• 沒有找到結果。

Chapter 3 Methodology

3.4. Pragmatic Strategies in Agreement

3.4.6. Concession

personal judgment. In other words, when the speaker in T2 agrees with a strong bias or applies a personal view to agree on T1, he or she is supporting the agreed party through agreement and showing empathy. In example (10), Speaker B in T2 uses a personal judgment on the different preferences between Speaker A and boys to agree on Speaker A’s decision. By the application of a strong bias in T2, Speaker B supports Speaker A’s opinion through agreement.

(10)

The strategy Concession applied when the agreeing party who has the opposite view at first concedes that the other speaker is totally or partially correct. In example (5), repeated here as example (11), Speaker A considers their friend is “super” ultra at

5 This is a metaphor to visualize how thick the notebook is.

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

first. But after the persuasion from Speaker B in T1, Speaker A concedes and partially admits that the friend grows mature by the term “也是啦.” By the application of concession, the original confrontation is averted.

(11)

A: (0)她超偏激的

T1B: ..唉唷她長大了啦..有啦她比較[[長大了]]

T2A: [[也是啦]][有啦]

Base on the previous studies, several linguistic features can be found frequently occurred in the structure of agreement, while some features which were considered as forms or structures are more like pragmatic strategies. For example, reasoning through elaboration in Baym’s study (1996) are more like the ways of speaking so that it is taken as pragmatic strategies in this study.

After the general patterns of linguistic features and pragmatic strategies of agreement are analyzed, how social constraint—gender can influence categories of agreement and pragmatic strategies of agreement would also be investigated. Three kinds of gender constraints would be examined: speaker’s gender only, hearer’s gender only, and both speaker’s and hearer’s gender.

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

Chapter 4

Data Analysis (1): Constructions of Agreement

This chapter presents the results of the quantitative analyses of agreement collected from the 8 conversations. In this chapter, general findings of constructions of agreement and the influence of gender are provided and analyzed. They can be generally divided into the following three parts: agreement tokens, categories of agreement, and degrees of agreement. Additionally, because in the previous studies (Pomerantz, 1984; Kotthoff, 1993; Kuo, 1994; Mulkay, 1985; Baym, 1996; Rattai, 2003), how the social factor—gender influences people’s constructions of agreement is left unexamined; the influences of gender will be presented after each single discussion on the usage of agreement. Three kinds of influences of gender are made:

by speaker’s gender, by hearer’s gender, and by both speaker’s and hearer’s gender.

4.1. Agreement Tokens as a Whole

152 tokens of agreement are found in the collected data. When the influence of gender is taken into consideration on agreement tokens, the following sections present some findings.

4.1.1. Agreement Tokens as a Whole by Gender

This section is divided into three parts: agreement tokens by speaker’s gender, by hearer’s gender, and by both speaker’s gender and hearer’s gender. The distributions of the 152 tokens of agreement by gender are summarized in Table 4. In this table,

“MM” stands for Male to Male, “MF” stands for Male to Female, “FF” stands for Female to Female, and “FM” stands for Female to Male. Related analyses and discussions are given after the presentation of Table 4

Table 4. Agreement tokens as a whole by gender

(MM= Male to Male; MF= Male to Female; FF= Female to Female; FM=

Female to Male; -=No significant difference found in any two of these four gender groups.)

Agreement

Gender orientation % (Frequency) P

Speaker's

4.1.1.1. Agreement Tokens as a Whole by Speaker’s Gender

In Table 4, results of the statistic tests indicate that speaker’s gender is not a significant factor that would determine the emergence of agreement in face-to-face conversation. This result is against the findings in many previous studies, which claim that women would agree with their interlocutors more frequently than men do (Kalcik, 1975; Leet-Pellegrini, 1980; Edelsky, 1981; Coates, 1989; Holmes, 1995: 60). It seems that superficially, male and female speakers do not differentiate from each other on the tokens of agreement. However, when the investigation goes deeper on how the two genders construct agreement (see section 4.2 and 4.3), gender differences are revealed.

4.1.1.2. Agreement Tokens as a Whole by Hearer’s Gender

Like the consequence of speaker’s gender, Table 4 also indicates that hearer’s gender is not a significant factor to influence speakers’ production of agreement. In other words, despite of hearer’s gender, the subjects yield similar amount of

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

agreement tokens. However, the influence of hearer’s gender is located in speakers’

choice of linguistic devices and pragmatic strategies (see section 4.2, 4.3, and chapter 5).

