Chapter 3 Methodology
3.3 Data Analysis
3.3.3 Mother-child Interaction—Topic Episode
國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
together be investigated to see if their tendencies of conversational behaviors were consistent with those maternal interactional styles distinguished in previous studies.
3.3.2 Children’s Topic-maintaining Competence
In order to investigate how children related their utterances to preceding utterances produced by their mothers in the process of ongoing conversational exchange, all the topic-maintaining utterances of each child were first identified and then analyzed in accordance with communicative intents (Wanska & Bedrosian, 1986; Huang, 2004).
These topic-maintaining utterances were further analyzed in terms of topic collaboration and topic incorporation (Huang, 2004). In such a way, frequency of adding new
information in children’s topic-maintaining utterances could be observed. In addition, children’s rate of topic change was examined to see how frequent they change topics in the conversational interaction (McDonald & Pien, 1982; Olsen-Fulero, 1982; Lin, 2006).
3.3.3 Mother-child Interaction—Topic Episode
In order to investigate topic maintenance in mother-child interaction, data were examined based on topic episode, or conversational exchanges conforming to a
propositional content. For the purpose of our study, a topic episode is defined as a stretch of discourse that contains utterances that continue a discourse topic. A stretch of
discourse may contain a series of linked utterances and the utterances may be linked in a least two ways (Keenan & Schieffelin, 1983): First, two or more utterances may share the same discourse topic, which is considered collaborating discourse topics by Keenan and Schieffelin (1983). Second, discourse topics may take some presupposition of the
immediately preceding discourse topic and/or the new information provided relevant to the discourse preceding and use it in a new utterance, which is considered incorporating
‧
discourse topics by Keenan and Schieffelin (1983). In addition, both collaborating and
incorporating discourse topics are continuous topics.The following scheme presents how a topic episode is identified in our study:
Figure 1: Preliminary topic episode identification scheme
U1 topic introducing / reintroducing discourse topic A U2 collaborating / incorporating discourse topic A ((
))
Un-1 collaborating / incorporating discourse topic A Un topic introducing / reintroducing discourse topic B
A minimum topic episode is consisted of at least two utterances, a topic introducing / reintroducing utterance and a topic continuing utterance. If the immediately following utterance continues the discourse topic, the topic episode will become longer. Un represents that in theory the same discourse topic can be continued or maintained endlessly, which is nevertheless hardly possible in daily conversation due to physical limitation of the speakers. Suppose that Un discontinues the discourse topic and introduces a new discourse topic / reintroduces a previously introduced discourse, then the present topic episode ends. In sum, a topic episode begins with an utterance with a discourse topic that is continued by the immediate following utterance and ends with an utterance with a discourse topic that is discontinued by the immediate following
utterance.
Here it should be noted that the above model is in theory utterance-based but not speaker-based. If such a topic episode involves only one speaker, it is realized as a
T opi c E pi sode
‧
monologue or narration. However, in real conversational exchanges, a discourse topic is usually maintained through several utterances by at least two speakers. To capture the nature of mother-child interaction of topic maintenance in our study, the identification scheme of a topic episode should thus be presented as speaker turns consisting of subordinated utterances that continue a discourse topic:
Figure 2: Topic episode identification scheme involving two speakers
S1 TURN 1 U1 topic introducing/reintroducing discourse topic A The above scheme shows that in our study, for a topic episode a discourse topic is
maintained or continued by at least two speaking turns and each speaking turn contains at least one utterance. This model is in theory necessary since in conversational exchanges speakers frequently elaborate their statements or add more comments when maintaining
T opi c E pi sode
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
discourse topics in a single speaking turn. In addition, if in a given speaking turn the speaker makes a topic introducing / reintroducing utterance at a certain point, the given speaking turn will serve as the closure of the topic episode:
Figure 3 Speaking turn with topic introducing / reintroducing utterance
S1 Turn U1 topic introducing / reintroducing discourse topic A ((
Un collaborating / incorporating discourse topic A S2 Turn U1 collaborating / incorporating discourse topic A
((
Un-1 collaborating / incorporating discourse topic A Un topic introducing / reintroducing discourse topic B The total number of topic episode were identified and analyzed to investigate the interaction of maternal interacitonal styles and topic maintenance.
