Chapter 2 Literature review
2.4 Control acts
2.4.3 Politeness
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
“Stop that” (Lampert & Ervin-Tripp, 1993). The evidence shows that many prohibitions on a child’s behavior cause negative outcomes (Maccoby, 1980;
Maccoby & Martin, 1983).
Gleason, Ely, Perlmann, and Narasimhan (1996) examined variation in parents’
use of prohibitions with their daughters or sons. They discovered three findings. One was that parental prohibitions in the play sessions declined as children’s ages
increased. Another was that boys tended to hear more repeated or clustered
prohibitions than girls. The other was that prohibitions were more commonly occurred at dinner time.
Schaffer and Crook (1979, 1980) found that only about 4% of the controls were prohibitions during a laboratory-directed play situation. The majority of the controls that the mothers in their studies used were to tend to propose a new activity instead of prohibiting the child’s present activity. Like Schaffer and Crook, McLaughlin (1983) also found that the vast majority of control utterances were directives rather than prohibitions in nature. Only 5 % of all controls were prohibitions for both mothers and fathers. Thus, the use of prohibitions was much rarer than that of directives.
2.4.3 Politeness
Since control acts are utterances designed to bring about a change in the other’s behaviors, they are inherently face-threatening (Brown & Levinson, 1987). Thus, politeness becomes a major consideration (Blum-Kulka, 1990). According to Brown and Levinson, three parameters need to be considered during any face-threatening act produced: (1) the ranking of the degree of imposition; (2) the social distance between a speaker and an interlocutor; and (3) the power differential between a speaker and an interlocutor. In earlier studies on traditional directness perspective, directness was
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
associated with impoliteness and indirectness with politeness (Ervin-Tripp, 1976;
McLaughlin, Schutz, & White, 1980). However, this perspective has been rejected for the reason that politeness cannot be equated with indirectness.
Blum-Kulka (1990) considers that the contribution of mitigation to politeness was placed as secondary to indirectness in the literature. In addition, according to Blum-Kulka’s view of Brown and Levinson’s model, the different forms of mitigation are considered as sub-strategies of positive and negative politeness and do not justify the centrality of mitigation in indexing politeness, at least for family discourse.
Therefore, studies in analyzing the use of the politeness need to consider the overall mitigation (Blum-Kulka, 1990; Aronsson & Thorell, 1999).
Two studies were conducted on politeness by examining the proportions of mitigation in mothers’ regulatory language. Halle and Shatz (1994) found that only 16% of British mothers’ regulatory language is mitigated, so they consider that British mothers seem not to favor polite, mitigated, or indirect forms. Blum-Kulka (1990) examined parental speech acts of control around the dinner table in middle-class Israeli, American, and American-Israeli families. The finding show that the language of parental control is richly mitigated, so family discourse is essentially polite (39% in Israeli, 32% in American immigrants, and 26% in American). The author found that mitigated directness is used to redress the hearer’s positive face in the context of family discourse. The term mitigated directness represents solidarity politeness according to the term of Scollon and Scollon (1981), or positive politeness according to the term of Brown and Levinson (1987). Although forms of indirectness encode a self-face-saving element which allows for the denial of requestive intent, the use of mitigated directness is hearer-oriented in order to enhance the hearer’s positive face
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
by appealing to in-group membership or by giving reasons and justifications.
Blum-Kulka (1990) proposes an index of politeness: (1) IMPOLITE, whose point is to completely disregard face-needs by the use of aggravated-directness; (2)
NEUTRAL, whose point is to not to index directness either politeness or impoliteness in the domain-specific requirements of the family code; (3) MITIGATED
DIRECTNESS, whose point is to take any forms of mitigation; (4) HINTS, whose point is to use nonconventional indirectness to show regard for a child’s face.
‧
3.1 Subjects and dataIn this study, the subjects were a Mandarin-speaking boy and his mother. They live in Taipei City, Taiwan. H2 is the only child in the family, and his mother, M, has a master’s degree and works in advertising. The data of M’s regulatory language was analyzed when H’s ages were 2;1, 2;7, 3;1, and 3;73. In the collected data, we found Mandarin Chinese was mainly spoken. Southern Min and English were used
occasionally.
