漢語兒童「然後」的使用 - 政大學術集成

全文

(1)國立政治大學語言學研究所碩士論文 National Chengchi University Graduate Institute of Linguistics Master Thesis. 指導教授: 黃瓊之博士. 政治大. Advisor: Dr. Chiung-chih Huang. 立. ‧. ‧ 國. 學 sit. y. Nat. 漢語兒童「然後」的使用. io. n. al. er. The Use of Ranhou in Mandarin Child Language. Ch. engchi. i n U. 研究生：黃建銘撰 Student: Chien-ming Huang 中華民國 103 年 6 月 June, 2014. v.

(2) The Use of Ranhou in Mandarin Child Language. By Chien-ming Huang. 立. 政治大. ‧. ‧ 國. 學 er. io. sit. y. Nat. A Thesis Submitted to the. n. Institute of Linguistics aGraduate linCPartial Fulfillment of then i v hengchi U Requirements for the Degree of Master of Arts. June 2014.

(3) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. Copyright © 2014 Chien-ming Huang All Rights Reserved iii. i n U. v.

(4) Acknowledgements. 在我語言所碩士班的求學過程中，首先要感謝的是我的論文指導教授黃瓊之老師，在我困惑的時候指引我的方向，使我能克服瓶頸，順利完成論文。我也要感謝論文的口試委員尤雪瑛老師、張麗香老師，在忙碌之中撥空前來，給予我寶貴的建議。在所上的學習過程中，我要感謝所上的老師們教導我浩瀚的語言學知識。感謝黃瓊之老師、莫建清老師、蕭宇超老師、何萬順老師、徐佳慧老師、萬依萍老. 治政大姊在行政方面的協助，解答我的疑難雜症。立. 師、尤雪瑛老師、張郇慧老師、林佑瑜老師等的教導。另外，也感謝助教惠鈴學. 我要謝謝工作室的易儒學姊、侃彧學姊、美杏學姊，在我遇到問題時是我詢. ‧ 國. 學. 問的好對象；謝謝薏淩、亭伊幫忙一起錄語料；也謝謝惟珍接下我報帳的工作。. y. Nat. 加充實。. ‧. 我也很開心能認識所上的同學們，與你們一起學習與成長，使我的研究所生活更. er. io. sit. 此外，我非常感謝薏淩，一直在我身旁陪伴著我，與我走過校園的各個角落，跟我分享生活中的點點滴滴。在我沮喪、失落時，給予我安慰、支持與鼓勵；她. al. n. v i n 也是我努力、前進的動力。她在我寫論文時幫了我許多忙，在我論文口試時，她 Ch engchi U 也幫我安排許多事務，我非常感謝她！. 最後，感謝爸爸、媽媽尊重我的決定，支持我唸研究所。在我寫論文的這段期間，你們適時地給予我勉勵和督促，使我能順利完成學業，謝謝你們！我終於畢業了！. iv.

(5) Table of Contents Acknowledgements…………………………………………………………………..iv Table of Contents……………………………………………………………………..v List of Tables and Figures…………………………………………………………vi Chinese Abstract……………………………...……………………………………viii English Abstract……………………………...……………………………………...ix Chapter 1. Introduction……………………………..……………………………...1 1.1 Background and motivation………………………..……………………….1 1.2 Purpose……………………………………………..………………….……5 1.3 Organization of the study...………………………………….…………….5 Chapter 2. Literature Review……………………………………….……………...7 2.1 Connectives and discourse markers……………………………..………….7 2.2 Research on the acquisition of connectives and discourse markers…...…...9 2.3 Previous studies on ranhou……………..…………………………..…...….15 2.4 Research on the acquisition of ranhou………………..…..…………….…..18. 立. 政治大. ‧. ‧ 國. 學. Chapter 3. Methodology……………………………………..……………………20 3.1 Subjects and data…………………………………………..………………20 3.2 Analytical framework………………………………………..…………….22 3.2.1 Ranhou as the connective…………………………...…..………….23 3.2.2 Ranhou as the discourse marker……………………..…..………...28. n. al. er. io. sit. y. Nat. Chapter 4. Results……………………………………..…………………………..33 4.1 The distribution of ranhou…………………..…………………………….33 4.2 Ranhou as the connective use………………….…….……………………38 4.2.1 Temporal function…………….……….…..……………………….39 4.2.2 Additive function……………………..…….……………………...43 4.2.3 Causal function…………….……….……….……………………..48 4.2.4 Contrast function………………….….………..…………………...51 4.3 Ranhou as the discourse marker use……………………….………….…..54 4.3.1 Verbal filler function……………………….……………...…….…54 4.3.2 Topic succession function………….……….….………………..…59 4.3.3 Resumptive opener function………….……....………..………..…62. Ch. engchi. i n U. v. Chapter 5. Discussion and Conclusion……………….……………………...….66 5.1 The distribution of ranhou…………….……………..……………....66 5.2 Connective use of ranhou………….…………………..………….…68 5.3 Discourse marker use of ranhou……….…………………..………...71 5.4 Limitations and suggestions……….……………..……………..……73 Appendix. Transcription Conventions and Gloss Abbreviations…………….…..74 References………………………………………………………………………….75 v.

(6) List of Tables. Table 1 Subject Information in Group 1 and Group 2…………...…………..……….21 Table 2 The Frequency of Ranhou in the Two Age Groups……..………………........34 Table 3 The Distribution of Ranhou as Connective Use or Discourse Marker Use…37 Table 4 The Proportion of Different Functions of Ranhou as the Connective….........38 Table 5 The Tokens of Temporal Function of Ranhou for All Subjects in the Two Groups……………………………………………………………….....…...39 Table 6 The Tokens of Additive Function of Ranhou for All Subjects in the Two Groups.…………………………………………………………………...…44 Table 7 The Tokens of Causal Function of Ranhou for All Subjects in the Two Groups………………………………………………………………...…….48 Table 8 The Tokens of Contrast Function of Ranhou for All Subjects in the Two Groups……………………………………………...……………………….52 Table 9 The Proportion of Different Functions of Ranhou as the Discourse Marker…………………..…………………………………………………54 Table 10 The Tokens of Verbal Filler Function of Ranhou for All Subjects in the Two Group……..…...……………..…………………..………………………….55 Table 11 The Tokens of Topic Succession Function of Ranhou for All Subjects in the Two Groups………..……..…………..…………………...………….……60 Table 12 The Tokens of Resumptive Opener Function of Ranhou for All Subjects in the Two Groups……….……….…………………………………….…..…63. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. vi. i n U. v.

(7) List of Figures. Figure 1. The Framework of the Classification of Ranhou in This Study………….23. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. vii. i n U. v.

(8) 國立政治大學研究所碩士論文提要. 研究所別：語言學研究所論文名稱：漢語兒童「然後」的使用指導教授：黃瓊之博士研究生：黃建銘論文提要內容：. 治政大本研究旨在探討漢語母子對話中兒童「然後」的使用與其功能。研究語料來立. 自於八位年齡介於兩歲七個月到五歲四個月的兒童，並分成兩組：年紀較小之兒. ‧ 國. 學. 童(平均兩歲十一個月)與年紀較大之兒童(平均四歲十個月)。本研究主要根據 Su. ‧. (1998)的分類，把兒童「然後」分為連接詞(connective)及言談標記(discourse marker). y. Nat. 的使用。連接詞的使用有時間性(temporal)、遞增性(additive)、因果性(causal)及. er. io. sit. 轉折性(contrast)四項功能。言談標記的使用有填補語(verbal filler)、主題接續(topic succession)及主題再開(resumptive opener)三項功能。研究結果顯示年紀較大之組. al. n. v i n 的兒童使用「然後」的次數較多。隨著兒童年齡增長，言談標記使用的增加比連 Ch engchi U. 接詞的增加還多。「然後」當連接詞時，時間性功能使用得最多。與 Yeh (2011) 「然後」在敘述文中的研究比較，其顯示出「然後」在敘述文中出現得比在對話中還多。另外，時間性功能隨著年紀減少，而遞增性及因果性隨著年齡增加，可用 Winskel (2003)的「概念複雜性」(conceptual complexity)來解釋。「然後」當言談標記時，兒童填補語使用得最多，其隨著年紀增加而減少使用。主題再開兩組都使用得少，而主題接續只出現在年紀較大之組別。總結，兒童在「然後」的使用上有發展上的不同。. viii.

