JCL Monograph Series NO.22 专著系列 22 卷 – 2007

Aspect(s) of Chaoshan grammar: A synchronic description of the Jieyang variety
潮州话揭阳方言语法研究
Hui Ling Xu 许慧玲 著

Chapters 章节

Abstract 摘要
1.1 Aims, scope and outline of the thesis The Chaoshan dialect refers collectively to a group of mutually intelligible dialects spoken in the Chaozhou-Shantou region, a coastal region in the eastern part of Guangdong province in the People’s Republic of China. Linguistically, the Chaoshan region borders the Min-speaking Fujuan province to the east, and within the Guangdong province the Hakka-speaking region to the north and the Yue-speaking region to the west (see 1.4 for discussion on linguistic affiliation). The Chaoshan dialect is one of the most conservative dialects in China, retaining many archaic linguistics features which have disappeared in other Sinitic languages (see 1.5). However, previous studies of this important dialect group are sketchy and many attempts have been confined to phonology (see 1.6). The task of this thesis is to provide a more comprehensive analytical and functional description of important aspects of Chaoshan grammar, aimed at filling this gap. The dissertation focuses on data from the Jieyang, the grammatical topics and features addressed in this thesis are characteristic of Chaoshan as a whole, since the variation among the Chaoshan subdialects is predominantly of a phonetic nature. This thesis is not a comprehensive grammar of the Jieyang dialect. Rather, it focuses on certain important aspects and provides a more analytical and in depth treatment. The topics and constructions chosen for discussion and description are core grammatical structures including those forming questions, negations, comparisons and expressing temporal relationships of events. These syntactic areas are known to have distinct Chaoshan characteristics which diverge from other Sinitic languages, particularly the official language of China, Mandarin. Many other linguistic issues would not be specific to the Chaoshan area. For example, word-classes (such as verb versus adjective or verb versus preposition); basic grammatical relationships or syntactic functions (such as subject and object); topicalization; serial verb constructions, and morphological issues (such as the question of word-hood) arise more or less equally for all Sinitic languages. Therefore, these general issues have not been covered here, as analyses and research on these topics are readily available, especially those that focus on Mandarin data.

This thesis is divided into 11 chapters: introduction (1); phonology (2); morphology (3); pronouns (4); numeral classifiers, possessives and relative clauses (5); the aspectual system (6); passives and the ‘k’eʔ2i33 construction’ (7); the pre-transitive construction (8); negation (9); interrogatives (10) and the constructions of comparison (11). In the remainder of this introductory chapter, I will present some background information about the demography, linguistic affiliation and previous research on the Chaoshan dialect, as well as describing the methodology and convention used in the present study.

1.1 Aims, scope and outline of the thesis
1.2 Geographic location and population distribution of the Chaoshan region
1.3 A brief account of the history of the region and aspects of its culture
1.4 The dialect
1.5 Typological linguistic features of the Chaoshan dialect
1.6 Previous study on the dialect
1.6.1 Early works
1.6.2 Works by Chinese scholars
1.7 Significance
1.8 Methodology
1.8.1 The data
1.8.2 The linguistic consultants
1.8.3 Conventions used in presenting the data
1.8.4 Approach to data analysis

Abstract 摘要

2.2 The sound system of Jieyang
2.2.1 Initial consonants
2.2.2 Finals
2.2.3 Tones
2.2.3.1 Basic tones
2.2.3.2 Tone Sandhi

Summary of section

This section has shown that tone sandhi is a very important and an integral part of the sound system in Jieyang. However, TS rules are not straightforward. In existing literature on Chaoshan phonology, two processes of TS are noted, which can be called anterior and posterior TS. The former is characterized as having the syllabus except the right-most one surfaced in sandhi forms in a string of two or more syllables. While the rules of TS for a string of two syllables in the Jieyang dialect are documented (see Cai Junming 1991; Lin Lunlun and Chen Xiaofeng 1996), how these elementary processes interact to product the final output of a string of more than two syllabus have yet to be researched. Posterior TS is basically a process of lowering the right most syllable. However, as in Mandarin, in many cases it is an unpredictable process, which is basically lexically idiosyncratic.

Abstract 摘要

3.1 Introduction
It is generally held that Sinitic languages are typologically isolating, that is, each word is made up of one morpheme. However, as Tsao Feng-fu(2001) points out, the isolating label suits Classical Chinese better than modern Chinese because in Classical Chinese, ‘most Chinese characters correspond to monosyllables which are both morphemes and words’, whereas in modern Sinitic languages, ‘most Chinese characters correspond to morphemes only’ (p.286), as most words now consist of two or more morphemes. In fact, the trend towards polysyllabicity of Chinese words is corroborated by findings of a study conducted in the 1980s which shows that among the first 9000 most productive or frequently used words in Mandarin, only one quarter is monosyllabic (Su Xinchun 1995: 160). The morphological complexity of words in modern Sinitic languages are achieved through derivational processes such as reduplication, affixation and compounding, with compounding being the most productive. Inflectional morphological processes, however, are generally not found in Sinitic languages. Being one of the most conservative and oldest dialects, the Jieyang dialect preserves more monosyllabic words than its northern counterparts such as Mandarin. However, like other modern Sinitic languages, it also make use of compounding as well as reduplication and, to a lesser extent, affixation. In this chapter, I describe affixation, reduplication and compounding in the Jieyang dialect, making references to Sinitic languages in general and Mandarin in particular wherever necessary. Due to the limitation of space, examples given in each subsection are far from exhaustive. They only serve to illustrate the discussion.