4.1.1.3. Agreement Tokens as a Whole by Both Speaker’s and Hearer’s Gender Although FM, on the surface, seems to use less agreement than the other three groups as what are shown in Table 4, according to the results of statistic tests, no significant gender difference is located in any two of the four gender groups. Neither in same gender groups (i.e. MM and FF) nor in cross gender groups (i.e. MF and FM) is gender found to be an influential factor. However, again, it is the ways which the groups choose their linguistic devices and pragmatic strategies that show significant gender difference.

4.2. Categories of Agreement

In the following sections, agreement structure and comparisons among these categories of agreement are examined. One by one, the categories of agreement are presented first, followed by comparisons among these categories of agreement, and in turn followed by comparisons among their subcategories. To be specific, this section includes: (1) Head Act Alone (HA), (2) Supportive Moves Alone (SM), (3) Head Act with Supportive Moves (HA+SM), (4) Comparisons among Head Act Alone (HA) vs.

Supportive Moves Alone (SM) vs. Head Act with Supportive Moves (HA+SM), (5) Subcategories of Supportive Moves (SM), (6) Subcategories of Head Act with Supportive Moves (HA+SM), and (7) Inventory of All Six Subcategories of Agreement. Besides, the influence of gender will be presented after each of the section above.

Because the terminology of these categories of agreement is very copious, in the following sections, their abbreviations are used to replace them. For reader’s reference, please see List of Abbreviations in page xvi.

4.2.1. HA (Head Act Alone)

This section presents how many tokens of HA are found and its gender distributions.

4.2.1.1. HA by Subjects as a Whole

In this study, 65 tokens of HA are found in the collected data. The following sections introduce how gender influences HA’s distribution.

4.2.1.2. HA by Gender

In this section, the usages of HA by speaker’s gender, by hearer’s gender, and by the four gender matrix are analyzed. Table 5 below shows the distributions of HAs by gender. Following Table 5, related analyses and discussions are given.

Table 5. Head acts alone by gender

(Numbers in parentheses are frequencies; -=No significant difference found in any two of these four gender groups.)

Head Act

Gender orientation % (Frequency) P

Speaker's

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

4.2.1.2.1. HA by Speaker’s Gender

According to statistic results shown in Table 5, speaker’s gender is not an influential factor to people’s choice of using HA to express agreement. It seems that both male speakers and female speakers consider using agreement marker alone as the most conventional way or most efficient way to make agreement. The pragmatic goal and the social goal of making agreement are convivial. It means that there is no threat to hearer’s positive face (Brown and Levinson, 1978), and thus, it is appropriate for people to abundantly express agreement with a direct speech act. Therefore, male and female speakers do not differentiate from each other on frequently proffering HAs.

4.2.1.2.2. HA by Hearer’s Gender

In Table 5, statistic results indicate that hearer’s gender alone is not a significant factor to influence the frequencies of the speakers’ use of HA. As mentioned above in 4.2.1.2.1, when it comes to Brown and Levinson’s face theory (1978), agreement is as a speech act with no harm to hearer’s face so that people can perform it directly and simply. AM is the type which express agreement efficiently and effectively. Therefore, people’s preference of applying AM is not affected by hearer’s gender, either.

4.2.1.2.3. HA by Both Speaker’s and Hearer’s Gender

According to statistic results given in Table 5, no significant gender difference is found in any two of the four gender groups. It means that when both speaker’s and hearer’s genders are taken into consideration at the same time, the speakers, male as well as female, use similar amounts of HA to show consensus to their interlocutors despite their gender. Like what have been discussed above, it is probably because that HA is the most efficient way to express agreement and to fulfill hearers’ positive face wants (Brown and Levinson, 1978), so that all four gender groups apply it frequently.

4.2.2. SM (Supportive Moves Alone)

In this section, tokens of SM and how gender influences distributions of SM are investigated.

4.2.2.1. SM by Subjects as a Whole

54 tokens of SM are found in 8 conversations collected in this study.

4.2.2.2. SM by Gender

After the discussion on the usage of HA, this section introduces the frequencies of using supportive moves by speaker’s gender, by hearer’s gender, and by both speaker’s and hearer’s genders. The distributions of supportive moves as a whole by gender are presented in Table 6. Related analyses and discussions are given after the presentation of Table 6.