(9) Abstract. The study investigated Mandarin-speaking children’s different use of ranhou at different ages in mother-child conversations. Eight subjects aged from 2;7 to 5;4 were divided into two groups; a younger group (mean age: 2;11) and an older group (mean age: 4;10). In this study, the children’s use of ranhou was categorized into connective use and discourse marker use, mainly based on Su’s (1998) classification. The connective use of ranhou includes temporal, additive, causal, and contrast functions.. 政治大 resumptive opener functions. The results showed that older children used more tokens 立. The discourse marker use of ranhou contains verbal filler, topic succession, and. of ranhou than younger children. As children got older, the discourse marker use of. ‧ 國. 學. ranhou increased much more than the connective use of ranhou. In the connective use. ‧. of ranhou, temporal ranhou was the most frequent function. Compared with Yeh’s. sit. y. Nat. (2011) study of ranhou in narratives, it may be suggested that ranhou occurs more in. io. er. narratives and less in conversations. Additionally, the temporal function decreased with age whereas the additive and causal functions increased with age, which could be. al. n. v i n explained by Winskel’s (2003) C conceptual complexity.UIn the discourse marker use of hengchi. ranhou, the verbal filler function was used the most and decreased with age. The tokens of the resumptive opener function were few in both groups, and the topic succession function only occurred in the older group. To conclude, the results suggest that there is a developmental difference in children’s use of ranhou.. ix.

(10) 1. Chapter 1 Introduction. 1.1 Background and Motivation Ranhou was traditionally defined as a connective in Mandarin, which connected two related things in sequence to indicate that one thing is followed by the other (Lü,. 政治大. 1991; Wang et al., 1996). Connectives are defined as functional words to join. 立. linguistic units such as words, phrases, sentences, and paragraphs (Crystal, 2003).. ‧ 國. 學. They function as cohesive devices to indicate semantic relations between two or more. ‧. propositions (Halliday & Hasan, 1976). Since there are several semantic relations in. Nat. io. sit. y. connectives, from a developmental perspective, investigation of the acquisition of. al. er. connectives is important as it enables us to understand the multiple semantic relations. n. v i n in language acquisition (Bloom,CLahey, Lifter & Fiess, 1980). Besides, one of h e nHood, gchi U. the important issues of connectives in language acquisition is children’s choice of connectives as it explains the mapping of form and functions and the acquisition of multiple functions (Sprott, 1992). The acquisition of connectives is not an isolated process since it involves children’s overall cognitive and linguistic development. There are three components which are related to the process of the acquisition of connectives: children’s.

(11) 2. conceptual or cognitive development (Piaget, 1969), syntactic development (Slobin, 1973) and the linguistic mapping between concepts and words (Wagner, 1998). Each of these components may be used to explain certain characteristics of the process of connective acquisition. Children’s use of connectives starts at an early age. Bloom et al. (1980) investigated the use of connectives in English-speaking children and observed that. 政治大. children as young as two years old started to use connectives in their speech. It occurs. 立. either within a phrase or between two related utterances. In Mandarin, Hsu (2003). ‧ 國. 學. observed that Mandarin-speaking children started to use connectives such as yinwei,. ‧. suoyi, and ranhou at the age of 2 to 3. Children were able to use connectives like. Nat. io. sit. y. adults by the time they reach age 5 to 6. In addition, Bloom et al. (1980) explored the. al. er. acquisition order of English connectives in a longitudinal study of children between. n. v i n age 2;1 to 3;5. It was observed C thaththere was a developmental sequence in children’s engchi U. use of connectives. The earliest use of connectives was found to be additive, and then followed by temporal, causal, and adversative. As ranhou is used to connect two related things in sequence (Lü, 1991; Wang et al., 1996), it was regarded as a connective, indicating a temporal sequence between the two utterances. Nowadays, in addition to the connective use, ranhou is also referred to function as a discourse marker in spoken discourse of Modern Chinese (Su,.

(12) 3. 1998; Wang, 1996). The process, from connective use of ranhou to discourse marker use of ranhou, suggests that the use of ranhou may have evolved into a multifunction (Wang & Huang, 2006). Since grammaticalization generally refers to the linguistic process where lexical items become more and more grammatical (Hopper & Traugott, 1993), the use of ranhou from a connective to a discourse marker would imply that ranhou has undergone the process of grammaticalization.. 政治大. Ranhou in Mandarin serves various functions. According to Su (1998), the. 立. ideational use of ranhou, namely the connective use of ranhou, can be further divided. ‧ 國. 學. into three functions. These functions are temporal, consequence, and lists functions.. ‧. Besides, connectives in discourse often lose their primary semantic meaning in. Nat. io. sit. y. connecting two propositions and are instead used as discourse markers (Schiffrin,. n. al. use of ranhou, can also be. er. 1987). According to Su, the interactive use of ranhou, namely the discourse marker. v i n C further three subcategories. h e divided n g c hintoi U. These are. condition or concession, verbal filler and topic-succession functions. Since ranhou is a highly grammaticalized connective and serves many functions as a connective or a discourse marker, how children use ranhou would indicate the developmental trend of multiple functions of connective use and discourse marker use in Mandarin-speaking children. The emergence of ranhou can be examined by its form-function mapping and the order of acquisition of the semantic relations (Bloom.

(13) 4. et al., 1980; Slobin, 1973). Furthermore, for children, while it is true that ranhou has a semantic value, it also has a pragmatic value. Children’s discourse marker use of ranhou could achieve some pragmatic functions in narratives (Yeh, 2011). Some studies have investigated the use of ranhou in spoken discourse of adults (Shen, 1994; Su, 1998; Wang, 1996; Wang & Huang, 2006). However, few studies have been conducted on children’s acquisition of ranhou (Su, 1999; Yeh, 2011). Su. 政治大. (1999) investigated children’s production of all Mandarin connectives including. 立. ranhou, but did not further investigate the multiple functions of ranhou. Yeh (2011). ‧ 國. 學. investigated how 5-year-old and 9-year-old Mandarin-speaking children use ranhou. ‧. as temporal coherence in a story-telling narrative. However, the use of ranhou for. Nat. io. sit. y. younger children before the age of 5 was not investigated to understand their use of. al. er. ranhou at the early age and the following developmental trend. Besides, since Laubitz. n. v i n (1987) indicated that the use C of h connectives may be e n g c h i U affected by different speech contexts, the use of ranhou in conversations could be different from that in story-telling narratives. Hence the use of ranhou could be further investigated when used by children in mother-child conversations. To conclude, younger children’s use of ranhou could be further explored to understand children’s acquisition of connectives and discourse markers and its various functions in conversations..

(14) 5. 1.2 Purpose The purpose of the present study is to investigate Mandarin-speaking children’s use of ranhou in mother-child conversations in order to understand the developmental difference of the connective use and discourse marker of ranhou and its multiple functions. A younger group (2;7 ~ 3;2) and an older group (4;0 ~ 5;4) are compared.. 政治大. The three research questions are listed below.. 立. (1) How do children in the two age groups use ranhou differently in terms of. ‧ 國. 學. the frequency of ranhou?. ‧. (2) How do children in the two age groups use ranhou differently in terms of. Nat. io. sit. y. the connective use of ranhou?. n. al. er. (3) How do children in the two age groups use ranhou differently in terms of. Ch. engchi. the discourse marker use of ranhou?. i n U. v. 1.3 Organization of this study The remainder of the thesis is as follows. Chapter 2 consists of the literature review of connectives and discourse markers, research on the acquisition of connectives and discourse markers, previous studies on ranhou, and the research on the acquisition of ranhou. Chapter 3 describes the methodology and analytical.