3.1 Introduction

3.2 Affixation
3.2.1 Prefixes
3.2.1.1 a33
3.2.1.2 lau35
3.2.1.3 ts’iu33_
3.2.1.4 toi35_
3.2.1.5 ho53_
3.2.1.6 mo53_
3.2.1.7 sio33_
3.2.1.8 huaŋ33_
3.2.1.9 u35_ & bo55_
3.2.2 Suffixes
3.2.2.1 -kiã53
3.2.2.2 –hue53
3.2.2.3 –t’au55
3.2.2.4 –ko213
3.2.2.5 –bo53, -kou33 & -heŋ55

3.3 Reduplication
3.3.1 Reduplication of adjectives
3.3.1.1 Forms
3.3.1.2 Semantic and syntactic features of reduplicated forms
3.3.2 Reduplication of measure words, classifiers & nouns
3.3.2.1 Reduplication of measure words/classifiers
3.3.2.2 Reduplication of nouns
3.3.3 Reduplication of verbs
3.3.4 Reduplication of onomatopoeic expressions
3.3.5 Reduplication of phrases

3.4 Compounding
3.4.1 Nominal compounds
3.4.2 Adjectival compounds
3.4.3 Verb Compounds

Summary of section
This chapter deals with three derivational processes in the Jieyang dialect: affixation, reduplication and compounding. In terms of affixation, the Jieyang dialect has prefixes as well as suffixes but no infixes. The prefixes and suffixes discussed here are mainly from the local lexicon. Those which have cognates in Mandarin have been shown to exhibit certain features not found in Mandarin. In the Jieyang dialect, reduplication is rich, as it is in other Sinitic languages. In terms of types of reduplication, it has been shown that complete reduplication, thatg is, the copying of an entire word, as well as partial reduplication which copies only part of a word are all common. Reduplication in the Jieyang dialect not only takes various forms, it also serves different purposes, such as expressing intensified meaning, as exemplified in adjectival reduplication; in deriving words, as exemplified in the NN reduplication which yields an adjective, and in denoting distributive meaning as shown in the reduplication of classifiers. Compounding is very productive, and the processes can result in three types of compounds: nominal, adjectival and verbal. A very significant feature of compounds is that the semantic connection between the meaning of a compound and the meaning of its constituents in many cases are not transparent because the connection is more often metaphorical, figurative, metonymical and inferential rather than literal.

Abstract 摘要
In this chapter, I discuss personal pronouns, reflexives and nominal demonstratives. Interrogative pronouns are discussed separately in Chapter 10 on interrogatives so as to avoid overlap here.

4.1 Personal pronouns
4.1.1 Forms
4.1.2 Usage
4.1.2.1 Singular pronouns: ua53, lƜ53,i33
4.1.2.2 Plural forms: naŋ53, uaŋ53, neŋ53, i33naŋ55-11
4.1.2.3 The possessive forms3.1 Introduction

4.2 Reflexive pronouns
4.2.1 ka33ki11 as a pronoun
4.2.2 ka33ki11 as an adverbial
4.2.3 The emphatic function of ka33ki11

4.3 Nominal demonstratives
4.3.1 tsi53 and hio53
4.3.2 tsio53 and hio53
4.3.2.1 Semantic and syntactic features
4.3.2.2 Deictic and anaphoric functions
4.3.3 tsia53 and hia53
4.3.3.1 Semantic and syntactic features
4.3.3.2 Deictic and anaphoric functions
4.3.3.3 Other functions
4.3.4 tsia53-35 kai55 and hia53-35 kai55

Summary of chapter

In this chapter, I have discussed personal pronouns and nominal demonstratives in the Jieyang dialect. Personal pronouns do not vary for gender but do vary for number: singular versus plural. The first person plural pronouns also have two forms: inclusive and exclusive naŋ53incl. and uaŋ53excl.. The pronouns are invariant in all positions in the sentence. Possessive relation is expressed analytically with the particle kai55 or a classifier, both of which are placed after the pronouns. However, there are two fused forms which have developed into fully-fledged genitive forms, used only with kinship terms. In terms of usage, the third person singular pronoun is found to be of particular interest because of its various functions. For example, it can be used anaphorically to refer to non-human but animate reference. It can also be used in opposition to the subject NP, the usage of which is also anaphoric. Syntactically, it can also serve as a dummy for the agent rolein the ‘k’eʔ2i33 construction’, and the semantic patient role in the pretransitive construction (see Chapter 7 and 8 for detailed discussion on these two constructions). The system of the nominal demonstratives is quite complex, not only in their semantic features but also in their different deictic, anaphoric and syntactic functions. The paradigm of the demonstratives shows a binary distinction: distal vs proximal, with each sub-paradigm consisting of four forms. Within each sub-paradigm, there is a number distinction between singular and plural (e.g. tsi53 ‘this’ vs tsio53 ‘these’). Some forms are distinguished by whether they can occur with individual entity or denote a kind (e.g. tsi53 ‘this’ vs. tsia53 ‘this.kind.of’ and tsia53-55 kai55 ‘this.kind.of NP’), and syntactically, by whether they can serve as full NPs or only as determiners or both. Tsia53 and hia53 are also found to have discourse-pragmatic functions, serving as conjunctions or as gap fillers in oral discourse.