Table 6. Supportive moves alone by gender

(SM= Supportive Moves; Numbers in parentheses are frequencies; *=P<.05) SM Alone

Gender Orientation % (Frequency) P

Speaker's

4.2.2.2.1. SM by Speaker’s Gender

Statistic result in Table 6 shows that men and women do not differ from each other on their choices of SM. Agreement made by SM is as an indirect speech act.

Because the interlocutors in this study are either close friends or couples, both male

and female speakers may think hearers can receive their intention of agreement based on much of shared background knowledge and the knowledge of CP through inference (Grice, 1975; Searle, 1975). Thus, male and female speakers both frequently apply SM.

4.2.2.2.2. SM by Hearer’s Gender

According to Table 6, statistic results indicate that, when hearer’s gender alone is considered, no significant difference is found between the SMs received by male and those by female hearers. Like what have been mentioned above, people may think that as close friends or lovers to themselves, male and female hearers have ability to make inference based on mutually shared background knowledge. Hence, both male and female hearers frequently receive SM.

4.2.2.2.3. SM by Both Speaker’s and Hearer’s Gender

When both speaker’s and hearer’s genders are considered, Figure 3 below shows percentages of SMs used by four gender groups.

Figure 3. Percentages of supportive moves by both speaker’s and hearer’s gender

0.0%

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

According to Table 6 and Figure 3, FF uses SM more frequently than the other three gender groups. Furthermore, Table 6 reveals that the only significant gender difference is located in FF’s and FM’s frequencies of using SM (P=.032). It means that female speakers’ usage of SM is significantly influenced by hearer’s gender. In other words, women are more willing to use SM to show agreement in same-sex conversations than in cross-sex conversations. Perhaps it is because that for women, the purpose of using SM is to show involvement in conversations. Female speakers may want to express because they fully comprehend what the previous speaker says, they can build on the previous context by SM (such as by extension, specification, explanation and so on). By showing involvement and support, women could establish solidarity between selves and other (Tannen, 1986). In same-sex contexts, women presume that their interlocutors, like themselves, would put emphasis on establishing solidarity and avoiding disagreement (Tannen, 1990). By contrast, women may think that male hearers may not value solidarity as much as female hearers do. According to Tannen, solidarity is a drive to be friendly, and is related to symmetrical relationship.

It means that in female-female conversations, women’s tendency to show social equality and similarity between self and other is stronger than in cross-sex conversations. Therefore, female speakers use more SM to agree with female hearers than with male hearers.

4.2.3. HA+SM (Head Act with Supportive Moves)

In this section, the findings of HA+SM are presented by subjects as a whole and by the influence of gender.

4.2.3.1. HA+SM by Subjects as a Whole

In the collected data, 33 tokens of HA+SM are found.

Table 7 below presents the distributions of HA+SM by speaker’s gender alone, by hearer’s gender alone, and by both speaker’s and hearer’s genders. Related analyses and discussions are given after the presentation of Table 7.

Table 7. Head act with supportive moves by gender

(HA+SM= Head Act with Supportive Moves; -=No significant difference found in any two of the four gender groups)

HA+SM

Gender Orientation % (Frequency) P

Speaker's

4.2.3.2.1. HA+SM by Speaker’s Gender

Distributions of statistic test indicate that speaker’s gender is not an influential factor that would affect their use of HA+SM. To be specific, both male speakers and female speakers apply few HA+SMs. When speech act theory is considered (Searle, 1975), HA is a direct manifestation of the illocutionary force of agreement. Perhaps this is because once HA, which is already a clear indication of agreement, emerges, the addition of SM to HA is redundant, for both male and female speakers.

4.2.3.2.2. HA+SM by Hearer’s Gender

When hearer’s gender is examined, the pattern similar to those in the preceding section is found. That is, no significant difference is located. In other words, HA+SMs received by male hearers and female hearers are equally few. Like what have been

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

mentioned above, people may think for their interlocutors no matter as male or female, SM behind HA sounds redundant. Thus, HA+SMs are not very frequently received by both male and female hearers.

4.2.3.2.3. HA+SM by Both Speaker’s and Hearer’s Gender

When both speaker’s and hearer’s genders are taken into consideration, again, no significant differences are located in the comparison between any two of the four gender groups. In other words, for the use of HA+SM, gender is not an influential factor. Again, it is probably because that HA alone is clear enough to express agreement, and thus, it is not necessary to have SM behind.

