General distributions of [s]% by the advancement

Chapter 4 Data Analyses

4.3 Distributions of [s]% by the Advancement of the Preceding Vowels

4.3.1 General distributions of [s]% by the advancement

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Tables 13 to 15, only one contrastively more conspicuous difference is located: the speakers of lower education level in Chungli show more [s]% in open syllable (33.4%) than in closed syllable (23.0%). In sum, formality is not powerful in any social/geographical group’s and any subgroup’s use of [s] in the two syllable structures.

4.3 Distributions of [s]% by the Advancement of the Preceding Vowels

This section examines the influences of the advancement of the vowels

preceding /ɕ/. These vowels are categorized as [-back, +back] (including [i], [y], [e], and [a]), [-back, -front] (i.e.[ə]) , and [+back, -front] (including [u], [o], [ʅ], and [ɿ]).

4.3.1 General distributions of [s]% by the advancement of the preceding vowels

It was expected that the advancement of the following vowels would influences

on the phonetic manifestation of /ɕ/. To be specific, /ɕ/ is more likely to be realized as [s] when the preceding vowels are [-back,+front] or [-front,-back]. The results of data analyses are given in Table 16.

Table 16. General distributions of [s]% by the advancement of the preceding vowels (numbers in patterns are frequencies of [s] )

According to Table 16, the subjects as a whole use more [s] when the vowels preceding /ɕ/ are [-back] than when the preceding vowels are [+back] (66.6% vs.

33.4%, respectively). Further, among the vowels that are [-back], those that are

[-back] [+back]

[+front] [-front] [-front]

42.6% (374) 24.0% (210)

33.4% (291) 66.6% (584)

100% (875)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

[+front] induce more [s] than those that are [-front] (42.6% vs. 24.0%,

respectively). These results indicate that the advancement of the vowels preceding /ɕ/

is influential to /ɕ/’s being realized as [s]. In other words, forward assimilation of place of articulation seems to be a persuasive explanation.

4.3.2 Distributions of [s]% by the advancement of the preceding vowels and formality

The distributions of [s] by preceding vowels is expected to be controlled by situational formality. Since in reading characters there is no preceding vowels, in this section, only the use of [s] in CON and that in RP are analyzed. Related data are given in Table 17.

Table 17. Distributions of [s]% by the advancement of the preceding vowels and formality (numbers in patterns are frequencies of [s])

According to Table 17, [s], in either style, as expected, appears more often after vowels that are [-back] than after those that are [+back]. Again, the

advancement of the preceding vowels is effective, despite stylistic difference.

Moreover, in conversation, [s]% is higher when after vowels that are either [-back, +front] or [+back, -front]; however, in reading passages, [s]% still

scores the highest after vowels that are [-back, +front], but [s]% is about the same

Formality Conversation Reading passages

Advancement

of Preceding vowels

[-back] [+back] [-back] [+back]

[+front] [-front] [-front] [+front] [-front] [-front]

[s]%

38.9%

(204)

22.3%

(118)

38.8%

(206)

49.1%

(170)

26.5%

(92)

24.4%

(85) 61.2%

(322)

75.6%

(262)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

when /ɕ/ comes after the other two kinds of vowels. In here, situational formality is still effective.

4.3.3 Distributions of [s]% by the advancement of the preceding vowels and social/geographical factors

This section examine the influence of vowel advancement to the use of [s]

under social controls: the four social/geographical factors. The results of data analyses are given in Table 18.

Table 18. Distributions of [s]% by the advancement of the precedingvowels and social/geographical factors (APV= Advancement of the

Preceding Vowels; numbers in patterns are frequencies of [s]) APV

Social/

Geographical factors

[-back] [+back]

[+front] [-front] [-front]

Place Taoyuan 41.0% (78) 27.3% (52)

31.5% (60) 68.3% (130)

Chungli 43.2% (296) 23.0% (158)

33.7% (231) 66.2% (454)

Age Old 43.4% (333) 22.7% (174)

33.9% (259) 66.1% (507)

Young 37.6% (41) 33.0% (36)

29.3% (32) 70.6% (77)

Gender Male 40.9% (134) 25.9% (84)

33.3% (109) 66.5% (218)

Female 43.7% (240) 22.9% (126)

33.2% (182) 66.6% (366)

Education Level

Higher Education

44.0% (195) 23.7% (105)

32.2% (143) 67.7% (300)

Lower Education

41.4% (179) 24.3% (105)

34.2% (148) 65.7% (284)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Table 18, when both the advancement of the preceding vowels and the social/geographical factors are taken into considerations, it is found that all social/geographical groups, except the younger group, share a pattern: [s]% is higher when /ɕ/ comes after vowels of [-back] than when /ɕ/ follows vowels of [+back].