Abstract 摘要
This chapter deals with three topics that are closely related to the noun phrase: numeral classifiers, possessive structures and relative clauses.

5.1 Numeral classifiers
5.1.1 Mensural Classifiers
5.1.1.1 Collective classifiers
5.1.1.2 Classifiers denoting plurality and nonspecific quantity
5.1.1.3 Body parts as classifiers
5.1.2 Sortal classifiers
5.1.3 Special features of classifiers/measure words
5.1.3.1 Functioning as pronominals
5.1.3.2 Denoting definiteness
5.1.3.3 [CL + mueʔ5] construction
5.1.3.4 With adjectives in expressing vividness
5.1.3.5 In indicating possessive relations

5.2 Possessive construction

5.3 Relative clauses
5.3.1 Relative clause with kai55
5.3.2 The use of a resumptive pronoun in relative clauses
5.3.3 Relative clause with classifiers
5.3.4 Differences between relative clauses with kai55 and with classifiers

Summary of chapter

In this chapter, I have discussed three topics which are integral to the noun phrase: numeral classifiers, possessive constructions and relative clauses. Numeral classifiers in the Jieyang dialect consist of two types: mensural, which are also called measure words, and shape/sortal classifiers. The latter type are found to be more general in meaning and thus have a wider scope of use than those of Mandarin. Numeral classifiers play a very important part in the syntax of noun phrases in the Jieyang dialect, as they have a wide range of functions. In addition to their primary and major functions of enumeration and individuation where they occur between a numeral and a noun, they are also used as an alternative way to form possessive pronouns, in possessive constructions and relative clauses as the linking word between the head noun and the modifier. Another important feature of nominal classifiers in the Jieyang dialect is their ability to combine with a bare noun to mark definiteness in topic positions, a usage which is highly productive in the spoken language. These special functions of numeral classifiers are also found in some Southern and Southwestern dialects such as Cantonese, in Miao and Yao, and in some Southeast Asian languages (see e.g. Bisang 2000). Possessive constructions and relative clauses formed with classifiers denote individual entities while those formed by the grammaticalised morpheme kai55 (or ti33 for the written language) denote general reference. Relative clauses in Sinitic languages in general can only be of the restrictive type. But in the Jieyang dialect, due to tone sandhi, a relative clause can also be analysed as a pivotal construction which is functionally similar to a non-restrictive relative clause in English.

Abstract 摘要

6.1 Introduction
It is well-known that in Sinitic languages, tense is not a grammatically encoded category, but there are rich aspectual systems. While tense and aspect are both concerned with the temporal dimensions of verb phrases, they differ from each other in their semantic focus. Tense relates an event or the occurrence of a situation to the moment of speaking or to some other situation and is thus deictic. Aspect, on the other hand, makes reference to the properties of situations themselves, that is, it is concerned with ‘the internal temporal “structure” of a situation’ (Payne 1999:234). If a situation is presented as an unanalysable whole, without beginning or end points, it is said to be perfective whereas if only part of a situation is presented as an unanalysable whole, without beginning or end points, it is said to be perfective whereas if only part of a situation is presented, it is said to be imperfective (see discussion in Comrie 1976; Dahl 1985; Chung and Timberlake 1985, Smith 1994 among others). However Chappell (1989) proposes that the basic distinction for the category of aspect be divided into bounded and unbounded. Bounded aspect, which corresponds to perfective aspect, refers to markers which ‘encode that the event or state of affairs has either a definite beginning or endpoint’ while unbounded aspect, which corresponds to imperfective aspect, refers to markers which ‘have neither a limit placed on duration nor definite beginning or endpoints’ (p.96). Such a division enables the classification of inchoatives under the bounded aspect because it typically marks the beginning of a new state of affairs or an event. It is pointed out by Comrie (1976:12) that the term ‘perfective’ should be distinguished from ‘perfect’, because ‘perfective’ contrasts with ‘imperfective’ and denotes a situation in its entirety, while ‘perfect’ refers to a past situation that has current relevance. It will be shown in the ensuing discussion that this distinction is relevant to the discussion of the aspectual system of the Jieyang dialect. As in other Sinitic languages, the Jieyang dialect has a very rich aspect system. However, apart from grammatical categories, it also employs verbal complements as well as auxiliary verbs to encode the perfective meaning (see discussion in 6.3). The latter is a distinct characteristic of the Min dialect group (see Tsao 1998).