(15) 6. framework of this study. Chapter 4 presents the results of the distribution of ranhou, different functions of ranhou as a connective, and different functions of ranhou as a discourse marker. Discussions and conclusions will be given in Chapter 5.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v.

(16) 7. Chapter 2 Literature Review. 2.1 Connectives and discourse markers. The definitions and classifications of connectives have been researched by several scholars (Bloom et al., 1980; Crystal, 2003; Halliday & Hasan, 1976).. 治政 Halliday and Hasan (1976) claimed that connectives are 大cohesive devices to link two 立 ‧ 國. 學. or more propositions indicating semantic relations. Crystal (2003) further defined connectives as functional words which could link linguistic units such as words,. ‧. phrases, clauses, sentences, and paragraphs. The use of the term “connective” or. sit. y. Nat. io. n. al. er. “conjunction” was quite similar, but it was suggested that “connective” has a broader. i n U. v. definition than “conjunction” because “connective” could refer to the semantic and. Ch. engchi. syntactic connection while “conjunction” mainly refers to syntactic connection (Rudolph, 1996). Generally speaking, according to the semantic relations of connectives, they can be classified into the four types (Bloom et al., 1980; Halliday & Hasan, 1976): 1. Additive connectives are used to link utterances that involve two independent events or states, such as “and,” “or” and “and also.” For example, “Tom has three sons and John has two sons.”.

(17) 8. 2. Temporal connectives are used to connect utterances that involve a dependency between events indicating sequence or simultaneity, such as “then,” “next” and “and then.” For example, “Mary brushed her teeth and then had breakfast.” 3. Causal connectives are used to link utterances that involve a dependency, most often a cause-and-effect relation, such as “because,” “so” and “consequently.” For example, “David was sick, so he didn’t go to school.”. 政治大. 4. Adversative connectives are used to conjoin utterances that show a contrastive. 立. but didn’t do well in the final exam.”. 學. ‧ 國. relationship, such as “but,” “however” and “yet.” For example, “Susan studied hard. ‧. Several studies have investigated the uses and functions of discourse markers. Nat. io. sit. y. (Brinton, 1996; Fraser, 1990; Lenk, 1998; Schiffrin, 1987). Schiffrin (1987, p.31). al. er. defined discourse markers as “sequentially dependent elements which bracket units of. n. v i n talk.” Besides, she pointed outC that h ediscourse h i U serve the function of adding n g c markers discourse coherence, and proposed the discourse model which is composed of five structures: the ideational structure, the action structure, the exchange structure, the information state and the participation framework. Differing from Schiffrin’s point of view, Fraser (1990, p.387) defined discourse markers as “a class of expressions, each of which signals how the speaker intends the basic message that follows to relate to the prior discourse.” Besides, discourse markers are thought to appear in interactions.

(18) 9. between different speakers to achieve coherence (Lenk, 1998). Lenk (1998, p.246) pointed out that discourse markers are “used by participants in conversation for establishing an understanding of coherence in conversation.” Based on the above characteristics of discourse markers, they are considered to be empty of lexical meaning, syntactically unconstrained, difficult or not necessary to translate, and hence optional. It seems that they do not have propositional meaning or. 政治大. grammatical functions. Nevertheless, Brinton (1996) indicated that discourse markers. 立. have the following functions:. ‧ 國. 學. 1. They mark various kinds of boundaries (to initiate or end a discourse or to shift a. ‧. topic).. Nat. 3. They help the speaker acquire or hold the floor.. n. al. Ch. engchi. 4. They repair one’s own or other’s discourse.. er. io. sit. y. 2. They denote either new information or old information.. i n U. v. 5. They assist in turn-taking in oral discourse.. 2.2 Research on the acquisition of connectives and discourse markers Some studies have found that children’s emergence of connectives may follow a developmental sequence (Bloom et al., 1980; Clancy, Jacobsen, & Silva, 1976). Bloom et al. (1980) investigated English-speaking children’s acquisition order of.

(19) 10. connectives by conducting a longitudinal study of four children aged from 2;1 and 3;5. The study centered on the developmental trend of the connectives children produced and the semantic relations they encoded. The various connectives used by these children were coded based on four different types of semantic relations, which are additive, casual, adversative and temporal. The acquisition order of connectives was found that additive was the earliest connective acquired, followed by temporal, causal,. 政治大. and adversative. The observation also indicated that children at around 2 years old. 立. phrase or two related utterances.. 學. ‧ 國. started to use connectives in their speech. They could use them either to connect a. ‧. The first connective observed to be used by young English-speaking children. Nat. io. sit. y. was “and.” In other languages such as French, German, and Turkish, the connectives. al. er. similar to English “and” in additive meaning were also the first acquired. In English. n. v i n C halso the connectiveUthat was used most frequently the additive connective “and” was engchi. (Bloom et al., 1980; Clancy, Jacobsen, & Silva, 1976). Although the core meaning of “and” is to connect two events to mark an additive relation, “and” was also used to encode all temporal, causal, and adversative semantic relations and regarded as a multifunctional connective. Therefore Bloom et al. (1980) and Clancy et al. (1976) pointed out that the multifunctional “and” could be used for expressing various semantic relations even when there are more appropriate or precise connectives..

(20) 11. The acquisition order of connectives may be affected by the different degree of conceptual complexity (Brown, 1973; Clark & Clark, 1977; Winskel, 2003), which means that complex relations are acquired later than simple ones. Clark and Clark (1977, p. 338) claimed that conceptual complexity, or cognitive complexity, “sets the pace for acquisition, at least in part.” Winskel (2003) reported that in English, the connective “then” was relatively acquired earlier than the connective “since.” One of. 政治大. the major possible reasons was involved with their conceptual complexity in that. 立. “then” referred to “sequentiality” while “since” referred to “sequentiality plus. ‧ 國. 學. duration.” In addition, Spooren (1997) investigated the developmental trend of Dutch. ‧. connectives in children aged 6 to 12 by focusing on their ability to mark explicit. Nat. io. sit. y. relations (the lexical form fully corresponding with the semantic meaning). It was. n. al. connectives to mark. er. found that compared with younger children, older children were more able to use. i n C explicit h semantic relationsUand engchi. v. use connectives more. appropriately. The studies on the acquisition of connectives in Mandarin Chinese are relatively few. Hsu (2003) in his study of connectives observed that children at the age of 2 to 3 started to use connectives such as yinwei, suoyi, and ranhou, and children at the age of 5 to 6 were able to use connectives appropriately like adults do. Wu, Zhu and Miao (1986) investigated Mandarin-speaking children’s development of connectives. It was.

(21) 12. reported in their study that the 2;6 year-old children started to use connectives during free play and story-telling. Children’s development of connectives was not only revealed in the frequency of their occurrence, but also presented in the variety and complexity of the connectives. Miao and Zhu (1989), instead of production, tested the comprehension of compound sentences which included seven types of connectives in children aged 4 to 6. The results showed that most 4-year-old children were more. 政治大. capable of understanding instructions with additive connectives than other kinds of. 立. connectives.. ‧ 國. 學. Su (1999) made a comprehensive cross-sectional study of Mandarin-speaking. ‧. children’s acquisition of connectives in which 26 subjects from age 3;5 to 5;11. Nat. io. sit. y. participated. The study explored children’s use of connectives in syntax, semantics,. al. v i n C hConcerning semantics, adults. e n g c h i U the. n compared with those of. er. pragmatics and discourse organization. The use of connectives by children was also study indicated that. younger children used connectives to mark underspecified (general) relations more often than older children. Concerning pragmatics, the results suggested that children used connectives to signal a pragmatic relation more often than a semantic relation. Concerning discourse organization, the study showed that children used connectives both locally (connecting adjacent utterances) and globally (referring to nonadjacent utterances). Furthermore, the results also indicated that the contextual effect may.