However, when both [back] and [front] are considered, most of the eight

social/geographical groups, except the younger group, show high [s]% when the preceding vowels are [-back, +front], less when the vowels are

[+back, -front], and the least when the vowels are [-back, -front]. It seems reasonable to think that for the sake of assimilation, [s]% is higher when the

precedings are [-back, +front] than when the preceding vowels are [+back, -front].

However, it is unclear why [s]% is the lowest when the preceding vowels are [-back, -front]. In other words, it seems not the case that the fronter the preceding vowel is, the higher the [s]%.

4.3.4 Distributions of [s]% by the advancement of the preceding vowels, formality, and social/geographical factors

This section investigates the impact of vowel advancement on the distributions of [s] by the five external factors, (namely, formality, geographical factor, age, gender, and education level.) Related data are given in Table 19.

‧

Table 19. Distributions of [s]% by the advancement of the preceding vowels, formality, and social/geographical factors (APV= Advancement of the Preceding Vowels; the number in patterns are frequencies of [s])

According to Table 19, several patterns are located. First, when when the vowels preceding /ɕ/ are categorized into [+back] and [-back], all of the Formality

APV Social/

Geographical factors

Conversation Reading passages [-back] [+back] [-back] [+back]

[+front] [-front] [-front] [+front] [-front] [-front]

Place Taoyuan City

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

social/geographical groups show conspicuous stylistic variation. Next, when /ɕ/

following vowels that are [-back], all groups score considerably higher [s]% in RP than in CON. These findings are opposite to the researcher’s expectation that the percentage of non-standard forms would be decreased when formality increases.

Moreover, when the preceding vowels are [+back], all of the eight groups, as expected, decrease considerably their use of [s] when formality increases. This result may be explained by the fact that the density of the target characters is much higher in RP than in CON. For example, in RP, there is a sentence containing 9 target

characters at the same time.

4.3.5 Distributions of [s]% by advancement of the preceding vowels and the interactions of the social/geographical factors

This section examines the relationship between the impacts of the advancement of the vowels coming in front of /ɕ/ and the interactions of the external factors on the emergence of [s]. It was expected that subjects of different social backgrounds would use [s] differently when the advancement of the preceding vowels changes. Related data are given in Table 20.

‧

Table 20. Distributions of [s]% by the advancement of the preceding vowels and the interactions of social/geographical factors (APV= Advancement of Preceding Vowels; numbers in parentheses are frequencies)

APV

Social/Geographical factors

Geographical Area Age Gender Education Level

Taoyuan Chungli Old Young Male Female High Low

‧

Table 20 (continued)

APV

Social/geographical factors

Geographical Area Age Gender Education Level

Taoyuan

‧

Table 20 (continued)

APV

Social/geographical factors

Geographical Area Age Gender Education Level

Taoyuan

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Table 20, when the vowels preceding /ɕ/ are categorized by the feature of [back], it is noticed that none of the social/geographical groups and

subgroups show significant difference in conversation, as well as in their [s]% in front of the two vowel types. Exactly the same pattern is located when the vowels

preceding /ɕ/ are categorized as [+front] vs. [-front], and when the advancement of those vowels are grouped as [+front,-back], [-front,-back], and [-front,+back]. In other words, [s]% by the advancement of the preceding vowels is free from

social/geographical controls.

4.3.6 Distributions of [s]% by the advancement of the preceding vowels, formality, and the interactions of the social/geographical factors

This section examine the relationship between the use of [s], the advancement of the preceding vowels, and the interactions of social, geographical, and situational factors. Related data in CON are given in Table 21, and those in ready passages are listed in Table 22.