T opi c E pi sode
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Chapter 4 Results
Results of data analysis are presented in this chapter. In section 4.1, utterances of the
two mothers were analyzed in terms of the speech category level and the conversational parameter level. In addition, their conversational characteristics and interactional styles were compared on the basis of previous studies. Section 4.2 is concerned with the two children’s topic maintaining utterances and their topic maintaining competences. In section 4.3, the two dyads’ topic maintenance was examined and related to the twomothers’ interactional styles in interactional contexts.
4.1 Maternal Interacitonal Style
Results regarding the speech category and conversational parameter level are presented in section 4.1.1 and 4.1.2 respectively. The two mothers’ interacitonal styles are
distinguished according to previous studies.
4.1.1 The Speech Category Level
Results of the speech category level in the two mothers’ utterances are presented in table 3. Over one-third of LIN’s mother’s (35.4%) and LJW’s mother’s (38.9%)
utterances are questions. It suggests that questions are frequently adopted by adults to compensate for children’s immature conversational skills such as conversational exchanging and topic maintenance, which conforms to previous studies concerning mother-child conversation. As for the subcategories of questions, no significant difference is observed in two mothers except for the more frequent use of report questions and permission requests in LJW’s mother’s questioning utterances. While LJW’s mother adopted prompts (8.6%), attention devices (5.4%), and declaratives
‧
(23.7%) slightly more frequently than LIN’s mother, LIN’s mother (19.6%) adopted more directives than LJW’s mother (5.9%).
Table 3 Results of the speech category level in two mothers’ utterances
LIN’s Mother LJW’s MotherDirectives 260(19.6%) 172(5.9%)
Questions
Repairs 28(2.1%) 42(1.4%)
Test Questions 122(9.3%) 281(9.6%)
Real Questions 179(13.6%) 331(11.3%) Verbal Reflectives 49(3.7%) 76(2.6%) Action Reflectives 14(1.1%) 12(0.4%) Report Questions 48(3.6%) 268(9.7%)
Permission Requests/
Offer of Helps
26(1.9%) 129(4.4%)
Total 466(35.4%) 1139(38.9%)
Prompts 69(5.2%) 252(8.6%)
Attention Devices 43(3.3%) 159(5.4%)
Responses to Question/Directives 76(5.8%) 105(3.6%)
Acknowledgements 151(11.5%) 405(13.8%)
Declaratives 251(19.1%) 693(23.7%)
Total 1316(100%) 2925(100%)
In both mothers’ utterances, questions were the most frequently. The mother could keep
‧
the conversation continuing by manipulate different kinds of questions:
(1) MOT is questioning about where LJW’s bag is.
1 *MOT: 你的袋子咧?
As shown in example (1), LJW’s mother began with a real question that elicited LJW’s verbal response. LJW then provided the answer, in which the location requested was not explicit due to the use of referential expression ‘there’. Since LJW’s mother didn’t receive enough information to infer the location of the bag, she replied with a repair. In order to elicit the answer she needed, the repair was immediately followed by a new real question which served the same function to the first real question. Results presented in table (3) have illustrated that among all the speech categories, both LIN and LJW’s mother questioned their children the most frequently. What differed the most significantly in their speech categories was that while LJW’s mother appealed to declaratives the second frequently, LIN’s mother appealed to directives. In the middle of periods of ongoing conversations, it’s observed that LIN’s mother frequently used directives to
‧
control LIN’s physical behaviors rather than verbal behaviors:
(2) LIN is looking at a penguin on the cover of a videotape.
1 *MOT: 一綾姊姊請問你在幹什麼?
‘What are you doing, LIN?’
2 *LIN: 我在看企鵝.
‘I’m looking at the penguin.’
3 *MOT: 你在看企鵝.