The data examined in the present study were adopted from Professor
Chiung-chih Huang’s database4. The data involved 4-hour audio- and video-taped natural interactions between the child and his mother. The interactions were
transcribed in the CHAT (Codes for the Human Analysis of Transcriptions) format.
The data was coded and further computed in the CLAN (Child Language Analysis) program. Only maternal regulatory speech was coded for the present study. During the observation, the mother and child engaged in various activities, including playing cars, reading books, drawing, watching TV, and eating.
3.2 Procedures of data analysis
All maternal utterances that involved initiating, modifying, avoiding or stopping a child’s behavior were identified. Identification was made by viewing the video
records and reading the transcripts. The identified utterances were then analyzed and coded according to the analytical framework listed in Section 3.3. CLAN was then used to compute the frequencies. In this study, the results will be presented in the
2 H and M are subject codes.
3 H’s four ages are based on subjects’ ages in related studies (e.g., Schneiderman, 1983; Kuczynski &
Kochanska, 1995).
4 I am deeply grateful for Professor Huang’s generosity and kindness in allowing me to make use of the data.
‧
form of descriptive statistics to describe or characterize the data.
3.3 Coding system
This study focuses not only on maternal action directives but also action
prohibitions. Thus, utterances which are intended to initiate, modify, avoid or stop a child’s behavior will be investigated. Following Schneiderman (1983), fragments (e.g., On the table), stock expressions (e.g., Come on), calls for attention (e.g., See, child’s name), false starts (e.g., Do you want…uh…eat the…), test questions (e.g. Can you say cookie?), and routines5 (Where’s your mouth?) are excluded from analysis.
In the present analysis, each coding tier is composed of four levels to examine maternal regulatory language. The first level divides the types of control acts produced by the mother into directives and prohibitions. The second level indicates the types of syntactic directness. The third level indicates the semantic modifications.
The fourth level indicates the content. In other words, maternal regulatory language is
studied by analyzing maternal control acts and the distributions of syntactic directness, semantic modification, and content under two maternal control acts, as shown in Figure 2.
Figure 2. The coding system of control acts
The details of each coding category will be given and explained in the following
5 According to Schaffer and Crook (1979), another form of attention control belonging to interrogative attention controls mostly involved “Where is…” and “What is…” questions.
Syntactic directness
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
will be described in Section 3.3.2. Semantic modification will be described in Section 3.3.3. Content will be described in Section 3.3.4. Coding system and reliability will be described in Section 3.3.5.
3.3.1 Maternal control acts
Although Lampert and Ervin-Tripp (1993) proposed six types of control acts , as mentioned in Section 2.4, this research only focuses on directives and prohibitions to investigate how a Mandarin mother regulates her child’s behavior. Two types of control acts examined in this study are shown as below. Certain of the definitions are modified according to Searle (1976) and Olsen-Fulero (1982).
(1) Directives
Directives refer to utterances which elicit a hearer’s physical behavior
(Olsen-Fulero, 1982) to act and offer either goods or services. The speaker attempts to get the hearer to do something (Searle, 1976). Attention directives are not included in this study. Only action directives are analyzed. For example, Lai. ‘Come.’
(2) Prohibitions
Prohibitions refer to utterances which constrain a hearer’s physical behavior (Olsen-Fulero, 1982). The hearer is asked to avoid doing or to stop performing an undesirable behavior (Lampert & Ervin-Tripp, 1993). For example, Ni buyao zai qu wan. ‘Don’t play again.’
3.3.2 Syntactic directness
Syntactic directness measures how explicitly the structure of a sentence indicates that it is a control act and what that control act entails. In this study, the system of analysis of maternal syntactic directness is mainly adopted from Rue and Zhang’s
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(1975), Schneiderman’s (1983), Gao (1999), and Hays, Power, and Olvera’s (2001) classifications.
The following arrangement of three types of syntactic directness is from the most to the least direct, and each type is further classified into sub-categories.