‧

Table 21. Distributions of [s]% in Conversation by the advancement of the preceding vowels and the interactions of social/geographical factors (

APV=

Advancement of Preceding Vowels;

numbers in parentheses are frequencies)

APV

Social/geographical

factors

Geographical Area Age Gender Education Level

Taoyuan Chungli Old Young Male Female High Low

‧

Table 21 (continued)

APV

Social/geographical factors

Geographical Area Age Gender Education Level

Taoyuan

‧

Table 21 (continued)

APV

Social/geographical

factors

Geographical Area Age Gender Education Level

Taoyuan

‧

Table 22. Distributions of [s]% in Reading Passages by the advancement of the preceding vowels and the interactions of social/geographical factors (APV=

Advancement of Preceding Vowels;numbers in parentheses are frequencies)

APV

Social/geographical

factors

Geographical Area Age Gender Education Level

Taoyuan Chungli Old Young Male Female High Low

‧

Table 22 (continued)

APV

Social/geographical facotors

Geographical Area Age Gender Education Level

Taoyuan

‧

Table 22 (continued)

APV

Social/geographical factors

Geographical Area Age Gender Education Level

Taoyuan

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Tables 21 and 22, the advancement of the preceding vowels are found not to be effective in the productions of [s] by subjects of different social backgrounds, in different styles.

4.4 Distributions of [s]% by the Preceding Nasals

This section examines the relationship between the distribution of [s] by preceding nasals,(namely, [n] and [ŋ]) and the external factors.

4.4.1 General distributions of [s]% by the preceding nasals

It was expected that the nasals preceding /ɕ/ would affect the subjects’ use of the nonstandard form [s], with [n] being more influential than [ŋ]. Related data are given in Table 23.

Table 23. General distributions of [s]% by the preceding nasals (numbers in parentheses are frequencies)

According to Table 23, [s] is used more frequently when following [ŋ] than when following [n]. This finding is against the researcher’s expectation that [s]% would be higher when following [n] as a result of forward assimilation of place of articulation;

after all, [n] and [s] are both alveolar consonants. It is unclear why such result emerges.

4.4.2 Distributions of [s]% by preceding nasals and formality

When situational formality is taken into consideration, it was expected that [s]

would be used more often in CON than in RP, and this tendency would strengthened when [s] follows [n]. There is no RC due to no preceding nasals in it. The results of

Preceding

Nasals [n] [ŋ] Total

[s]% 46.1% (175) 53.9% (204) 100% (379)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

data analyses are given in Table 24.

Table 24. Distributions of [s]% by the preceding nasals and formality (numbers in parentheses are frequencies)

According to Table 24, in either style, [s]% is moderately higher when the preceding nasal is [ŋ]. Also, no significant stylistic difference is located.

4.4.3 Distributions of [s]% by the preceding nasals and social/geographical factors

It was predicted that the use of [s] by preceding nasals would change with the subjects’ social backgrounds. The results of data analyses are given in Table 25.

Style Conversation Reading passages

Preceding nasals [n] [ŋ] Total [n] [ŋ] Total

[s]% 46.8%

(101)

53.2%

(115)

100%

(216)

45.4%

(74)

54.6%

(89)

100%

(163)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Table 25. Distributions of [s]% by the preceding nasals and social/geographical factors (numbers in parentheses are frequencies)

According to Table 25, all of the eight social/geographical groups share the same pattern in using [s]: [s]% following [ŋ] surpasses [s]% following [n]. Among all, Taoyuan group and the younger group show greater gaps between [s]% following the two nasals (21.0% and 23%, respectively), but the gaps are milder in the male the male group and the lower education group (11% and 13%, respectively); and the gap is even weaker in the other four groups. In conclusion, preceding nasals, in combination with subjects’ social backgrounds are not influential to subjects’ use of [s].

Preceding nasals Social/Geographical factors

[n] [ŋ] Total

Place Taoyuan City

39.7% (33) 60.3% (50) 100% (83) Chungli

City

47.9% (142) 52.1% (154) 100% (296) Age Old 47.5% (155) 52.5% (172) 100% (327) Young 38.5% (20) 61.5% (32) 100% (52) Gender Male 44.8% (64) 55.2% (79) 100% (143)

Female 47.0% (111) 53.0% (125) 100% (236) Education

Level

Higher Education 49.0% (90) 51.0% (94) 100% (184) Lower

Education

43.6% (85) 56.4% (110) 100% (195)

‧

4.4.4 Distributions of [s]% by the preceding nasals, formality, and social/geographical factors

It was predicted that [s]% would change when social/geographical factors and situational formality are taken into consideration together. The results of data analyses are given in Table 26.

