Chapter 1 Introduction
1.2 Organization of the thesis
國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
markers now in Mandarin conversations (Tsai, 2001; Wang & Tsai, 2005; Wang et al., 2010; Yu, 2004; Zhang, 2006). This thesis therefore intended to investigate how Mandarin-speaking children manipulate the various functions of the two frequently appearing discourse markers – hao ‘okay’ and dui ‘right’ – especially when interacting with peers, since peers, as suggested in earlier studies, are influential in children’s development. Five-year-old children were chosen in the study as our subjects for they have developed the preliminary ability to use some discourse markers (Huang, 2000). Also, children at this age are usually considered to have fairly more chances than the younger ones to engage in peer interactions. The purpose of this thesis is to investigate how Mandarin-speaking children use the discourse markers hao ‘okay’ and dui ‘right’ in peer interactions in order to examine their communicative skills and moreover to see whether it reflects the characteristics of peer relation. The research questions are as followings.
(1) How do Mandarin-speaking children use the discourse marker hao in different discourse structures while interacting with peers?
(2) How do Mandarin-speaking children use the discourse marker dui in different discourse structures while interacting with peers?
(3) How do Mandarin-speaking children’s uses of the discourse markers hao and dui reflect the characteristics of peer interaction?
1.2 Organization of the thesis
The reminder of the thesis is laid out as follows. Chapter 2 consists the literature review of previous studies, which include the definition of discourse markers, discourse coherence, and earlier research about the multifunction of hao and dui in
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
Mandarin. Cross-linguistic studies about children’s use of discourse markers are included in chapter 2 as well. Chapter 3 describes the methodology and analytical framework of the present study. The results are presented in chapter 4. Finally, discussions and conclusions are made in chapter 5 and 6 respectively.
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
5 Chapter 2
Literature review
Cross-linguistic studies have revealed the dynamic functions of discourse markers in conversations, especially those related to discourse coherence (Fraser, 1990; Halliday, 1994; Schiffrin, 1987; Wang et al., 2010). Some of the studies investigated children’s acquisition of different markers (Andersen et al., 1999;
Kyratzis, 2005; Kyratzis & Ervin-Tripp, 1999; Sprott, 1992). In this chapter, related studies of discourse markers are reviewed and presented in the following order. In section 2.1 and 2.2, the existing literature on discourse markers and discourse coherence is reviewed. The studies about discourse markers hao ‘okay’ and dui ‘right’
in Mandarin Chinese and those concerning children’s acquisition of discourse markers are illustrated in section 2.3 and 2.4 respectively. Finally, in section 2.5, earlier research on the roles of peers in children’s development is discussed.
2.1 Discourse markers
Discourse markers have been widely investigated by many researchers in the past decades. Different but similar definitions have been provided (Fraser, 1990; Lenk, 1998; Schiffrin, 1987). Among them, the most well known definition was the one given by Schiffrin (1987). According to Schiffrin, discourse markers are “linguistic, paralinguistic, or non-verbal elements that signal relations between units of talk by virtue of their syntactic and semantic properties and by virtue of their sequential relations as initial or terminal brackets demarcating discourse units” (p. 40).
Discourse markers are used by speakers to make the relations between the previous
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
and ongoing texts salient in order to build the coherence of discourse units. That is, they have a coherence building function in discourse.
Schiffrin suggested that in addition to the semantic content about the real world, discourse markers also mark relations at other levels of discourse. She proposed a framework consisting of five components of discourse organization, which are the ideational structure, the action structure, the exchange structure, the participation framework and the information state. She claimed that discourse markers select and display both the meaning (i.e. contrastive, resultative etc.) and structural relations (i.e.
which utterance is a main/subordinate/coordinate unit) between the preceding and following utterances. Discourse markers therefore function as the “contextual coordinates” of an utterance at one or more levels of discourse organization. That is, they index an utterance to a certain plane of discourse in which it is produced and is to be interpreted. She argued that it is not the marker itself conveys social and expressive meanings but the discourse slots it situated. For instance, but itself does not have the meaning of ‘challenge’, however, in some contexts the utterance following it may be interpreted as a challenge. Schiffrin also identified several common features that qualify an expression to function as a discourse marker. These features could also be viewed as the characteristics of discourse markers. The features are listed as following.
It has to be syntactically detachable from a sentence.
It has to be commonly used in initial position of an utterance It has to have a range of prosodic contours
E.g. tonic stress and followed by a pause, phonological reduction
It has to be able to operate at both local and global levels of discourse, and on different planes of discourse.