‘You’re looking at the penguin.’
4 *MOT: 那請你的腳可不可以把他放出來.
‘Please take out your feet.’
5 *LIN: 唉唷我知道了啦 [= in loud voice]!
‘O.K. I knew it!’
In example (2), LIN was thinking about which videotape to watch and looking at one of the videotapes. LIN’s mother asked what she was doing and was provided with LIN’s proper response. However, LIN’s mother didn’t elicit more information about what LIN was focusing on but turned to control the place she thought her child’s feet should be in.
4.1.2 The Conversational Parameter Level
Table 4 Results of two mothers’ conversational parameter level
LIN’s Mother LJW’s Mother
Topic Maintenance 9.4 (952/101) 19.2(1973/103)
Rate of Topic Change 9.9% (131/1316) 2.7%(78/2925)
Results of the mothers’ conversational parameter are illustrated in table 4. It can be clearly observed that the average length of topic episode maintained by LJW’s mother and LJW was 19.2, which is much longer than that maintained by LIN’s mother and LIN
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
(9.4). As for the rate of topic change, results showed that LIN’s mother changed topics much more frequently than LJW’s mother. In general, the two mothers’ conversational behaviors in the conversational parameter differed to a great extent.
In section 4.1.1 and 4.1.2, the two mothers’ speech was analyzed in terms of the speech category level and the conversational parameter level. In the next section, results of the two levels are to be compared to those in McDonald and Pien (1982) and
Olsen-Fulero (1982)’s study. In addition, results of the conversational parameter level will be further examined in terms of the two dyads’ conversational interactions in section 4.3.
4.1.2 Maternal Interactional Style of the Two Mothers
By looking into the two mothers’ distribution of utterances in the speech category level and the conversational parameter level in the previous sections, pictures of the two mother’s conversational characteristics were presented. As hypothesized by McDonald and Pien (1982), determination of maternal underlying intention would be inferable from patterning of illocutionary acts of mothers. In terms of the speech category level, by adopting directives much more frequently than LJW’s mother, LIN’s mother seemed intending more to control LIN’s physical actions. In terms of the conversational
parameter level, LIN’s mother changed topics much more frequently than LJW’s mother and LIN’s dyad maintained topics through less turns than LJW’s dyad. Compared with McDonald & Pien (1982), LIN’s mother’s cluster of conversational behaviors conformed more to the directive mother distinguished in their study, and LJW’s mother’s cluster of conversational behaviors conformed more to the conversation-elicitatiing mother.
Compared with Olsen-Fulero (1982)’s study, LIN’s mother conformed more to the
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
directive mother, and LJW’s mother conformed more to the conversational mother. In the study, the interactional style of LIN’s mother would be considered more directive and LJW’s mother would be considered more conversational-eliciting in the study. With such intentions of the two mothers in mind, we are to relate them to the two dyads’
topic-maintaining in section 4.3.
4.2 Children’s Speech in Topic Maintaining Utterances
In this section, the two children’s topic maintaining competences are examined by means of communicative intents, topic collaboration / incorporation, and rate of topic change.
4.2.1 Communicative Intents
Table 5 Results of two children’s communicative intents in topic-maintaining utterances
LIN LJW
Declaratives 84(18.2%) 198(19.8%)
Questions 29(6.3%) 37(3.7%)
Requests 77(16.7%) 70(7%)
Acknowledgements 27(5.9%) 77(7.7%)
Responses 243(52.8%) 617(61.8%)
Total 460(100%) 999(100%)
Results of two children’s communicative intents in topic-maintaining utterances are presented in table 5. In both LIN’s and LJW’s topic-maintaining utterances, responses are the most frequently occurring communicative intent, taking up nearly half tokens in both data, which corresponds to their mother’s frequent uses of questions in their utterances.
‧
Nevertheless, LJW appealed to responses more often than LIN did when maintaining topics, which might suggest that LJW attended more to her mother’s questions or requests than LIN.
(3) MOT and LJW are talking about LJW’s going to the dentist in the morning.