Table 26. Distributions of [s]% by the preceding nasals, formality, and

social/geographical factors (numbers in parentheses are frequencies) Formality

Preceding Social/ nasals Geographical factors

Conversation Reading passages [n] [ŋ] Total [n] [ŋ] Total

Place Taoyuan City

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Table 26, when formality and social/geographical factors are taken into consideration at the same time, it is found that when the preceding nasal is [n], Chungli group, the older group, the female group, and the higher education group decrease their use of [s] as formality increases; while the other four groups behave the opposite. However, when [s] comes after [ŋ], it is Taoyuan group, the younger group, the male group, and the lower education group that decreases their use of [s] when formality increases.

Moreover, increase (or decrease) of [s]% behind [n] or [ŋ] are always given by the Taoyuan groups, not by the others. To be specific, Taoyuan group significant decreases its use of [s] when the preceding nasal is [ŋ], but it increases its use of [s]

when the preceding nasal is [n], with a gap over 20% in between. The former result is expected, but the latter is against the researcher’s expectation. This can be explained that their awareness increase in RP, their pronunciation will follow ease of articulation more; otherwise, they will feel some difficulties to read the passages.

4.4.5 Distributions of [s]% by the preceding nasals and the interactions of social/geographical factors

It was anticipated that the subjects’ social backgrounds would modify the subjects’ use of [s] behind different nasals. The results of data analyses are given in Table 27.

‧

Table 27. Distributions of [s]% by the preceding nasals and social/geographical factors (PN= Preceding nasals; numbers in parentheses are frequencies)

Social/geographical factors

Geographical Area Age Gender

Education Level

Taoyuan

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Table 27, when the nasal preceding /ɕ/ is [n], age, geographical area, and gender, individually, are influential in inducing the nonstandard variant [s];

while education level shows no significant effect. To be specific, it is those speakers of older age, those speakers of Chungli City, and the female speakers that show extremely higher [s]% (88.5%, 81.2%, and 63.5%, respectively). Whereas, the two education groups yield almost identical [s]% (51.4% by higher education level, 48.6%

by lower education level). Also, among the three influential social factors, age overrides geographical area, which in turn overrides gender. Moreover, exactly the same patterns are repeated when the preceding nasal is [ŋ].

When any two of the four social/geographical factors interact with each other, the following patterns are located. (1) In Chungli City, older speakers, female speakers, and speakers of higher education level offer significantly higher [s]%

(73.7%, 51.5%, and 48.0%, respectively). (2) When age interacts with gender and education level separately, it is always the older speakers who show higher [s]%, with older females giving higher [s]% than the older males (54.8% vs. 33.7%); however the older speakers of higher education level yield only slightly higher [s]% than the older speakers of lower education level (48.0% vs. 40.5%). It seems that among the older speakers, gender is more potent than education level. (3) When gender and education level are taken into considerations together, it is noticed that despite the educational difference, it is always the females, not the males, who present higher [s]% (30.8% from females of higher education level and 32.7% from female of lower education level; for the two male groups, both close to 20.0%). Again, gender is verified to be more powerful than education level.

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

4.4.6 Distributions of [s]% by the preceding nasals, formality, and the interactions of social/geographical factors

The following two tables present [s]% under the interaction of the external factors. It was hypothesized that the interaction of situational formality and social factors would affect [s]% by different preceding nasals. The result of data analyses are given in Table 28 and Table 29.

‧

Table 28. Distributions of [s]% in Conversation by the preceding nasals and the interactions of social/geographical factors (PN= Preceding nasals; numbers in parentheses are frequencies)

Social/geographical factors

Geographical Area Age Gender

Education Level

Taoyuan Chungli Old Young Male Female High Low

‧

Table 29. Distributions of [s]% in Reading Passages by preceding nasals and the interactions of social/geographical factors (PN= Preceding nasals; numbers in parentheses are frequencies)

Social/geographical factors

Geographical Area Age Gender

Education Level

Taoyuan Chungli Old Young Male Female High Low

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Comparing the data presented in Table 28 and Table 29, it was found that That situational formality is not influential to those subgroups’ use of [s] behind either preceding nasal.

4.5 [s]% in Lexical Diffusion

4.5.1 [s]% in lexical diffusion by character frequency and formality

It was predicted that characters of higher frequency are more likely to join the target variation. Tables 30 shows the results of data analyses.