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
This means that it either has to have no meaning, a vague meaning, or to be reflexive (of the language, of the speaker).
(Schiffrin, 1987, p. 328) In addition to Schiffrin, Fraser (1990) also defined a discourse marker as “a pragmatic marker which provides a commentary on the following utterance; that is, it leads off an utterance and indicates how the speaker intends its basic message to relate to the prior discourse.” In both Schiffrin’s (1987) and Fraser’s (1990) analyses, they discussed mainly about the relations between adjacent units at the local coherence level in discourse. Lenk (1998) further pointed out that discourse markers signal relations between larger discourse segments such as topics, situations or background and foreground knowledge, as well as smaller segments like adjacent units. In other words, except the local level, discourse markers can be used to establish coherence at the global level of discourse. Briefly speaking, the function of discourse markers, as previous scholars suggested, is to identify relations between segments in the context, both locally and globally, and thereby construct the coherence of the whole discourse (Fraser, 1990; Lenk, 1998; Shiffrin, 1987; Wang, Tsai, & Lin, 2007).
2.2 Discourse coherence
Because discourse markers serve the coherence-building function in discourse, earlier researchers have proposed various frameworks for explaining planes/levels of discourse (Halliday, 1994; Miracle, 1991; Schiffrin, 1987). Schiffrin (1987) established a framework of discourse coherence with five components, which includes the action structure, the exchange structure, the ideational structure, the participation framework and the information state. She suggested that discourse
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
markers signal meanings at the exchange structure, the action structure and participation framework as well as the ideational structure. The details of the five structures are illustrated below.
The action structure refers to a structure where speech acts are situated. Actions are considered as situated because they do not appear randomly. They occur in a constrained linear order. The action structure accounts for the order of actions, that is, what actions tend to precede, what are intended to follow and what actually follows, and also for the speakers’ identities and social setting. It is a structure about “ritual requirements of talk” (Goffman, 1981a), which concerns with “the management of oneself and others so as not to violate appropriate standards regarding either one’s own demeanor or deference for another” (Schiffrin, 1987). English discourse markers such as oh, well, and, but, so, and because, all serve their meanings in this structure.
In example (1), because functions at the action structure to modify or support a speech act which is a request. According to Schiffrin, causal connectives operate in the action structure serves the function as justification of various speech acts. In Redeker’s (1990) analysis, she named action level uses “pragmatic” uses.
(1) Mei (4;03): Can I have that daddo? Because I like him. (Kyratzis et al., 1990) The exchange structure is where the speaker establishes “conditionally relevant adjacent-pair parts”. Relevant adjacency pairs, for example, question-answer, request-compliance and greeting-greeting, are all built in the exchange structure. According to Schiffrin’s definition, this structure is “the outcome of the decision procedures by which speakers alternate sequential roles and define those alternations in relation to each other.” It parallels to that in Goffman’s (1981a) analysis named the “system constraints of talk”, which are the constraints for interlocutors to give appropriate and interpretable feedbacks to each other. In English, discourse markers such as well, and,
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
but, or, so, and you know all operate in this structure. Example (2) illustrates the use of because in young children’s discourse to introduce turns in the exchange structure.
Because in the example carries no meaning of events or speech acts.
(2) A: It’s mine [reaching for the toy that B is holding].
B: Because it’s mine [keeping toy out of A’s reach]. (Sprott, 1992)
There is also a structure concerning the organization of ideas within the discourse, which is named, in Schiffrin’s term, the ideational structure. In contrast to the action structure and the exchange structure, which are pragmatic, the relations in the ideational structure are semantic. The ideational structure consists of three types of relations, which are cohesive relations, topic relations and functional relations. The cohesive relations (coherence) of discourse are built when “interpretation of an element in one clause presupposes information from a prior clause because of the semantic relationship underlying a text” (Schiffrin, 1987, p. 26). The organization of topics and subtopics is also a part of the ideational structure. The other part of the ideational structure includes the functional relations between ideas. That is, the roles ideas play in relation to the others and within the whole text. English markers and, but, or, so, and because all function in this structure. In example (3), the connective because signals the causal relation between the two events in the real world. It functions as a semantic marker linking the propositional content of two clauses in the ideational structure of discourse.
(3) I sprained my ankle ‘cause I was hitting my father’s shoe. (7;05. Kyratzis et al., 1990)
The next plane of discourse in Schiffrin’s model is the participation framework. It is a term introduced by Goffman (1981b). The participation framework includes relations between both the speaker and the hearer, and the speaker and his utterances.