6.2 Situation types

6.3 Aspectual categories in the Jieyan dialect
6.3.1. The bounded aspect
6.3.1.1 The Perfective aspect
6.3.1.2 The inchoative aspect
6.3.1.3 The Experiential aspect
6.3.1.4 The Delimitative aspect and the Tentative aspect
6.3.2 The unbounded aspect
6.3.2.1 The Progress
6.3.2.2 The Continuous

Summary of chapter

The discussion of the aspectual system in the Jieyang dialect has been carried out in terms of the two main aspectual distinctions: perfective and imperfective. I have discussed six main aspectual categories in the Jieyang dialect: the Perfective, which also includes such meanings as the inception of a state; the Experiential; the Tentative; the Delimitative; the Progressive and the Continuous. Although these aspectual categories are similar to other Sinitic languages such as Mandarin, the discussion in this chapter shows that the syntactic means to express the different aspectual viewpoints, the range of meanings they possess and thus their scope of use can be different. For example, due to stratification and borrowing, the Jieyang dialect boasts more syntactic forms to encode the Experiential aspect than in Mandarin. The discussion also shows that the Jieyang dialect exhibits a lesser degree of grammaticalisation of aspectual markers than in Mandarin. In many cases, this is due to the retention of the source meanings of the markers. For example, in the Experiential, the maker pak2 cannot be used for denoting inferential evidentiality due to the retention of its basic meaning of ‘to know, have knowledge’. Other evidence includes the possibility of aspectual markers to appear after [Verb + Object] rather than attached to the verb, as well as the possibility of using the Experiential marker pak2 independently as an answer. The imperfective viewpoint in the Jieyang dialect also display some salient features. While the basic functions of the Progressive and the Continuous are to encode an on-going activity and an on-going resultative state respectively in a neutral way, both aspectual categories also possess a subset, marked by paŋ213-53-ko213, which encodes ‘intentional’ and ‘wilful’ actions in declarative sentences and ‘agentive involvement’ in imperative sentences.

Abstract 摘要

7.1 Introduction
In this chapter, I discuss a set of constructions which are collectively called the pretransitive construction. It resembles to a certain extent the BA construction in Mandarin, which is one of the most extensively studied constructions in Chinese linguistics because of its unique semantic, syntactic and pragmatic features (see discussions in Wang Li 1958; Y.R. Chao 1968; Thompson 1973; Cheung Hung-nin 1973; Ying-che Li 1974; Li and Thompson 1981; Chen Yuchin 1991; Chappell 1991; Sebesma 1992; Liu Yizhi 2000; Ziegeler 2000). Based on semantic and syntactic grounds, two subtypes of BA sentences have been identified in various studies: one of them is often called the ‘disposal’ form while the other is labeled the ‘causative’ form (see, for example, Y. R. Chao 1968; Chen Yuchin 1991; Sybesma 1992; Yue-Hashimoto 1993; Fan Xiao 2000 and Liu Yizhi 2000). Example (1) illustrates a disposal BA sentence while (2) illustrates a case of the causative type. However, in Chappell (1991), both types are said to be semantically causative:

(1) haizimen ba zongzi quan dou chi-quang le Children BA rice:dumpling fully all ear-bare ASP “The children ate up all the rice dumplings.”

(2) Ba Ping Ping dou huai ji-si le BA Ping Ping all soon upset-die ASP “It just about had Ping beside herself with anxiety.” As can be seen, the first type has a structure [Subj + BA + Obj. +VP] where the object, ie, the patient augment, is preposed to the topic position and is syntactically marked by the morpheme BA. Semantically, this type of sentences denote a set of meanings which are subsumed under the notion popularly known as ‘disposal’ or ‘處置’ in Chinese, introduced first by Wang Li (1958). The meanings of disposal include ‘how a person handled, manipulated, or dealt with the object; how something is disposed of ; how an affair is conducted’ (translated by Li Ying-che 1974: pp. 200-201). In most cases, this type of BA sentence entails affectedness of the patient argument which is reflected in a resultative state. As such, it must have a semantically transitive predicate with the subject interpreted as the agent and the object the affected undergoer of the action. In fact, BA sentences as exemplified by (1) are called the transitive type in Chappell (1991). The causative type, on the other hand, has a different syntactic configuration altogether, which is [BA – Subj.- VP], with the BA NP acting as the actor of the verb, or the experience, rather than the affected undergoer. In Chappell (1991), this type is called the intransitive type, as opposed to the transitive type.