(22) 13. affect children’s use of connectives. In the context of narratives, temporal connectives occurred more than other types of connectives, while in the context of conversations; additives were used more often than other types of connectives. Children’s acquisition of discourse markers have been studied by several researchers (Andersen et al., 1999; Kyratzis, 2005; Kyratzis & Evrin-Tripp, 1999; Sprott, 1992). Sprott (1992) investigated discourse markers in verbal disputes in. 政治大. English-speaking children aged from 2;7 to 9;6. A total of five discourse markers were. 立. examined in terms of their discourse functions: “because,” “so,” “and,” “but,” and. ‧ 國. 學. “well.” The results showed that children use these discourse markers differently at. ‧. different ages. Younger children organized discourse coherence mainly at the local. Nat. io. sit. y. level. Older children started to construct the relations at the global level. Therefore it. al. er. was claimed that as children get older, discourse markers are used to express more. n. v i n functions, or they may lose theirCmain U other functions. Kyratzis and h efunctions n g c hto iserve Evrin-Tripp (1999) conducted a research on English-speaking children’s acquisition. of discourse markers in peer conversations. Discourse markers “because,” “so,” “well,” “ok,” and “now” were identified and coded based on Shiffrin’s discourse model (1987). The results showed that the factors of age and context may affect children’s use of discourse markers. Older children used more global relations than younger children. And different contexts, such as doll play or story re-telling, may prompt.

(23) 14. different discourse markers. Some studies researched into children’s use of discourse markers in a sociolinguistic point of view (Andersen et al., 1999; Kyratzis, 2005). Andersen et al. (1999) investigated the acquisition of discourse markers in a cross-linguistic study and explored how these markers serve as markers for social relation. The study compared findings in American English, French, and Spanish-speaking children at the. 政治大. age of 4 to 7. The results showed the developmental differences in all the three age. 立. groups. Children used more discourse markers as they grew up. The results also. ‧ 國. 學. indicated that there seemed to be cross-linguistic similarities in that children used. ‧. discourse markers to convey social meaning and to manipulate social power relations.. Nat. io. sit. y. Kyratzis (2005) investigated how English children used discourse markers as a. al. er. collaborative stance in peer interaction. The findings showed that children used. n. v i n C h or collaboration.UThe results also demonstrated discourse markers to mark solidarity engchi gender differences in children’s use of discourse markers. Relatively little research has focused on Mandarin-speaking children’s acquisition of discourse markers. Huang (2000) investigated Mandarin-speaking children’s use of the discourse marker, “hao.” Children aged 3 to 5 were divided into three age groups and children’s conversations with adults and peers were observed. The results suggested that there is a developmental trend of Mandarin-speaking.

(24) 15. children’s use of hao as a discourse marker. In general, children at ages 3 and 4 use hao mostly at the local level, while children at age 5 could use hao both at the global level and local level. Older children were more capable of using more functions of hao than younger children.. 2.3 Previous studies on ranhou. 政治大. Several researchers have investigated the use of ranhou in adults’ spoken. 立. discourse (Chen, 2011; Shen, 1994; Shen, 2009; Su, 1998; Wang, 1996; Wang &. ‧ 國. 學. Huang, 2006). Shen (1994) discussed the function of ranhou in discourses based on. ‧. Schiffrin’s (1987) discourse model, which consists of five components: the ideational. Nat. io. sit. y. structure, the action structure, the exchange structure, the information state and the. al. er. participation framework. It was pointed out that ranhou can function as a connective. n. v i n C h that ranhou served or a discourse marker. Shen claimed e n g c h i U functions in three structures: the ideational structure, the information state and the action structure. The ideational structure is composed of linguistic units with semantic content, such as propositions. The information state is concerned with the speaker and hearer’s cognitive capacities to organize their meta-knowledge. The action structure refers to a structure in which speech acts are situated. The findings indicated that ranhou occurred in the ideational structure in most cases. It seldom appeared in the information state and the action.

(25) 16. structure. That is, ranhou mostly appears in only one structure in Schiffrin’s discourse model. Wang (1996) discussed the pragmatic use of ranhou in oral discourse. The sources of data were collected from monologues, dialogues, and conversations. Wang claimed that the primary meaning of ranhou is to denote “what is next.” Besides, the functions of ranhou were divided into two categories: temporal function and. 政治大. non-temporal function. Temporal function means that ranhou could be used to mark. 立. the temporal relation between events, and it was used to convey sequentiality of the. ‧ 國. 學. speaker’s event time. Wang further explained that event time was anaphoric because it. ‧. shows a relation between linguistic events and indicates the relation between two. Nat. io. sit. y. propositions. Non-temporal function means that ranhou could be used as a. al. er. continuation of an action, and it was used to denote speaker’s discourse time. It was. n. v i n C h to the temporal U explained that discourse time refers e n g c h i relation between utterances in a discourse. In a nutshell, Wang made a brief categorization of the functions of ranhou. However, she did not further explore the use and functions of ranhou in detail. Su (1998) investigated the use of ranhou in Chinese adult spoken discourse. Su pioneered in the analyses of ranhou by dividing her findings into two major categories: the ideational and the interactive use of ranhou. The ideational use refers to the connective use of ranhou, including temporal, consequence and lists functions..

(26) 17. The interactive use refers to the discourse marker use of ranhou, including condition or concession, verbal filler, and topic succession functions. In the ideational use, the temporal function of ranhou is to indicate the temporal sequence of two utterances indicating two events as they occur at different reference times. In other words, the starting point of one occurrence is followed by the starting point of another occurrence. The consequence function of ranhou pays more attention to the outcome. 政治大. resulted from the previous proposition and it shows the logical causal relationship. 立. between the cause and effect. The third function of ranhou is the lists function, which. ‧ 國. 學. are used to connect a group of entities that share the same feature. In the interactive. ‧. use, the conditional or concessional function of ranhou does not have semantic. Nat. io. sit. y. meaning; rather, the meaning of ranhou derives from the sentences or from the. al. er. context. The verbal filler function of ranhou occurs after a pause, and it indicates that. n. v i n the speaker is having difficultyCinhretrieving the linguistic e n g c h i U unit or target. The topic succession function of ranhou is regarded as a mechanism for turn-taking in order to achieve conversation coherence in interactions. The study also investigated the frequency of the use of ranhou in ideational use and interactional use. The results showed that the ideational use, namely the connective use of ranhou, is the primary use in spoken discourse. Wang and Huang (2006) investigated three connectives yinwei, suoyi, and.

(27) 18. ranhou in Mandarin Chinese using a corpus-based study. As to the functions of ranhou, five different functions proposed by Su (1998) were found in the data except for the condition or concession function. In addition, two other functions were pointed out, both of which were not mentioned in previous studies. One is the resumptive opener function, and the other is the additive function. When ranhou is used as a resumptive opener, it resumes an old, digressed topic rather than serves as the. 政治大. connective use. Ranhou here functions as a strategic device so that the speaker can. 立. trace back old topics; in other words, ranhou functions as a discourse marker to. ‧ 國. 學. organize discontinuous discourse units. The other function of ranhou newly. ‧. discovered is the additive function. The additive function of ranhou does not express. Nat. io. sit. y. a temporal relationship, but is more like adding new information to the old one. From. al. er. another viewpoint, it links a series of related events or actions together. The findings. n. v i n Cofh ranhou was the most showed that the additive function e n g c h i U frequent function, followed by temporal function and verbal filler function of ranhou.. 2.4 Research on the acquisition of ranhou Little research has been done on children’s acquisition of ranhou. Yeh (2011) investigated the use of ranhou in Mandarin-speaking children in story-telling narratives. The subjects were 5-year-old and 9-year-old children. Eight different.