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
Goffman (1981b) suggested that there are various levels of identity a “speaker” and a
“hearer” could have through talk. He differentiated the term “speaker” for different
“production formats”, such as the speaker who presents talk (an animator) from the one who is presented through talk (a figure), and the one who takes responsibility for the content and implication of talk (a principal). The term “hearer” is also categorized for various “reception formats”, which includes addressees and overhearers (the intended or unintended recipients), passive participants or the expected contributors of talk. The relations between the speaker and the hearer also include institutional relations such as teacher/student and doctor/patient, and interpersonal differences of power and solidarity. Another aspect of the participation framework is how a speaker could relate to the utterances (the proposition, actions and turns) he produced. For example, speakers may evaluate, commit to or distance from their ideas; they may perform their actions indirectly or directly; or they may claim or relinquish their turns.
The way a speaker relates to his/her utterances influences the relation between the speaker and hearer. Discourse markers in the participation frameworks “serve to shift the roles in the talk of participants, or introduce upcoming talk as relating to concerns or questions raised in prior talk” (Schiffrin, 1987). This shift of roles can be seen as suggesting a new episode or phase in an exchange or activity. For instance, the marker well is often used to signal upcoming talk as contradicting expectations rose in prior utterances or questions.
The last component of discourse in Schiffrin’s model is the information state. The information state refers to interactions between speakers and hearers in their cognitive states. Namely, it focuses on the organization and management of both the speaker and hearer’s knowledge and meta-knowledge. In other words, both the speaker and hearer need to understand what they know and what other interlocutors know
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
respectively, and what information they share in order to give sufficient information for their hearers to interpret their messages. The interactions in the information state are interactive and dynamic processes that would change frequently throughout the conversation.
Similar to Schiffrin, Halliday (1994) proposed another model for explaining discourse coherence. He also assumed that language is multifunctional and metafunctional. In his model, he divided discourse into three levels: the ideational level, the textual level and the interactional level. The ideational level is where
“language serves to express logical and experiential meanings”. The textual level is where links between situations are made and where the cohesive texts are constructed.
The interactional level is where the interlocutors use languages as a set of actions on each other to build up relations between themselves and their addresses. It is where social relationships are established and maintained.
Generally speaking, previous researchers suggested that language is multifunctional. A linguistic unit, such as a discourse marker, may operate in more than one discourse structures at the same time. Theoretically it is possible for each discourse marker to function in all structures. However, in real world situations, the structures where a certain discourse marker could situate are restricted. It is due to the limited discourse slots one marker could locate. For one marker, some functions in certain discourse structures are more easily found than the others. For instance, in English, well, okay and now only function in the action structure, the exchange structure and the participation frameworks, but are seldom found in the ideational structure. Because serves as a marker in the action structure, the exchange structure and the ideational structure, but is hardly found in the participation frameworks (Kyratzis & Ervin-Tripp, 1999).
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
Moreover, previous studies suggested that discourse markers could be used to build up both the local and global relations (Lenk, 1998; Schiffrin, 1987). Local relations consist of those between adjacency pairs. According to Sacks and colleagues (1974), the organization of conversations is generally based on “adjacency pairs”.
Adjacency pairs are “consecutive, contingently related utterances produced by two different speakers” (Wang et al., 2007, p. 683). They can be both reciprocal like greeting-greeting, and non-reciprocal such as question-answer. An adjacency pair consists of two parts. The “first pair part” projects an adjacency pair can carry out various types of the “second pair part”. For example, a statement as the first part in an adjacency pair may be followed by another statement, an agreement, a refusal, a counter-statement, or nothing at all. Among the wide range of possible second pair parts, there are certain ones that are preferred by speakers than the others. These preferred second pair parts demonstrate the alignment between interlocutors. They are spoken out without any signals of hesitation and are usually made up of simple sentence structures. On the other hand, dispreferred second pair parts usually follow a pause or other hesitation markers, such as um. In addition, for being polite, they are usually initiated by signals of agreement or acceptance, an apology or appreciation and are usually with explanations (Levison, 1983; Pomerantz, 1975). Well, for instance, is one of the most common used discourse markers in adult conversations as the marker of dispreferred responses in adjacency pairs (Pomerantz, 1984). The relations of adjacency pairs construct the local structure of discourse. Moreover, discourse markers can also mark global relations. Global relations are the relations between larger chunks of discourse, topics and phases. Schiffrin (1987) and Polanyi and Scha (1983) have proposed that so in English could function at the global level of discourse to signal returns of digressed topics. It is, in their term, a “pop-marker”. In
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
other words, a global-level discourse marker is a boundary marker indicates changes of topics or episodes. In brief, according to earlier researchers, discourse markers are multifuntional. They may serve more than one functions at different discourse structures at the same time.