(1) Direct
The direct category includes language where the speaker’s apparent intent is conveyed to the hearer. The sub-categories for direct, including imperatives,
performatives, obligation statements, and want statements are directly adopted from Rue and Zhang’s (2008).
(a) Imperative6
Imperatives conventionally signal a control act. The illocutionary force is explicitly marked, and the force is transparent. In Mandarin, the imperative consists of a subject ni ‘you’ or a subjectless verb phrase to describe the action. For example, Bu yao hua! ‘Don’t draw!’
(b) Performative
The speaker expresses the illocutionary intent by using an overt
illocutionary verb such as the giving of an order and the making of plea or begging. Gao (1999) has found that compared with English, speakers of Chinese have more performative verbs such as rang ‘let’, yaoqiu ‘request’, qingqiu ‘sincerely request’, kenqiu ‘plead’, qiu ‘beg/ask’, qiuqiu ‘pleadingly ask’, zhishi ‘direct’, mingling ‘order’, and jiao ‘ask’. For example, Wo
mingling ni bu yao hua. ‘I order you not to draw.’ In the example, there is an
6 One kind of imperative which directly names the desired object is not discussed in this study. This follows Schneiderman’s (1983) exclusion of fragments for analysis.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
obvious illocutionary verb mingling ‘order’ before the imperative sentence ni bu yao hua ‘don’t draw.’ Thus, the construction of the performative is formed by an illocutionary verb with an imperative sentence.
(c) Obligation statement
The illocutionary intent is conveyed by directly stating moral obligation.
The verb yinggai ‘should’ is typically used. For example, Ni yinggai zao dian huilai. ‘You should come back earlier.’
(d) Want statement
The speaker conveys a particular want, need, desire or wish. The verbs xiang ‘want’ and yao ‘need’ are typical usages. For example, Wo yao heshui.
‘I want to drink.’
(2) Conventionally indirect
The definitions of conventionally indirect are adopted from Rue and Zhang (2008). The conventionally indirect category includes language where the speaker induces the addressee’s compliance with regard to a desired act by invoking the ability or willingness of the addressee. The conventionally indirect category is produced in the form of questions. In fact, a question about ability or willingness is conventional enough to be recognized as a control act after undergoing an idiomatic process.
The question category includes requests for information (yao-bu-yao
‘want-not-want’, yao…ma ‘do you want’) and embedded imperatives (neng-bu-neng, ke-bu-keyi, keyi ‘can you’) with a focus on the listener’s activity when the speaker elicits specific information which he or she does not have but wants by the use of questions.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(3) Non-conventionally indirect
The non-conventionally indirect category includes language where the speaker states partially relevant acts rather than using explicit statement to express the desired act. The language in the non-conventionally indirect category is produced in the form of questions or declaratives. The language in the non-conventionally indirect category consists of hints.
The process of understanding non-conventionally indirect is more complex than direct, so Searle (1975) proposed 10 steps in comprehending an non-conventional indirect. Searle mentioned that in order to derive the speaker’s meaning in
non-conventionally indirect, the listener has to rely on nonlinguistic and linguistic background information, use of the principles of conversation, and make inferences about the speaker’s intention. Although the illocutionary intent is not overtly
expressed, the speaker provides clues (Rue & Zhang, 2008) or statements of condition (Hays, Power, & Olvera, 2001) for the hearer to deduce the content. The form is non-imperative and the desired action is not named (Schneiderman, 1983). For example, Zhe ni de lianluobu. ‘This is your contact book.’ (Intent: asking the hearer not to draw on the contact book.)
3.3.3 Semantic modification
Words or phrases that change the meaning of control acts and cannot be described in terms of syntactic directness will be classified into semantic modification.
The system of maternal semantic modification is described based on Hays, Power, and Olvera’s (2001) and Rue and Zhang’s (2008) classifications. The categorization in these two studies includes mitigation and aggravation parts, and
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
there is no categorization for bare semantic modification. Thus, in order to satisfy the concept of a mitigation-aggravation continuum, another category of bald from Brown and Levinson (1987) is added. Maternal semantic modifications are shown as below.