(28) 19. functions of ranhou were coded based on the classification of Su (1998) and Wang and Huang (2006). The results showed that the frequency of ranhou in narratives decreased with age. Besides, the results suggested that children at age five had the abilities to use the temporal and additive functions of ranhou in their narrative. Besides the conventional temporal use of ranhou, most of the five-year-old children also used ranhou to add new information. Except for the temporal and additive uses,. 政治大. nine-year-old children were able to use ranhou to mark consequence. Moreover, it. 立. was pointed out that the detail of a nine-year-old children’s story was also presented. ‧ 國. 學. by their consecutive manipulation of the additive function of ranhou. In children’s. ‧. narratives, the most salient use of ranhou was as predicted, the temporal function to. Nat. io. sit. y. connect clauses or events. Other meanings, such as speech planning, appeared. al. er. relatively less. The temporal function of ranhou decreased with age while the additive. n. v i n C h with age. TheUfindings implied that although and consequential functions increased engchi children could not manipulate all functions of ranhou, they started to understand the multifunction of ranhou as a connective and a discourse marker at age five. Moreover, it also indicated a developmental difference in children’s use of the multifunction of ranhou..

(29) 20. Chapter 3 Methodology. 3.1 Subjects and data The data in this study were adopted from the database of Language Acquisition Lab in National Chengchi University1. The subjects in this study were 8 children aged. 政治大. from 2;7 to 5;4 who lived in Taipei. All subjects were native speakers of Mandarin. 立. Chinese. The major language used for communication at home for children was also. ‧ 國. 學. Mandarin Chinese. Among the eight subjects, five of them were boys and three were. ‧. girls. According to Hsu (2003), Mandarin-speaking children started to use ranhou at. Nat. io. sit. y. age 2 to 3, and children at age 5 to 6 were able to use connectives as adults do. Thus,. al. er. children aged from 2;7 to 5;4 were chosen. Subjects were further divided into the. n. v i n C hfor comparison: (1) younger group and the older group e n g c h i U Group 1: 2;7~3;2 (mean age:. 2;11); (2) Group 2: 4;0~5;4 (mean age: 4;10). Table 1 presents the subject information in Group 1 and Group 2.. 1. The lab is directed by Professor Chiung-chih Huang. I am deeply grateful to her contribution..

(30) 21. Table 1 Subject Information in Group 1 and Group 2. Group 1. Subjects. Gender Age. SEN TIN YIN BUO. M F F M. Mean Group 2. 2;7 2;10 3;1 3;2 2;11. XUN QIN LEE YOU. M M F M. 立. Mean. 4;0 4;11 5;2 5;4. 政治大 4;10. ‧ 國. 學. Mother-child conversations were videotaped at the children’s homes. Mothers. ‧. were told to do whatever activities they liked, only to make sure that they have. sit. y. Nat. io. n. al. er. spontaneous speech in mother-child conversations. The common activities for these. i n U. v. mother-child dyads were playing toys, reading books, chatting, role-playing, eating. Ch. engchi. food, and doing housework. The total length of time videotaped for each subject was about 2 hours. Thus, the data in this study included 8 children with a total time of approximately 16 hours. The data collected were transcribed following the CHAT (Codes for the Human Analysis Transcriptions) conventions. In order to ensure the reliability for the different functions of ranhou, one fourth of the data of ranhou were randomly chosen and coded by another person. Cohen’s Kappa was utilized to test the inter-rater reliability. The reliability for the coding of.

(31) 22. ranhou was 0.81.. 3.2 Analytical framework According to previous studies, ranhou could function as a connective or a discourse marker in spoken Mandarin Chinese (Su, 1998; Wang, 1996; Wang & Huang, 2006). Based on Su (1998) and Wang and Huang’s (2006) research, their. 政治大. classification of the use of ranhou was slightly modified as the main analytical. 立. framework of this study. First, Su (1998) indicated that the connective use of ranhou. ‧ 國. 學. includes temporal, consequence, and lists function. The lists function proposed by Su. ‧. was not adopted in that this use could be included in the additive function. The. Nat. io. sit. y. discourse marker use of ranhou includes the functions of conditional and concessional,. al. er. verbal filler, and topic succession. Wang and Huang (2006) later proposed two. n. v i n C not functions of ranhou, which were U One is the additive function, h ementioned n g c hbyi Su. which belongs to the connective use. The other is the function of resumptive opener, which belongs to the use of discourse marker. In summary, temporal, additive, consequence (named causal), verbal filler, topic succession, and resumptive opener functions were adopted in the study. Based on the Halliday and Hasan’s (1976) classification of connectives, the contrast function of ranhou as a connective use was added. Therefore, the total seven functions of ranhou in this study are classified into.

(32) 23. two major categories, as the connective use or as the discourse marker use2. There are four functions in the connective use of ranhou and three functions in the discourse marker use of ranhou, as shown in Figure 1.. Connective. Temporal Additive Causal Contrast 政治大. Ranhou. Discourse marker. 學. ‧ 國. 立. Verbal filler. ‧. Topic succession. er. io. sit. y. Nat. Resumptive opener. Figure 1. The Framework of the Classification of Ranhou in This Study. n. al. Ch. engchi. i n U. v. 3.2.1 Ranhou as the connective (1) Temporal: When ranhou serves as the temporal function, the notion of reference time is crucial to the discussion of the temporal function of ranhou. The starting point of one event precedes the starting point of another event. So the two sentences are highly 2. Although in some cases, the token of ranhou may seem to carry more than one function. However, in this study, only the primary function was chosen and coded. Thus, each token was coded as one function only..

(33) 24. connected by time sequence. Example 1 shows children’s temporal function of ranhou. In Example 1, YIN said “you wash this tomorrow” and then used the temporal function of ranhou to say “and then air it.” The time point of “washing” occurred before “airing,” and the two actions are connected by a time sequence.. Example 1 (YIN, 3;1): YIN was playing with the pacifier. *YIN: 明天妳把這個洗一洗.. 立. 政治大. 學. ‧ 國. *YIN:. mingtian ni ba zhe ge xiyixi. tomorrow you BA this CL wash “You wash this tomorrow.” 然後晾乾. . n. al. er. io. sit. y. Nat. (2) Additive:. ‧. ranhou lianggan. then air “And then air (it).”. i n U. v. When ranhou serves as the additive function, it links together a series of related. Ch. engchi. events or actions, and does not manifest as an overt time sequence. Each of them is encompassed within a larger discourse frame. Example 2 demonstrates children’s additive function of ranhou. In Example 2, when YIN was playing with toys and pretended to cook something, she used the additive function of ranhou to connect two related events, “one pot is to cook grapes” and “one pot is to cook corns,” which do not have an overt indication of time sequence..

(34) 25. Example 2 (YIN, 3;1): YIN pretended to cook something. *YIN: 媽媽還有一個鍋子要煮 mama hai. you yi ge guozi yao. 葡萄. 的.. zhu putao de.. mom also have one CL pot be going to cook grapes NOM “Mom, and a pot is going to cook grapes.” *MOT: 恩. en.. *YIN:. um “Um.” 然後再. 一個. 鍋子. 是. 要. 煮 # 玉米的.. . zhu # yumi de.. ranhou zai yi ge guozi shi yao. and again one CL pot COP be going to cook corn NOM “And a pot is going to cook corn.”. 立. 政治大. ‧ 國. 學. One thing to note is that the lists function proposed by Su (1998) was not adopted because this function is similar to the additive function. Example 3 was. ‧. excerpted from Su’s example of lists function. In Example 3, Su mentioned that the. sit. y. Nat. io. n. al. er. speaker begins by listing a member of the set (e.g. chorus, folk band music) and then. i n U. v. adds another entity (military band) to the list. However, this “adding another entity” is. Ch. engchi. similar to the additive function, which links successive ideas or related events. This function is therefore not included in the study.. Example 3 (pp.173-174 of Su) ADU: 除了體育活動. ADU:. 以外. 還. 有. 其他. chule tiyu huodong yiwai hai you qita besides sports activity else still have else “Besides sports, there are other (activities).” 比如說有合唱隊. biru. shuo you hechang dui.. 的. de. NOM.