1 *MOT: 去哪裡看牙齒?
‘Where did you go check your teeth?’
2 *LJW: 去 -: 醫院.
‘Went to the hospital.’
3 *MOT: 去醫院.
‘Went to the hospital.’
4 *MOT: 去萬芳醫院對不對?
‘Went to Wan Fang Hospital, right? ’ 5 *LJW: 嗯.
In example (3), LJW’s mother asked LJW about the incident of seeing dentist earlier in the morning. Instead of leaving LJW to describe the whole incident by herself, LJW’s
‧
mother eased the task by constantly asking questions which elicit just one piece of information at a time. It can be observed in the example that LJW continued catching up with her mother’s questions and providing relevant replies to her mother’s inquiries about the incident of seeing the dentist. The whole topic episode concerning seeing the dentist in fact continued for more than 40 speaking turns, to which the mother contributed mostly questions and LJW mostly responses.
While LJW adopted more responses (61.8%) than LIN (52.8%) when maintaining topics, LIN adopted slightly more questions (6.3%) and much more requests (16.7%) than LJW (3.7 % and 7%). Compared to results in previous study (Wanska & Bedrosian, 1986; Huang, 2004), LIN’s frequent performance of maintaining discourse topics by requesting is salient and thus deserves investigation.
(4) MOT is asking for the toy bear.
In example (4) , LIN’s mother was requesting for the toy bear that was brought in by the observer ‘Uncle’. LIN’s mother began with claiming the possession of the toy bear. After the possession was denied by LIN, the mother continued to reissue the possession by
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
providing evidence regarding the source from which the toy bear came. LIN then appealed to a request that displayed her refusal. In terms of communicative intents, frequent occurrence of requests in LIN’s topic maintenance might be explained in two aspects. From the child’s angle, the intention of requesting in topic maintenance implied that LIN was egocentric in some way that when she was interacting with her mother, she put more attention on her personal needs than the conversation itself. In other words, instead of eagerly participating in the conversational interaction, LIN tented to relate the present conversation to her own desires or intentions. From the angle of interacting with the mother, LIN’s mother’s intention to control her child’s physical behaviors might lead to the dyad’s less interest of participating in the conversation. That is, if the mother frequently directs the child to do things, it might turn out that continuing of a
conversation or maintaining of a discourse are overlooked or even ignored by the dyad.
In sum, in either points of view it suggested that compared with LJW, LIN’s intention of requesting revealed her greater interest in her own desires rather than being engaged in the conversational interaction. As for interaction of the two dyads, more discussions will be conducted after the conversational parameters are examined in section 4.3.
Besides a great number of responses among LIN and LJW’s topic-maintaining utterances, their second frequent uses, declaratives, are fairly equal (18.2% and 19.8%).
As reported in Huang’s (2004) study, prevalence of declaratives, termed as informatives in Huang‘s study, might be explained by the less interactive nature of informatives in the present point of age. This suggested that in the current case, being inadequately
interactive, the two three-year-olds children were not mature enough regarding the conversational skills and still required their mothers’ elicitation in the interaction.
‧ 國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
4.2.2 Topic Incorporation/Collaboration
Table 6 Results of two children’s topic incorporation/collaboration in topic-maintaining utterance
LIN LJW
Topic Incorporation 251(54.6%) 506(50.6%)
Topic Collaboration 209(45.4%) 493(49.4%)
Total 460(100%) 999(100%)
Concerning the information provided by the two children, results of two children’s topic incorporation/collaboration in topic-maintaining utterance are presented in table 6.
In half of the time, both LIN (54.6%) and LJW (50.6%) incorporated new information when maintaining topics at about age 3, suggesting that by means of the mother’s
elicitation they were already capable of taking propositions in the previous utterances as foundations and adding new materials to them. At this point, quantitative analysis of two the children’s topic incorporation/collaboration in topic-maintaining utterance suggested that LIN and LJW displayed nearly equal sophistication in providing new materials when maintaining discourse topics. Since understanding of children’s topic-maintaining
capacity cannot be done without exploring how new materials are added in continuing of discourse topics, qualitative analysis of LIN and LJW’s topic maintenance was conducted with respect to topic incorporation/collaboration.