(1) Mitigation
Mitigation means that speakers mitigate or soften control acts. The following are six sub-categories in mitigation.
(i) Minimization
Adverbial modifiers are used to diminish or under-represent the task. The typical lexical forms are understaters (yidian ‘a little’, yixie ‘some’), delimiters (zhiyou ‘only’), and particles ne, le, ba, ma, o, luo, a, la, and ye
‘particles’.
(ii) Politeness markers
Politeness markers (qing, baituo, qiu ‘please’) are added to seek cooperation from the hearer.
(iii) Communal orientation
An inclusive women ‘we’ form is used, and the speaker means to solicit the approval or agreement of the hearer.
(iv) Tag question
Tag questions (…xing ma?, …keyi ma?, ...hao ma?, …haobuhao, … xingbuxing? ‘…OK?’) are typically used at the end of an utterance to elicit agreement.
(v) Justification
The speaker provides reasons, explanations, and justifications.
(vi) Bargain
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
The speaker promises a reward to get the hearer’s compliance.
(2) Bald
The character of a bald utterance is that of one without mitigation or aggravation.
As long as the three types of syntactic directness in 3.2.2, are not modified by any mitigation or aggravation categories, they will be coded as bald.
(3) Aggravation
Aggravation means that speakers increase the coerciveness of control acts. The following are three sub-categories in mitigation.
(i) Immediacy
Time limited or constrained phrases (ganjin, gankuai, kuai ‘hurriedly’, mashang ‘right now’) are used to stress urgency.
(ii) Repetition
A control act utterance is repeated literally or paraphrased. Those repetitions occurring immediately after the original utterance in the transcript are coded.
(iii) Threat
The hearer is threatened with punishment if he/she does not comply with the control act.
3.3.4 Content of maternal control acts
To obtain a more complete and systematic investigation of maternal regulatory language usages in varied contents, we will adopt Kuczynski and Kochanska’s (1995) classification and state the categories of content as follows.
(1) Competent action (a) Prosocial action
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Benefit another person or mutual benefit including taking on caretaking responsibilities or doing chores. (e.g., “Take care of your sister.” “Share your candies.” “Put the magazine back for me.” “Set the table.” “Get the
cookies.”) (b) Cognitive/play
Regulation of a child’s cognitive and play activities (e.g., “Look, what’s this for?” “Let’s play with that.” “Read this book.”)
(2) Appropriate behavior (a) Social regulation
Regulation of a child’s personal and interpersonal behaviors and also performance of social interaction and social conventions (e.g., “Don’t hit.”
“Keep quiet.” “Say thank you.” “Tell Cindy about the present.”) (b) Use of objects
Protection of objects and the environment from being affected by messy or inappropriate behavior (e.g., “Don’t draw on the floor.” “Don’t spill juice on the table.” “Stop throwing crayons around.”)
(3) Caretaking (a) Physical care
Care about dress, cleanliness, and feeding. (e.g., “Wash your hands.” “Eat your noodles.”)
(b) Monitoring
Regulation of a child’s location, orientation, and physical safety (e.g.,
“Careful on the stairs.” “That’s dangerous, don’t.” “Come here.” “Play in the other room.”)
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
3.3.5 Coding system & Reliability
Directives (1) Maternal control acts Prohibitions
____________________________________________________________________
Imperative Direct Performative
Obligation statement (2) Syntactic directness Want statement
Conventionally indirect Non-conventionally indirect
____________________________________________________________________
Minimization Politeness marker
Mitigation Communal orientation Tag question
Justification Bargain (3) Semantic modification Bald
Immediacy Aggravation Repetition Threat
____________________________________________________________________
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Competent action Cognitive/play Social regulation
(4) Content Appropriate behavior Use of objects Physical care Caretaking Monitoring
Figure 3. The four classifications of maternal regulatory language: Maternal control acts, Syntactic directness, Semantic modification, and Content
Sixty-minute transcriptions were selected to be coded by another Mandarin Chinese speaker. Cohen’s Kappa was used to estimate the inter-rater reliability of the coding transcripts. In the present study, the reliability for maternal control acts
reached 0.87, for syntactic directness reached 0.87, for semantic modification reached 0.84, and for content reached 0.92.