The disposal type of BA sentence contrasts with a SVO sentence both structurally and semantically. Compare the following two sentences:

(3)
a. ta chi le yi wan fan
b. ta ba na wan fan chi le

The most salient difference between a canonical disposal type of BA sentence, such as (3b), and a SVO sentence, as represented in (3a), is that the patient argument in a disposal sentence has been preposed to a secondary topic position with the predicate appearing in clause-final position. The BA morpheme, which is treated as a first verb in Y. R. Chao (1968:342), is thus used to disambiguate the direction of action of the main verb towards the preverbal NP (ibid:345). As a result, (3a) can be used to answer ‘what did he do?’ whereas (3b) can be used to answer ‘what did he do to that bowl of rice?’ In other words, a SVO sentence such as (3a) is a neutral report of an event, without particular attention being paid to either the agent or the patient. If any, the semantic focus tends to be laid on the influence of the verb on the agent. A BA sentence such as (3b), on the other hand, shifts the attention on the fate of the rice and how it is affected: it is gone because it has been eaten. In the Jieyang dialect, the pretransitive construction shares certain semantic and syntactic features with the disposal subtype of BA construction in Mandarin, but there is no corresponding causative type or intransitive type. Example (4) illustrates a Jieyang version of a disposal type of BA sentence in which the patient argument is preposed to the topic position and is syntactically marked, here, by one of the vernacular forms t’aŋ213 (see 7.3 for detailed discussion). The sentence denotes what the agent did to the patient argument with the implication that the patient argument was thus affected:

(4)
tsiaʔ2 ŋiau33 t’aŋ213-53 bue53-35 hƜ55 tsiak5 k’Ɯ213-21
CL cat PRET CL fish eat go The cat ate the fish.

Despite the narrower scope of use of the pretransitive construction in the Jieyang dialect, it is nevertheless much richer in its syntactic forms, resulting from syntactic stratification. Altogether, there are six markers which can all serve to mark the preposed object NP and can be said to be functionally similar to the BA morpheme in the Mandarin BA construction. These markers are tsiaŋ33, pa53-35tsiaŋ33, t’aŋ213 , tui213, kai35-21-i33, which appears after the preposed object NP, all other markers appear before the object NP (see 7.3 for discussion on their similarities and differences). Furthermore, the pre-object and post-object markers can also combine to form a hybridized construction. Thus, there are three structures of the pretransitive construction, which are shown in Table (7.1).

Table (7.1) Pretransitive Constructions in the Jieyang Dialect

(a.) Subj. + tsiaŋ33 /pa53 / pa53-35tsiaŋ33/ t’aŋ213 /tui213 + Obj.NP + VP
(b.) (Subj.) + Obj.NP + kai55-11-i33 + VP
(c.) Subj. + tsiaŋ33/t’aŋ213/tui213 + Obj.NP + kai55-11-i33 + VP

A can be seen, the first structure is very similar to the disposal type of BA sentence. Pattern (b) has the marker occurring after the object NP, which is in the topic position. The third pattern combines the first and the second, which is not very commonly found in other Sinitic languages. It will be shown that the second and the third structures may come from a Southern Min source. It will also be shown that the six syntactic markers are not altogether free variants. They are distinguished either by register, literary vs. spoken, or based on semantic grounds such as high or low degree of transitivity, as well as on syntactic grounds, such as their position in the sentence (as shown in the above table). In the following sections, I first describe the semantic functions of the pretransitive construction (7.2) before explaining the differences and similarities of the syntactic forms in relation to the semantic properties (7.3). Two other features associated with the pretransitive construction, namely, the definiteness of the object NP (7.3.1) and the complexity of the VP (7.3.2), will also be touched upon. It will also be shown that the pretransitive construction can co-occur with the passive. This is typical of cases when only part of an entity is being affected (7.4).

7.2 The semantic functions of the pre-transitive construction 7.2.1 Resultative type of pre-transitive sentence
7.2.1.1 Semantically transitive verbs
7.2.1.2 The adversity semantics
7.2.2 The telic type of pre-transitive sentence

7.3 Syntactic makers of the pre-transitive constructions
7.3.1 tsiaŋ33, pa53 and pa53-35tsiaŋ33
7.3.2 t’aŋ213
7.3.3 tui213
7.3.4 kai55-11-i33
7.3.4.1 The lexical and grammatical usage of kai55-11-i33
7.3.4.2 The functions of kai55-11-i33 pre-transitive sentences
7.3.5 The co-occurrence of pre-transitive markers

7.4 Other grammatical features: definite NP and complex VP
7.4.1 Definiteness of the object NP
7.4.2 Complexity of the VP