(35) 26. ADU:. example say have chorus team “For example, there is a chorus team.” 一個禮拜活動一次.. ADU:. yi ge libai huodong yici. one CL week activity once “Once a week.” 有民樂隊.. ADU:. you min yuedui. have folk band “There is a folk music band.” 那些二胡和什麼笛子.. ADU:. naxie erhu han shenme dizi. those erhu fiddles and what flutes “Such as erhu fiddles and flutes.” 然後還有軍樂隊. . 立. 政治大. ‧ 國. ‧. io. sit. y. Nat. (3) Causal:. 學. ranhou hai you jun yuedui. and still have military band “And there is a military band.”. er. The function of ranhou for causal meaning pays more attention to the outcome. al. n. v i n C h It focuses moreUon the logical cause-and-effect resulting from a previous proposition. engchi between two events or actions. Example 4 shows children’s causal function of ranhou. In Example 4, LEE said “JEE was pulling LIN’s pants at that time,” and used the causal function of ranhou to say “so it caused LIN’s pants to fall off.”. Example 4 (LEE, 5;1): LEE was talking about an interesting thing at school to her mother. *MOT: 你給 JEE 脫褲子喔? ni. gei JEE tuo kuzi. o?.

(36) 27. *LEE:. you to (name) took pants Q “You took off JEE’s pants?” 沒有!. *LEE:. mei you! NEG have “No!” 是 JEE # 以前. *LEE:. shi JEE # yiqian la zhe LIN de kuku. COP(name) before pull DUR (name) GEN pants “JEE was pulling LIN’s pants at that time.” 然後害 LIN 的褲子掉下來! . 拉. 著. LIN 的褲褲.. ranhou hai LIN de kuzi diao xialai! so cause (name) GEN pants fall off “So it caused LIN’s pants to fall off.”. 立. ‧. ‧ 國. 學. (4) Contrast:. 政治大. When ranhou serves as a contrast function, its basic meaning is “contrary to. sit. y. Nat. io. n. al. er. expectation.” It connects the phrases or utterances which the logical relation of the. i n U. v. latter utterance is opposite to the expectation of the former one. Example 5. Ch. engchi. demonstrates children’s contrast function of ranhou. In Example 5, when YIN and her mother were reading a storybook, YIN said “Xiao-bu used the big one,” and used the contrast function of ranhou to show the opposite of her expectation that “but Nu-nu used the small one.”. Example 5 (YIN, 3;1): YIN and her mother were reading a storybook. *YIN: 為什麼小布用大的然後努努用小的? weishenme xiao-bu yong da de ranhou nu-nu yong xiao de?. .

(37) 28. why (name) use big NOM and (name) use small NOM “Why did Xiao-bu use the big one and Nu-nu use the small one?” *MOT: 阿努努比較大還是小布比較大?. *YIN:. a nu-nu bijiao da haishi xiao-bu bijiao da? uh (name) more big or (name) more big “Was Nu-nu bigger or was Xiao-bu bigger?” 小布. xiao-bu . (name) “Xiao-bu.”. 政治大. 3.2.2 Ranhou as the discourse marker. 立. (1) Verbal filler:. ‧ 國. 學. When ranhou functions as verbal filler, it indicates that the speaker is thinking of. ‧. what they are going to say or has some difficulty retrieving the words they wish to say.. Nat. io. sit. y. So here ranhou functions to fill the empty time slot. It may occur after or before a. al. er. pause, or it may end in an incomplete sentence and is then followed by another. n. v i n C h verbal filler function ranhou. Example 6 shows children’s e n g c h i U of ranhou. In Example 6,. TIN’s verbal filler function of ranhou was followed by the sound lengthening and pause, indicating TIN was thinking of what she was going to say.. Example 6 (TIN, 2;10): TIN and her mother were reading a storybook. *TIN: 等什麼? deng shenme? wait what “Who was (it) waiting for?” *MOT: 等奴奴啊!.

(38) 29. deng nu-nu a! wait (name) PRT “(It) was waiting for Nu-nu.” *MOT: 等奴奴來啊!. *TIN:. deng nu-nu lai a! wait (name) come PRT “(It) was waiting for Nu-nu to come.” 然後 -: # 那奴奴為什麼要. . 來?. ranhou-: # na nu-nu weishenme yao lai? well x (name) why need come “Well, why did Nu-nu come?” *MOT: 奴奴要過來幫忙刷油漆啊! nu-nu yao guolai bangmang shua youqi a! (name) need come help brush paint PRT “Nu-nu came to help to paint.”. 立. 政治大. ‧ 國. 學. In Example 7, YOU’s use of ranhou was not followed by any words, indicating. ‧. ranhou here has the verbal filler function rather than the connective function. And it. sit. y. Nat. io. n. al. er. was followed by another ranhou to connect the following utterance.. Ch. engchi. i n U. v. Example 7 (YOU, 5;3): YOU and his mother were reading a storybook about color changing of the caterpillar. *YOU: 是黑的然後是黑白黑白的. shi hei de ranhou shi hei bai hei bai de. COP black NOM then COP black white black white NOM “(It) is black and then black and white.” *MOT: 黑白黑白的喔.. *YOU:. hei bai hei bai de o. black white black white NOM PRT “Oh, black and white.” 然後 +…  ranhou +… well.

(39) 30. *YOU:. “Well…” 然後它變綠. 的. 時候. 就. 長成. 這. 樣子.. ranhou ta bian lü de shihou jiu zhangcheng zhe yangzi. then it turn green NOM time just grow this way “Then when it turned green it grew like this.”. (2) Topic succession: When ranhou serves as a topic succession function, it is used so that the transition from one speaker to another seems justified because one carries on the topic. 治政 where the other speaker stops. It is a mechanism for turn-taking 大 so that the interaction 立 ‧ 國. 學. in the conversation seems coherent. Example 8 demonstrates children’s topic succession function of ranhou. In Example 8, LEE’s mother was talking about. ‧. washing and cutting vegetables. When the mother stopped and it was LEE’s turn, LEE. sit. y. Nat. io. n. al. er. used the topic succession function of ranhou to carry on the topic about cutting vegetables so that the turn-taking seems smooth.. Ch. engchi. i n U. v. Example 8 (LEE, 5;2): LEE and her mother were washing vegetables. *MOT: 好. 先洗一洗.. hao xian xiyixi. ok first wash “Ok, wash (it) first.” *MOT: 等一下把它切小一點.. *LEE:. dengyixia ba ta qie xiao yidian. later BA it cut small few “Later on cut it to the small size.” 然後一直切一直切切切切切. ranhou yizhi qie yizhi qie qie qie qie qie.. .

(40) 31. well keep cut keep cut cut cut cut cut “Keep cutting and cutting.” *MOT: 對. dui. yes “Yeah.”. (3) Resumptive opener: When ranhou is functioning as a resumptive opener, it does not establish a. 政治大. sequential relationship, but rather resumes the old topic just mentioned. Example 9. 立. demonstrates children’s resumptive opener function of ranhou. In Example 9, BUO. ‧ 國. 學. was playing with a toy car, fire engine. BUO’s mother asked BUO a question, but. ‧. BUO did not immediately respond to that question. Instead, he used the resumptive. Nat. sit. n. al. er. io. about.. y. opener function of ranhou to go back to the old topic “toy car” he wanted to talk. Ch. engchi. i n U. v. Example 9 (BUO, 3;2): BUO and his mother were playing with the toy car (fire engine). *MOT: 好了沒啊?. *BUO:. hao le mei a? ok PFV NEG Q “Is (it) ok?” 好了.. *MOT:. hao le. ok PFV “Ok.” 好了那消防車 hao le na. 要. 回去了.. xiaofangche yao huiqu le..

(41) 32. *MOT:. ok PFV x fire engine need back CRS “Ok. The fire engine is going to go back.” 今天自己吃飯還是給李老師餵?. *BUO:. jintian ziji chifan haishi gei li laoshi wei? today self have lunch or to Lee teacher feed “(You) had lunch by yourself or Ms. Lee fed (you) today?” 今天給 # <自己吃> [//] 李老師餵.. *BUO:. jintian gei # <ziji chi> [//] li laoshi wei. today to self eat Lee teacher feed “Ms. Lee fed (me) today.” 然後它就會失火 [% playing with the fire engine].. *BUO:. ranhou ta jiu hui shihuo [% playing with the fire engine]. well it just will fire “It would catch fire.” 它 # /dung -: /.. 立. ‧. ‧ 國. 學. ta # /dung -: /. it bang “It banged.”. 政治大. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. .