‧ 國
立 學
‧
N a tio na
l C h engchi U ni ve rs it y
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter 4 Results
In this chapter, we will seek to discover the distributions of two types of Mandarin maternal control acts (Section 4.1) and three distributions of syntactic directness (Section 4.2), semantic modification (Section 4.3), and content (Section 4.4) under two maternal control acts, namely directives and prohibitions.
4.1 Two types of Mandarin maternal control acts
Two types of control acts investigated in this study refer to directives and prohibitions. Table 1 displays the frequency of directives and prohibitions after analyzing maternal utterances to H at four ages.
Table 1
The frequency of directives and prohibitions by H’s ages
Directives Prohibitions Total
Age N. (%) N. (%) N. (%)
2;1 91 (81.3) 21 (18.8) 112 (100.1)
2;7 52 (77.6) 15 (22.4) 67 (100)
3;1 65 (75.6) 21 (24.4) 86 (100)
3;7 36 (69.2) 16 (30.8) 52 (100)
Total 244 (77) 73 (23) 317 (100)
As shown in Table 1, the data identified includes 317 control acts, of which, 244 are directives and 73 are prohibitions. The number of the directives is more than
‧
three times that of the prohibitions. Thus, directives comprise most of the maternal control acts. This finding is correspondent with that of Schaffer and Crook’s (1979) study where the majority of maternal controls belonged to directives rather than prohibitions. Kuczynski and Kochanska (1995) might explain the results. They propose that directives rather than prohibitions are considered to be beneficial maternal control acts. Therefore, the mother chooses to use more directives rather than prohibitions to regulate her child.
As for directives, in the previous chapter it was mentioned that directives are used by the speaker to get the hearer to perform an act. According to the results listed in Table 1, the mother was found to employ 244 directives. Among 244 directives, 91 directives (81.3%) were identified at 2;1, 52 directives (77.6%) at 2;7, 65 directives (75.6%) at 3;1, and 36 directives (69.2%) at 3;7. The results show that the frequency of directives decreases with H’s ages. Example (1) is extracted from H at 2;1 to illustrate the use of directives.
(1) M is asking H to drink water.
Example (1) displays three directives in Lines 1, 3, and 4. The mother asked the child to drink water (Line 1). Then, she regulated the child’s orientation to be closer to the cup (see Line 3). As soon as the child drank some water, she further ordered him to hold the cup by himself. This example clearly shows how the mother gave three directives to get the child to perform the act.
‧
The other type of control acts discussed in this study is prohibitions. Prohibitions are used by the speaker to get the hearer to stop, avoid, inhibit or prevent undesirable behaviors. Among 73 prohibitions observed in Table 1, there were 21 prohibitions (18.8%) at 2;1, 15 (22.4%) at 2;7, 21 (24.4%) at 3;1, and 16 (30.8%) at 3;7. The results show that the frequency of prohibitions increases with H’s ages. It seems that the results show developmental patterns. The frequency of directives decreases with the child’s ages. On the other hand, the frequency of prohibitions increases. Example (2) from H at 3;7 shows the use of maternal prohibitions.
(2) M asks H not to pick up the video recorder.
%sit: H touches Ya-ting’s video recorder.
The mother ordered the child not to pick up the equipment in Line 1. In Line 2, she provided a reason for him to keep away from the video recorder. She warned him not to touch the equipment. If he touched the equipment, it would get broken. The mother stopped the child from touching the equipment.
4.2 Syntactic directness
As mentioned in the previous chapter, syntactic directness is divided into direct, conventionally indirect, and non-conventionally indirect by measuring how explicitly the structure of a sentence indicates that it is a control act. After analyzing the data for
As mentioned in the previous chapter, syntactic directness is divided into direct, conventionally indirect, and non-conventionally indirect by measuring how explicitly the structure of a sentence indicates that it is a control act. After analyzing the data for