7.5 The pre-transitive construction and the passive

Summary of chapter

The pretransitive construction resembles to some extent the much studied BA construction in Mandarin both semantically and syntactically. However, the pretransitive construction is by no means an equivalent of the BA-construction. Its range of functions is much more limited, which is evidenced by the absence of an equivalent of the causative or intransitive type of BA sentences in the Jieyang dialect and in the fact that many transitive type of BA sentences cannot be replicated in the pretransitive construction if the sentences do not denote physical or intrinsic affectedness of the thematic undergoer. The Jieyang data also show that OV topic –comment order without overt marking is quite common, which also accounts for the more limited use of the pretransitive construction in the Jieyang dialect. Based on semantic functions, the pretransitive construction can also be subdivided into two subclasses. While both types are about an agentive action carried out with regard to a known entity, the first type encodes a change of state on the part of the undergoer brought about by the agentive action. This change of state can either by physical or in terms of change of location. The second type does not encode this meaning but rather the potential achievement of a goal. This meaning derives from the telic situations which occur in the second type. Except for sentences which denote displacement in the metaphorical sense, the direct object of a pretransitive sentence must represent a concrete and physical entity, which must be syntactically realized as a definite NP (except for a generic NP). As with the BA construction, it has also been shown that the pretransitive construction requires a complex VP in the form of post verbal modification and other elements such as locative expressions, aspectual markers and purposive clauses to indicate resultativity and perfectivity. A unique feature associated with the pretransitive construction in the Jieyang dialect is the rich repertoire of markers, six in all, resulting from syntactic stratification: tsiaŋ33 and pa53 are cognate with jiang and ba from Mandarin and come from the literary stratum. They are mainly used in the formal and written register. The existence of a hybridized form pa53-35tsiaŋ33 is typologically very interesting, as it is not commonly found in other Sinitic languages. It may represent a stage after pa53 was introduced to the dialect before it became fully independent. The local forms t’aŋ213, tui213 and kai55-11-i33 on the other hand can only be used in the spoken language. Among them, further division of labor is observed: t’aŋ213 can only be used in the resultative type of pretransitive sentence. In particular, it encodes physical and intrinsic affectedness. Tui213 (even though not a common marker in the Jieyang dialect) tends to denote ‘intentional action’, whereas kai55-11-i33 can be used in both types of pretransitive sentence and in particular, in imperative sentences. Finally, it has also been shown that the source meanings of some of the pretransitive markers also influence their compatibility with certain types of pretransitive sentences. For example, tsiaŋ33 and pa53-35tsiaŋ33, having derived from an instrumental usage of ‘take hold of [object]’ in a serial verb construction in classical Chinese, are most commonly used when displacement, especially in the metaphorical sense, is involved, whereas tui213, with its lexical function as a preposition ‘to, towards’, is used when direction of an action or intention of an action is implied.

Abstract 摘要

8.1 Introduction
Passive constructions are widespread in the world languages (Keenan 1985:243). Functionally speaking, a passive construction serves to ‘topicalize’ the patient argument which is not normally in the topical position in the active. A passive sentence such as ‘John was slapped’ in English represents what is called the ‘basic’ passive in Keenan and is said to be the most common type cross-linguistically (1985: 247). Characteristic of basic passives is that the main verb is transitive. It represents an activity, taking agent subjects and patient objects, but the agent phrase is not present. However, the agent phrase can also be present, but is relegated to an oblique status (ibid: 250). Passives of non-transitive verbs are also found in some languages in the world (see Keenan 1985: 272-280). In this chapter, I describe two constructions in the Jieyang dialect which can be called passives (following the functional definition of Keenan given above) and one which has passive morphosyntax but is built on intransitive verbs (see also Matthews, Xu and Yip, 2005). The syntactic structures of these constructions can be represented respectively in the following three schemata:

(i) Patient NP + k’eʔ2 + Agent NP + VPvt
(ii) Patient NP1 + k’eʔ2 + Agent NP + VPvt + Patient NP2
(iii) Subject NP + k’eʔ2 +i33 + VPvi

Schema (i) represents a construction which will be called the ‘long passive’, following terminology used for Mandarin grammar. According to James Huang opposed to short passive where the agent phrase is not present. The Jieyang dialect only exhibits the long passive (see 8.3). It is based on transitive verbs, taking an agent and a patient argument, and the passive characteristic is analytically marked by the morpheme k’eʔ2. Schema (ii) has two patient arguments, one preposed to preverbal position while the other retained in postverbal position. I call this construction ‘indirect passive’ (8.4), a term used in Mandarin grammar, but which is in turn borrowed from the literature on passives in Japanese (James Huang 1999:34) where analogous constructions exist. The two NPs in the indirect passive typically represent a part-whole or possessor-possessed relationship. The third pattern is typologically interesting in that although it bears morphosyntactic similarity to the first pattern, the long passive, it is nevertheless based on intransitive verbs. Paradoxically, the slot after the marker k’eʔ2, which in the prototypical passive would be the agent slot, is filled by an invariable pronominal i33. It will be shown in 8.5 that this construction is not passive in meaning but the marker has become overt marking of unaccusativity, encoding a change of state, typically an adverse change of state. This construction differs syntactically and semantically from ‘impersonal passives’ found in some Indo-European languages, which also derive from intransitive verbs (8.5.4). For lack of a better term and parallel constructions cross-linguistically, I will call the construction represented in schema (iii) above the k’eʔ2-i33 construction.

8.2 The marker k’eʔ2

8.3 The long passive construction
8.3.1 The agent requirement
8.3.2 The adverse connotation
8.3.3 Lexical strategy: tek2kau213 and siu35

8.4 The indirect passive construction

8.5 The k’eʔ2i33 construction
8.5.1 Unaccusative predicates and the syntactic role of the two arguments
8.5.2 The formation of unaccusative predicates in the Jieyang dialect
8.5.3 Other semantic properties of the k’eʔ2i33 construction
8.5.4 The k’eʔ2i33 construction in discourse
8.5.5 The k’eʔ2i33 construction and the ‘impersonal passive’