2.3 Discourse markers in Mandarin Chinese
In terms of discourse markers in Mandarin, some studies have focused on the use of hao ‘okay’ and dui ‘right’ in conversations. Lu (1980/2004) suggested that hao in Mandarin serves to express different moods. It is a marker of agreement, conclusion and counter-expression that resembles an interjection. His analysis captured the essential meaning of hao, however, it did not explain clearly the multifunction of hao in discourse. Miracle (1991) was the first researcher who analyzed the various functions of hao in spoken discourse based on four discourse levels. He observed that hao marks three types of relations: (1) the development and closure of social and physical actions, (2) the speaker’s assertion of previous utterance and (3) the transition to a new topic or social activity. Table 1 shows Miracle’s analysis of hao in different discourse levels.
Table 1. Discourse use of hao (Miracle, 1991, p. 56) Aspect of discourse Function (marking)
Social action structure Closure s-act (= social action) Closure physical act (s-act) Completion of subsidiary action
Turn structure Appreciation of assertion
Idea structure Completion of idea-transition
Information state Completion of internal deliberation
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
In the social action structure, hao is related to the development and closure of commissive or requestive actions. Hao also marks assertions in the turn structure. In the idea structure, it functions as a marker of idea management indicating transitions to a new topic or a social activity. According to Miracle, the core function of discourse marker hao in Mandarin is to mark closures and transitions.
Some researchers have analyzed the various functions of hao in terms of its grammaticalization process (Biq, 2004; Bybee, 1994; Haiman, 1994; Wang, 2004;
Wang, 2005). Biq (2004) investigated the use of hao in certain constructions conveying stances. She pointed out that hao in such constructions has been de-categorized from a lexical item (a stative verb) to a marker of mood and subjectivity.
Wang (2004) analyzed the various functions of hao and its directionality of different changes from both synchronic and diachronic perspectives. She claimed that hao has changed from a predicative adjective to both a discourse marker of agreement/acceptance and a marker of transition/closure. Bybee (1994) and Haiman (1994) have also come to the same conclusion of the grammaticalization process of hao. They further pointed out that the semantic change of hao from a predicative adjective to a discourse marker of agreement and a marker of closure went through the conventionalization of implicature and ritualization. Wang (2005) investigated the multifunction of hao in relation to its grammaticalization process. She argued that the grammaticalization process of hao went through three different paths. It developed from (1) a predicative adjective to a resultative complement then to a phase marker, (2) a predicative adjective to an intensifier, and (3) a predicate to a discourse marker. She pointed out that these meanings of hao are closely related to each other because they all derived from the primary predicative meaning. Hao in Mandarin originally meant
‘good’. During the grammaticalization process, it obtained the meaning of ‘very’,
• 國
立 政 治 大
㈻㊫學
•
N a tio na
l C h engchi U ni ve rs it y
‘satisfaction’, and ‘finishing doing something’, that is, hao lost some of its lexical meaning and became more grammatical.
Differing from the studies considering the grammaticalization process, Shao and Zhu (2005) analyzed the discourse functions of hao in Mandarin. They found out that hao serves three types of discourse functions, which are (1) the active answering function (appraisal, promise, and affirmation), (2) the passive answering function (acceptance of indirect refusal, concession, and irony), and (3) the discourse coherence function (discourse boundary marking, transition, and closure). They also discussed the functions of hao when it appears with particles such as a, ba and le as an independent unit (i.e. without combing with other discourse elements). Hao a signaled the speaker’s appraisal toward the previous utterance. Hao ba marked
Differing from the studies considering the grammaticalization process, Shao and Zhu (2005) analyzed the discourse functions of hao in Mandarin. They found out that hao serves three types of discourse functions, which are (1) the active answering function (appraisal, promise, and affirmation), (2) the passive answering function (acceptance of indirect refusal, concession, and irony), and (3) the discourse coherence function (discourse boundary marking, transition, and closure). They also discussed the functions of hao when it appears with particles such as a, ba and le as an independent unit (i.e. without combing with other discourse elements). Hao a signaled the speaker’s appraisal toward the previous utterance. Hao ba marked