4.2.4. HA vs. SM vs. HA+SM (Head Act alone vs. Supportive Moves Alone vs.

Head Act with Supportive Moves)

This section compares distributions of HA, SM, and HA+SM to see which one is preferred by Mandarin speakers. And this section can be divided into the following three sub-sections: (1) HA vs. SM, (2) HA vs. HA+SM, and (3) SM vs. HA+SM.

After each comparison among HA, SM, and HA+SM by subjects as whole, the influence of gender is discussed. And because HA alone by gender, SM alone by gender, and HA+SM by gender have been discussed in the previous sections, they are not repeated in this section.

4.2.4.1. HA vs. SM

This section depicts comparisons between HA versus SM. After the discussion of HA vs. SM by subjects as a whole, the influence of gender is examined.

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

4.2.4.1.1. HA vs. SM by Subjects as a Whole

Table 8 presents the comparisons between HA and SM and the related statistic result. Following Table 8, related analyses and discussions are given.

Table 8.Comparisons between head act alone and supportive moves alone (HA= Head Act; SM=Supportive Moves; Numbers in parentheses are frequencies.)

HA SM TOTAL P

54.6% (65) 45.4% (54) 100.0% (119) .333

According to the data in Table 8, over half of data are AMs. However, according to statistic results, HA is not significantly different from SM (P=.333). It means that HA and SM are both frequently applied to make agreement. It is an unexpected result.

In this study, it is hypothesized that HA is more preferred than SM (see Hypothesis A-1 in page 2). This result indicates that Hypothesis A-1 is not verified.

HA, as the core of agreement act, is more explicit and more effective, while SMs are only adjuncts to modify the force of agreement. At first, it is presumed that people should have applied much more HA than SM. One possible reason to explain why this presumption is overruled is that SM, although indirect, could still effectively express agreement through inference. As mentioned in 4.2.2.2, when the speaker makes an agreement by repeating the agreed evaluation, by reasoning, or by adding extra information for the discussed referent, the hearer can receive the speaker’s intention based on their mutually shared background knowledge, the knowledge of CP, and the ability to make inference (Grice, 1975; Searle, 1975). Because interlocutors in 8 conversations are either close friends or couples, they must have shared much background knowledge. Thus, even though agreement is frequently sent by SM, which is more indirect than HA, hearers can receive speakers’ intention through inference.

This section discusses the comparisons between HA and SM by speaker’s gender, by hearer’s gender, and by both speaker’s and hearer’s genders.

1. HA vs. SM by Speaker’s Gender

Table 9 presents the distribution of the frequencies of HA and SM by speaker’s gender alone, by hearer’s gender alone, and by both speaker’s and hearer’s genders.

Related analyses and discussions are given after the presentation of Table 9.

Table 9. Head act alone and supportive moves alone by gender

(HA= Head Act; SM=Supportive Moves; Numbers in parentheses are frequencies; *=P<.05; -=No significant difference found in comparison between HA versus SM or in any two of the four gender groups.) Categories

(1) According to Table 9, significant difference is located in the comparison between men’s HA and SM (P=.031). To be specific, men use significantly much more HA than SM. For men, efficiency of information exchange seems to be the first priority for communication. Additionally, men may not put much emphasis on interpersonal relationship rhetoric. Hence, male speakers prefer to use HA which expresses agreement directly and simply.

(2) When HA and SM are compared, no significant difference is found in women’s usage. It means that women frequently apply both HA and SM. Perhaps for

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

women, HA and SM, which are both important, serve different functions. HA can be used to express agreement efficiently. By contrast, as mentioned in 4.2.2.2.3, SM may be taken as a special way to show involvement and thus to establish solidarity or rapport. Women may want to express that because they listen carefully to what the previous speakers say, they can repeat or elaborate the previous contexts by SM. Thus, female speakers frequently use either HA or SM for efficiency of showing agreement and for showing involvement to establish solidarity which is revered by women, according to Tannen (1990).

2. HA vs. SM by Hearer’s Gender

When the social factor changes to hearer’s gender, as Table 9 indicates, no significant difference is located. To be specific, when talking to men as well as women, the amount of HA and that of SM used by the speakers are not significantly different. Perhaps it is because in this study, interlocutors are close friends or couples who share lots of background information with each other; therefore, hearers in both

When the social factor changes to hearer’s gender, as Table 9 indicates, no significant difference is located. To be specific, when talking to men as well as women, the amount of HA and that of SM used by the speakers are not significantly different. Perhaps it is because in this study, interlocutors are close friends or couples who share lots of background information with each other; therefore, hearers in both