(42) 33. Chapter 4 Results. This chapter presents the findings of the three research questions in the present study, respectively. The first research question, concerning the distribution of ranhou, including its frequency and its distribution as connectives or discourse markers, was. 政治大. shown in section 4.1. In section 4.2, regarding the second research question, the. 立. quantitative and qualitative analyses of children’s ranhou as connectives were. ‧ 國. 學. analyzed, including the temporal, additive, causal, and contrast functions of ranhou.. ‧. In section 4.3, with regard to the third research question, the quantitative and. Nat. io. sit. y. qualitative analyses of children’s ranhou as discourse markers were analyzed,. n. al. er. including the verbal filler, topic succession, and resumptive opener functions of ranhou.. Ch. engchi. i n U. v. 4.1 The distribution of ranhou Table 2 presents the frequency of ranhou used by all subjects in the two age groups. As seen in the table, all utterances in Group 1 and Group 2 were 4864 and 5105 respectively. Among the utterances, the utterances with ranhou in Group 1 and Group 2 were 82 and 130 respectively. As to the proportions of utterances with.

(43) 34. ranhou in all utterances, Group 1 was 1.69% while Group 2 was 2.55%. It shows that children’s use of ranhou in conversations increased with age. The results were inconsistent with Yeh’s (2011) findings, which suggested that the frequency of ranhou decreased with age.. Table 2. 政治大 Utterances with All utterances % of utterances 立 ranhou with ranhou. The Frequency of Ranhou in the Two Age Groups. 4864. 19 33 49 29. 961 1356 1478 1310. al. n. Total. io. XUN QIN LEE YOU. Ch. 130. i engch 5105. 1.69%. y. 82. 1.98% 2.43% 3.32% 2.21%. sit. Total. 0.35% 1.38% 4.04% 1.31%. er. 1420 1084 1139 1221. ‧. 5 15 46 16. Nat. Group 2. SEN TIN YIN BUO. 學. Group 1. ‧ 國. Subjects. i n U. v. 2.55%. As shown in Table 2, younger children used less ranhou than older children. However, YIN’s proportion of utterances with ranhou in all utterances was 4.04%, which was far more than other children in the same group. The possible reason may be that YIN used several consecutive uses of ranhou, which may increase the frequency of ranhou, and this use was not usually seen in other children’s data..

(44) 35. Example 10 shows YIN’s consecutive use of ranhou. In Example 10, YIN used ranhou three times in a row to connect her utterances, which were “then he goes to school in Taipei,” “and I waited for him at home” and “then he sent a letter to me.”. Example 10 (YIN, 3;1): YIN and her mother were playing toys. *MOT: 這麼棒喔.. 政 治大 ranhou ta qu taibei shangxue. 立 then he go Taipei go to school. 學. io. ranhou ta fa yi ge xin gei wo. then he sent one CL letter to me “Then he sent a letter to me.” *MOT: 喔他發一個信給妳喔.. n. al. Ch. engchi. y. sit. . er. *YIN:. ranhou wo zai jiali deng ta. and I at home wait him “And I waited for him at home.” 然後他發一個信給我.. Nat. *YIN:. “Then he goes to school in Taipei.” 然後我在家裡等他. . ‧. ‧ 國. *YIN:. zheme bang o. so great PRT “So great.” 然後他去台北上學.. i n U. v. o ta fa yi ge xin gei ni o. oh he send a CL letter to you PRT “Oh, he sent you a letter.”. It appears that children in the younger group used less ranhou. However, SEN’s proportion of utterances with ranhou in all utterances was only 0.35%, which was far less than other children in the same group. The reason why SEN used fewer ranhou was not that clear. Several reasons which might have led to SEN’s few tokens of.

(45) 36. ranhou were proposed. First, in other children’s data, the children’s mother sometimes used utterances with ranhou to ask children questions to induce their children’s use of ranhou, but SEN’s mother rarely used utterances with ranhou in this way. The only example in SEN’s data can be seen in Example 11. In Example 11, SEN’s mother used ranhou to ask SEN a question, and SEN answered the mother’s question with ranhou. Second, ranhou could be used to connect events or to make the conversation. 政治大. more coherent. But in SEN’s data, most of his utterances were just a short reply to his. 立. mother, so ranhou might not occur in such circumstances.. ‧ 國. 學 ‧. Example 11 (SEN, 2;8): SEN and his mother were reading a storybook. *MOT: 他在睡覺.. y. Nat. ta zai shuijiao.. n. al. er. “He is sleeping.” 對.. io. *SEN:. sit. he DUR sleep. dui. yeah. “Yeah.” *MOT: 然後咧?. Ch. engchi. i n U. v. . ranhou lie? then. Q. “And then?”. *SEN:. 然後他 /ji/ # 孫悟空. 就. <被> [/] 被綁. ranhou ta /ji/ # SunWu-kong jiu <bei> [/] bei bang. then he (name) just BEI “Then Sun Wu-kong was kidnapped.”. 了.. . le.. BEI kidnap CRS. In order to understand the distribution of the use of ranhou, the use of ranhou as.

(46) 37. the connective or discourse marker for the two age groups was investigated. Table 3 shows the distribution of ranhou as the connective or discourse marker use. The proportion of utterances with connective use of ranhou in Group 1 and Group 2 was 1.40% and 1.55%, respectively. The proportion of utterances with discourse marker use of ranhou in Group 1 and Group 2 was 0.29% and 0.67%, respectively. Both groups used the connective use more than the discourse marker use of ranhou. It. 政治大. shows that the connective use slightly increased with age (from 1.40% to 1.55%), and. 立. the discourse marker use increased with age (from 0.29% to 0.67%), which increased. ‧ 國. 學. more than the connective use.. ‧. io. sit. y. Nat. Table 3. n. al. er. The Distribution of Ranhou as Connective Use or Discourse Marker Use Subjects. Ch. i n U. v. Ranhou as connective use. Ranhou as discourse marker use. Utterances % of all. Utterances. engchi. All utterances. % of all. Group 1 SEN TIN YIN BUO. 3 12 40 13. 0.21% 1.11% 3.51% 1.06%. 2 3 6 3. 0.14% 0.28% 0.53% 0.25%. 1420 1084 1139 1221. Total. 68. 1.40%. 14. 0.29%. 4864. Group 2 XUN QIN LEE YOU. 16 21 40 19. 1.56% 1.11% 2.23% 1.22%. 3 12 9 10. 0.31% 0.88% 0.61% 0.76%. 961 1356 1478 1310. Total. 96. 1.55%. 34. 0.67%. 5105.

(47) 38. 4.2. Ranhou as the connective use In order to further understand the use of ranhou, the distribution of each function of ranhou as the connective is investigated. Table 4 shows the proportion of different functions of ranhou as connectives in the two age groups. In Table 4, the temporal function in Group 1 was 86.76%. The additive function and the causal function in Group 1 only accounted for 8.82% and 2.94%. In Group 2, the temporal function. 政治大. decreases to 59.38% and the additive function and the causal function increase to. 立. 23.96% and 12.50%, respectively. The contrast function was 1.47% in Group 1 and. ‧ 國. 學. 4.17% in Group 2, respectively. It shows that the temporal function decreased with. ‧. age while the additive and causal functions increased with age. The contrast function. Nat. n. al. er. io. sit. y. was few in both groups.. Table 4. Ch. engchi. i n U. v. The Proportion of Different Functions of Ranhou as the Connective Group 1. Group 2. Categories. tokens. %. tokens. %. Temporal Additive. 59 6. 86.76 8.82. 57 23. 59.38 23.96. Causal. 2. 2.94. 12. 12.50. Contrast. 1. 1.47. 4. 4.17. Total. 68. 100. 96. 100.