At most of the times, maintaining of discourse topics was eased by their mother’s successive elicitations, such as questions in similar structure and content that were sometimes partially repetitive:
‧
(5) MOT and LIN are talking about voting for popular cartoon characters.
1 *MOT: 你早上不是說要投給 -: 皮什麼 +…
‘Didn’t you say in the morning that you would vote for someone named /pi/…?’
2 *LIN: ++ <皮卡丘> [= cartoon character].
‘Pokemon.’
3 *MOT: 啊妹妹投給誰?
‘Who did Sister vote for?’
4 *LIN: <kitty> [= cartoon character] .
‘Who did Mother vote for?’
10 *LIN: 也投給哈姆太郎.
‘Hamtaro, too.’
11 *MOT: 我也投給哈姆太郎.
‘I voted for Hamtaro, too.’
In example (5), LIN and her mother were talking about which cartoon character to vote for. By continuing asking which character did a specific person vote for, structure and content of the questions were partially repeated in line 3, 6 and 9 as ‘who did X vote for’.
Within the clear context where the discussion of which character to vote for was attended
‧
by the dyad, the verb ‘vote for’ could even be omitted by the mother in line 6 without causing any problem in LIN’s understanding. Given elicitations in repetition, the child could thus produce relatively more substantial contributions and maintain topics for more turns (Huang, 2004).
Generally, due to the unsophisticated conversational skills, maintaining of discourse topics for children relies heavily on mothers’ elicitations. As a result, the parents tend to become the dominant speakers in the parent-child conversation interaction. Intending to engage their children in the conversation, the parents would be likely to elicit their children with successive questions like what was illustrated above (Huang, 2004). Or, if their intentions are not to elicit their children’s verbal participation but to control their children’s physical behaviors, they may adopted a lot of directives or devote themselves to long monologue (McDonald & Pien, 1982). In one way or the other, the parents are usually the dominant speakers who control the development of topics. Nevertheless, in our data, situations where the child took good control of the developed topics were observed:
(6) LIN claims that she want to raise pigs at home.
1 *MOT: 一綾啊 # 那個你如果看到有那個小豬在跑 # 那是電視裡面的.
‘LIN, if you saw pigs running, that happened only in TV.’
2 *MOT: 那個不能養你知道嗎?
‘Do you know that you can not actually raise them?’
3 *MOT: 我們家哪有地方可以養.
‘There is no space for it in our place.’
4 *MOT: 養你們兩隻就受不了了.
‘Raising you two is enduring enough.’
5 *LIN: 養在下面好了.
‧
‘Just raise it down there.’
6 *MOT: 哪個下面啊?
‘Down where?’
7 *LIN: 不知道 # 養在門口好了.
‘I don’t know, then raise it at the entrance.’
8 *MOT: 養在門口啊?
‘Raise it at the entrance?’
9 *LIN: 嗯.
‘Yes.’
10 *MOT: 養在門口你不怕他跑走嗎?
‘Won’t you be afraid that it runs away?’
11 *MOT: 就不見啦.
In the above example, LIN expressed her great interest in raising a small pig at home. In the beginning, the mother discouraged LIN by appealing to several reasons and even mentioned that she already got LIN and her sister to raise. In spite of being strongly discouraged by her mother, LIN still continued to argue for raising the pig by providing several solutions that she thought would be workable. During the discussion, LIN kept contributing to the topic of raising the pig and gradually took control of the topic of raising the pig. A similar case was observed in another conversation of a pretend play where LIN and the mother were playing soldiers:
(7) LIN and the mother are playing soldiers.
‧
‘I still want to play the soldier.’
3 *MOT: 士兵 # 好.
‘Soldier, fine.’
4 *MOT: 請問士兵是…
4 *MOT: 請問士兵是…