Summary of chapter
This chapter describes two passive construction and the k’eʔ2-i33 construction which has similar syntactic frame with the passive in the Jieyang dialect. The two passives are the long passive and the indirect passive, both of which are based on transitive verbs, taking agent and patient arguments and that the passive character is overtly marked by k’eʔ2, which appears before the agent NP. Like other Southern dialects such as Cantonese or Taiwanese Southern Min, the Jieyang dialect does not have agentless passives, or short passives. The indirect passive is characterized as having retained two patient NPs, one in topic position and one after the VP, whose relationships are typically those of whole-part and possessor-possessed. The passive marker k’eʔ2 in the Jieyang dialect has grammaticalised from a lexical verb ‘to give’. It is also a marker for the dative and the causative. The argument structure of the source verb ‘give’ seems to be retained in its grammaticalised use as a passive marker, as is evidenced from the absence of short passive in the Jieyang dialect. This ‘ agent requirement’ feature also accounts for the fact that the third construction covered in this chapter, the non-transitive k’eʔ2-i33 construction, also has two overt NP arguments. The k’eʔ2-i33 construction is shown to be typologically unusual in that it allows agentless and non-volitional unaccusative predicates to appear in a morphosyntactic form parallel to the passive. It has been shown that this construction is not passive in meaning. Rather, it encodes that the subject NP, which is the thematic theme or patient, undergoes a change of state. The morphemes k’eʔ2-i33 have thus been analyzed as an overt marker for unaccusativity in the Jieyang dialect. This phenomenon of using a passive marker to indicate unaccusativity is not common within Sinitic languages, but cross-linguistically, it is pointed out by Levin and Rappaport (1987) that there are syntactic and morphological phenomena that class unaccusative verbs and passive verbs together (p.11). Although the k’eʔ2-i33 construction does not express the passive meaning, it is interesting to note that it exhibits semantic properties very similar to real passives. In particular, like the passives, it carries a strong sense of adversity. Furthermore, the k’eʔ2-i33 construction also implies that the adverse change of state is caused by some circumstances or a prior event. It is these properties which make the k’eʔ2-i33 construction distinct from other syntactic devices in the Jieyang dialect which also encode the inception of a new state.

Abstract 摘要

9.1 Introduction
As in many other languages in the world, negation in the Jieyang dialect can be divided into two types: clausal and constituent, the distinction being one of scope. According to Trask (1993), scope refers to the portion of a particular sentence ‘which is interpreted as being affected by a certain operator… such as a negative’ (p.248) (note: the bold font is in the original). In clausal negation, the negative grammatical element takes an entire clause in its scope and asserts ‘that some event, situation state of affairs does not hold’ (Payne 1997:282). Constituent negation, on the other hand, takes only certain constituents of the clause under the scope of the negative grammatical element.

9.2 Modal auxiliaries and their negative counterparts

9.3 Negative markers- an overview

9.4 Types of negation
9.4.1 Lexical negation
9.4.2 Clausal negation
9.4.2.1 m351 in negation copula clauses
9.4.2.2 m351 negating stative predicates
9.4.2.3 boi35 and m351 negating adjectives
9.4.2.4 The functions of bo55
9.4.2.5 bue11 indicating that an event has not yet occurred
9.4.2.6 Negating verbal complements
9.4.3 Indefinite negation
9.4.4 Negating modal meanings
9.4.4.1 Negating epistemic modality
9.4.4.2 Negating deontic modality
9.4.4.3 Negating futurity and volition
9.4.4.4 Summary of section

Saummary of the chapter
In this chapter, I have discussed five types of negation: lexical negation, clausal negation, negation of postverbal complements, indefinite negation and negation of modal meanings. What has been shown in this chapter shows that the Jieyang dialect has a relatively complex negation system which is reflected in several aspect: first of all in its large array of negative markers. Among them, the majority are fused forms which are contracted from a negative element and a modal auxiliary verb. Thus, negation is closely related to modality. For example, gradable adjectives, whose characteristics or semantics are ‘relative’ to subjective views, are negated by boi35 which expresses epistemic modality. Related to the first feature is the wide diversification of functions of the negators. As can be seen, the negators are task-oriented in the sense that different negators are called for different functions and purposes, which has been shown to closely interact with various factors. Apart from modality, as mentioned above, aspectual properties, temporal references and types of predicates also play an important role in the choice of negators. This clear division of labour, which is a feature shared by other Southern Min dialects such as Taiwanese, is not surprising because, as L.L. Chen (1991:1) puts it, ‘the grater the number of contradistinctive negatives in a language, the more specialized each negative is in its function’.

Abstract 摘要
10.1 Introduction
In this chapter, I discuss three basic types of interrogatives in the Jieyang dialect: Wh-questions (10.2.), Yes-No questions (10.3) and Alternative or Choice questions (10.4). It will be shown that syntactically, while Wh-questions and Alternative questions show close similarities to other Sinitic languages such as Mandarin, the second type, that is, Yes-No questions, exhibits some features which are more characteristically Southern Min. The most noteworthy is the close interaction between question and negation. As discussed in Chapter 9, negation in turn is closely tied to modality and aspect.