(48) 39. 4.2.1 Temporal function Table 5 presents the tokens of children’s temporal function of ranhou in the two groups. As shown in Table 5, in Group 1 and Group 2, all subjects used the temporal function of ranhou. It may suggest that younger children are already capable of using the temporal function of ranhou, and it may also indicate that the temporal function of ranhou has such common usage in all functions of ranhou.. 政治大. 立. Table 5. ‧ 國. 學. The Tokens of Temporal Function of Ranhou for All Subjects in the Two Groups. Group 2. XUN QIN LEE YOU Total. io. n. al. Ch. 8 17 19 13 57. y. 2 9 35 13 59. sit. SEN TIN YIN BUO Total. er. Group 1. ‧. Tokens. Nat. Subjects. engchi. i n U. v. The qualitative analysis was conducted to understand the characteristics of children’s temporal function of ranhou. First, temporal function of ranhou occurred in a variety of situations, but the observation showed that it occurred more in a book-reading context than other contexts, as shown in Example 12. In Example 12,.

(49) 40. BUO and his mother were reading a storybook. BUO was asked to tell a story to his mother. After BUO said that the moon had come out, he continued his story by using the temporal function of ranhou to say that they wanted to go to sleep. Second, the temporal function of ranhou was often combined with a question particle such as “ne” or “lie” to form a question, as shown in Example 13. In Example 13, XUN’s mother asked XUN to eat their breakfast faster, but XUN kept eating it slowly. The mother. 政治大. said that she would keep waiting until her hair turned gray. XUN used the temporal. 立. function of ranhou plus the question particle “lie” to ask “and then,” wanting to know. ‧ 國. 學. what would happen next.. ‧ er. le.. night. n. all come out PFV amoon v i “Tonight the moonlhas come out.” n Ch engchi U 嗯. today. *MOT:. io. jintian wanshang yueliang dou chulai. sit. y. Nat. Example 12 (BUO, 3;2): BUO and his mother were reading a storybook. *BUO: 今天晚上月亮都出來了.. en.. *BUO:. um “Um.” 然後 /ne/ <他> [/] 他們 # /hum/ [= sigh] 想. 睡覺. 了.. ranhou /ne/ <ta> [/] tamen # /hum/ [= sigh] xiang shuijiao le then PRT. he. they. “Then they want to go to sleep.” *MOT: 嗯. en.. um “Um.”. want sleep CRS. .

(50) 41. Example 13 (XUN, 4;0): XUN’s mother asked XUN to eat their breakfast faster. *MOT: 你在長大我在變老 /ei/. ni. zai zhangda wo zai. bian. lao /ei/.. you DUR grow up I DUR become old PRT “You’re growing up and I’m getting old.” *MOT: 你不要 # 等到我頭髮都變白掉. ni bu yao # deng dao wo toufa dou bian bai diao. you NEG want wait reach I hair all become white fall. *XUN:. “I don’t want to wait until my hair turn gray.” 然後咧?  ranhou lie? then. Q. “And then?”. 的蟲蟲. 政都治長綠色大 ranhou na ge mianbao shangmian dou chang lüse de chongchong. 立 then that CL bread top all grow green NOM worm 那個麵包. 上面. 學. ‧ 國. *MOT: 然後. “Then the green worm will grow on the top of the bread.”. ‧. As to the difference of the temporal function of ranhou between Group 1 and. sit. y. Nat. io. n. al. er. Group 2, the observation was as follows. In Group 1, children usually used the. i n U. v. temporal function of ranhou in the here-and-now situation; that is, referring to the. Ch. engchi. referents which are present in the physical context of the conversation (Huang, 2003), as illustrated in Example 14. In Example 14, YIN and her mother were reading a storybook. YIN’s mother mentioned the full moon and crescent moon in the storybook, saying that they were all moons but with different shapes. YIN said that she covered the full moon in the storybook by her hands, and she used the temporal function of ranhou to say “and then I hang the moon (covered by hands) here.” Ranhou here connects YIN’s two actions of time sequence and is thus regarded as a temporal.

(51) 42. function. The referent is the moon in the storybook, which is a here-and-now referent. In Group 2, in addition to using the temporal function of ranhou in the here-and-now situation, children were more likely to use the temporal function of ranhou in the there-and-then situation; that is, referring to the referents which are absent in the physical context of the conversation (Huang, 2003), as illustrated in Example 15. In Example 15, QIN’s mother asked QIN if he remembered the show they watched the. 政治大. day before yesterday. QIN used the temporal function of ranhou to mention what the. 立. character “grandpa” did in the show. The referent “grandpa” is absent in the physical. ‧ 國. 學. context of the conversation. Thus he used the temporal function of ranhou in the. ‧. there-and-then situation.. sit. y. Nat. er. io. Example 14 (YIN, 3;1): YIN and her mother were reading a storybook. *MOT: 彎彎的半圓形圓形通通都是月亮. al. 阿.. n. v dou shi yueliang a. i wanwan de ban yuan xing yuan xing tongtong n Ch i Uevery all COP moon PRT e nround curve NOM half round shape shape gch. “The crescent moon and full moon are all moons.” *MOT: 是不是有時候都不一樣? shi-bu-shi you shihou dou bu yiyang? COP-NEG-COP have time all NEG same “Are (they) sometimes different?” *MOT: 是不是?. *YIN:. shi-bu-shi? COP-NEG-COP “Right?” <啊> [/] 啊媽咪 # 把那一 <a> [/]. a. mami # ba. na. yi. 個遮起來 [% sticking out hands to cover the full moon by half]. ge zhe qilai [% sticking out hands to.

(52) 43. cover the full moon by half].. *YIN:. ah ah mom BA that one CL cover up “Mom, cover it up (by hands).” 然後掛在這裡 [= besides the crescent moon].. *YIN:. ranhou gua zai zheli [= besides the crescent moon]. then hang at here “Then hang it here.” 就好像半圓形.. . jiu haoxiang ban yuan xing. just like half round shape “(It) looks like the crescent moon.”. 政治大 watched the day before yesterday. 立後來呢後來呢?. Example 15 (QIN, 4;11): QIN’s mother asked QIN if he remembered the show they. *QIN:. na houlai ne houlai ne? x later Q later Q “What’s going on later?” 然後 <爺> [/] 爺爺有一次就 #. *QIN:. ranhou <ye> [/] yeye you yici jiu # xu yiban. then grandpa grandpa have once just pee half “Then once the grandpa was peeing.” 然後豆太 [= the main character] 就尿一半.. *MOT:. ranhou dou-tai [= the main character] jiu niao yiban. then (name) just pee half “Then Dou-tai was peeing.” 喔 -: [=! nodding].. 噓一半.. . n. al. er. io. sit. Nat. y. ‧. ‧ 國. 那. 學. *MOT:. Ch. engchi. i n U. v. o -: [=! nodding]. oh “Oh.”. 4.2.2 Additive function Table 6 presents the tokens of children’s additive function of ranhou in the two groups. As shown in Table 6, in Group 1, only two subjects (TIN and YIN) used the.

(53) 44. additive function of ranhou. In Group 2, all of the subjects used the additive function of ranhou. It indicates that older children were more capable of using the additive function of ranhou.. Table 6 The Tokens of Additive Function of Ranhou for All Subjects in the Two Groups. SEN TIN YIN BUO Total. 0 3 3 0 6. 立. y. sit. n. er. io. al. ‧. 5 4 10 4 23. Nat. XUN QIN LEE YOU Total. 政治大. 學. Group 2. Tokens. ‧ 國. Group 1. Subjects. Ch. engchi. i n U. v. The characteristics of children’s additive function of ranhou were observed. Different from the temporal function of ranhou which occurred more in book-reading, children’s additive function of ranhou occurred more in a toy-playing context, as shown in Example 16. In Example 16, TIN and her mother were playing with some colored balls, and the mother asked TIN to identify the color of the balls. After TIN realized that “this is a blue ball,” she took a green ball and used the additive function of ranhou to say “and this is green.”.