compared to Mandarin. The paper also explains the motivation and mechanism of the grammaticalization of hou from a verb to a conjunction in terms of syntactic semantics, context, and invited inference. It gives indirect evidence supporting the analysis of hou as a conjunction.

There is a unique AAAA pattern in Hakka dialects. It is generated through quadruplication of a monosyllabic word (or morpheme), with tonal variation involving rises and falls in pitch. Such forms are typically used as adjectives that are characterized by vividness and emphasis. Comparative analysis shows that the tonal variation is not necessarily linked to tonal categories, but is related to tonal types. Combinations of suprasegmental features such as pitch and duration have solidified into prosodic patterns in Hakka dialects. With the expansion of the semantic range and the influence of analogical processes, AAAA words undergo lexicalization and become prosodic words. Subsequently, some verbs and a few adjectives enter this prosodic structure, and through quadruplication, they develop into adjectives that imitate the situation. While retaining some of the original semantic features of the base form A, the AAAA pattern acquires additional semantic properties, including [+imitative (including onomatopoeic)], [+dynamic], [+continuous process], and [+intensified degree]. AAAA patterns in Mandarin and other Chinese dialects serve as morphological devices. In contrast, they function as word-formation mechanisms in Hakka dialects.