10.2 Wh-questions
10.2.1 Question words in questions
10.2.1.1 ti11 – ‘which’
10.2.1.2 ti11tiaŋ55 – ‘who’
10.2.1.3 ti11tiaŋ33si55 ‘when’ kui53-35 tiam53/zioʔ5-2ts11tiam53 ‘what time’
10.2.1.4 ti11ko213 – ‘where’
10.2.1.5 meʔ2kai55 – ‘what’
10.2.1.6 tso213-53ni55 ‘why’/’how’
10.2.1.7 zioʔ5-2tsoi11/kui53 ‘how many/much’
10.2.2 Question words as indefinite pronouns

10.3 Yes-No questions
10.3.1 Neutral (non-presumptive) questions
10.3.1.1 V-Not-Vp
10.3.1.2 [VP-Neg] and {VP-Dis
10.3.1.3 Aux-Neg-VP
10.3.1.4 VP-QPRT
10.3.1.5 Responses to neutral questions
10.3.2 Presumptive questions

10.4 Choice questions

Summary of chapter
In this chapter, I have discussed three basic types of interrogative in the Jieyang dialect: Wh-questions, Yes-No questions and Choice questions. It has been shown that one of the major features of Wh-questions in the Jieyang dialect, which is also a common feature in other Sinitic languages, is the position of the question words: they are not placed at the beginning of the sentence like English, but remain in situ as in a declarative sentence. This means that their position in sentence depends on what syntactic role they play. Question words can also serve as indefinite pronouns, expressing such notions as ‘whoever’, ‘whatever’, ‘anything’, ‘everything’, etc. The Yes-No questions further distinguish two subtypes: Neutral and Presumptive questions. Neutral questions have been shown to share certain characteristics that are pan-Sinitic, such as employing the technique of Juxtaposing the affirmative and the negative of the predicate as a way of posing a question in the Jieyang dialect also display some features which are characteristically Southern Min, such as using negative modal auxiliaries as question particles in the [VP-(DIS)-Neg] form, which has a history dating back to Archaic Chinese. As a result of using negative modal verbs as sentence final question particles, it has been shown that there is a close relation between interrogatives, negation, modality and aspect. The Choice questions are shown to be different from the neutral question patterns of disjunctive nature in that the order of the disjoined constituents in a Choice question is flexible but that of the neutral question forms is fixed.

Abstract 摘要
In this chapter, I describe constructions of comparison which include comparatives and the superlative. The syntactic structures and the markers used the comparatives are one of the areas which are known to have a Northern versus Southern distinction (see Yue-Hashimoto 1993: 158; Ansaldo 1999). However, due to the increasing influence from Mandarin, many Southern dialects have come to use a pattern borrowed from Mandarin as a secondary choice (Ansaldo 1999: 103). This observation can be applied to the Jieyang dialect, which not only has two forms from the local stratum but also a loan stratum. Furthermore, the form modeled on the Northern pattern has become increasingly productive, reflected in its functional versatility. In the following sections, I fist look at prototypical comparatives, followed by a brief description of other schemes of comparative constructions. The superlative constructions will be discussed last.

11.1 Prototypical comparatives
11.2 Other schemes of comparison
11.3 The superlative degree

Summary of chapter
This chapter has described the prototypical schemes of comparison involving two participants and one property as well as other more complex schemes of comparison involving one participant in terms of two properties. It also describes the superlative constructions. Four degrees of comparison in relation to the prototypical comparatives are identified in the Jieyang dialect. They are the comparative constructions of superiority (CCS), inferiority (CCI), equaling (CCEQL) and equality (CCEQ). Semantically, the CCS is the contrary to the CCI, but syntactically, the CCI is the negative counterpart of the CCEQL. The CCEQL is a unique category in the Jieyang dialect (and Sinitic languages as a whole) in that it expresses approximation, which can be translated as ‘X is at least as … as Y’. The CCEQL is thus semantically distinct from the CCEQ which expresses equality and identical properties. In other words, the CCEQ is about whether or not the two participants are the same or identical or what identical feature or quality the two compared entities share. In terms of syntactic means to express the comparative constructions of superiority, it has been shown that the Jieyang dialect not only has close affinity with other Southern dialects in having the Surpass type, but also has preserved the very ancient form of the comparative: the Absent-marking construction. The CCS in the Jieyang dialect is also expressed through the Similarity construction, which is a cognate form with Mandarin. The co-existence of the three forms is a clear indication of syntactic stratification, with the local forms being used for syntactically simple(r) structures while the borrowed form, the pi53 construction being the most versatile in terms of function. This might reflect a trend similar to those observed in other dialects where the Northern pattern is playing an increasingly prominent role (see Yue Hashimoto 1993: 160; Li and Lien 1994:85). Comparison of one participant and two properties cannot be expressed using the prototypical comparative structures. In many cases, the sentences have to be stated in terms of comparing subjects or topics or through bi-clausal constructions. The comparative constructions have been shown to interact with negation (which is also observed in the interrogative constructions. See Chapter 10). This is reflected in the different negators used on whether ungradable or gradable adjectives as PARA are involved. The superlative degree is the most akin to other Sinitic languages in terms of its syntactic structures. Although it uses a cognate form with Mandarin as the marker, there exist several colloquial words which also serve as markers for the superlative.

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
zh_HKZH-HK