Table 30. [s]% in lexical diffusion by character frequency and formality (numbers in parentheses are frequencies)

Formality Character

Frequency

General Conversation Reading High

frequency

65.2%

(902)

40.1%

(555)

25.1%

(347) Low

frequency

34.8%

(481)

13.7%

(189)

21.1%

(245)

According to Table 30, in general, [s]% scores much higher in lexical items of high frequency (65.2%) than in those of low frequency (34.8%). Further, when stylistic formality is considered, it is noticed that the general pattern of the emergence of [s] is repeated in conversation, with [s]% being higher in lexical items of higher frequency (40.1%) than in those of low frequency (13.7%). However, in reading, [s]%

in high frequency lexical items is only slightly higher than that in low frequency ones.

This finding support Schuchardt (1928)’s study.

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

4.5.2 [s]% in lexical diffusion by syllable structures and formality

It was expected that not all syllable structures would experience the target phonetic variation in the same speed. Table 31 presents related data.

Table 31. [s]% in lexical diffusion by syllable structures and formality (numbers in parentheses are frequencies)

Formality Syllable structure

General Conversation Reading

/ɕi/ 10.7% (148) 5.9% (82) 4.8% (66) /ɕiɛ/ 2.0% (28) 1.4% (20) 0.4% (6) /ɕia/ 7.8% (108) 4.1% (57) 3.7% (51) /ɕiau/ 12.9% (178) 9.0% (125) 3.8% (53) /ɕiou/ 11.9% (165) 3.5% (49) 8.4% (116) /ɕin/ 13.5% (187) 5.9% (82) 7.6% (105) /ɕi ŋ/ 16.6% (230) 7.5% (104) 9.1% (126) /ɕiɛn/ 7.7% (107) 5.6% (77) 2.2% (21) /ɕia ŋ/ 16.8% (232) 10.7% (148) 6.1% (84)

Total 100% (1383) 100% (1383)

According to Table 31, [s]%, in general, scores higher in the syllable structures of /ɕiaŋ/, /ɕiŋ/, /ɕin/, /ɕiau/, /ɕiou/, and /ɕi/; lower in /ɕia/ and /ɕiɛn/; and even lower in /ɕiɛ/. Although no systematic pattern is found among these three groups of syllable structures, it is confirmed that the emergence of [s], a nonstandard form of /ɕ/, is diffusing across syllable structures.

When situational formality is taken into consideration, it is noticed, again, that the [s]%, in CON as well as in reading, is larger in some syllable structures, less in some others, and fairly small in the rest. In sum, with or without stylistic control, it is always /ɕin/, /ɕiɛn/, and /ɕiŋ/ that yield high [s]%; while /ɕiɛ/ always score the lowest.

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

4.5.3 [s]% in lexical diffusion by characters

It was expected that characters with different syllable structures and frequencies would have different [s]% . Table 32 presents related data.

Table 32. [s]% in lexical diffusion by syllable structures, character frequency, characters, and formality (for style, CON=Conversation; numbers in parentheses are frequencies)

Syllable Structure

Character Frequency

Characters [s]% in general

Formality CON Reading

passages

Reading characters /ɕi/ HF 西 (386)

25.1% (81) 27.8% (54) 21.8% (21) 18.7% (6)

戲 (274)

15.2% (22) 18.7% (3) 14.5% (14) 15.6% (5)

LF 息 (35)

33.3% (24) 25.0% (16) 4.1% (4) 12.5% (4)

惜 (20)

12.4% (21) 21.9% (9) 11.4% (11) 3.1% (1)

/ɕiɛ/ HF 寫(1373)

1.6% (3) 5.7% (3) 0% (0) 0% (0)

些(2536)

6.5% (13) 6.5% (11) 1.0% (1) 1.0% (1)

LF 謝 (73)

5.8% (10) 11.9% (5) 5.2% (5) 0% (0)

洩 (15)

1.2% (2) 3.7% (1) 0% (0) 3.1% (1)

/ɕia/ HF 下(5000)

18.1% (46) 22.4% (28) 14.5% (14) 12.5% (4)

夏 (151)

13.5% (22) 38.2% (13) 6.2% (6) 9.3% (3)

LF 蝦 (37)

12.7% (22) 17.7% (8) 13.5% (13) 3.1% (1)

瞎 (13)

10.7% (18) 20.0% (8) 10.4% (10) 0% (0)

/ɕiau/ HF 笑 (774)

18.6% (32) 36.3%(16) 12.5% (12) 12.5% (4)

小(5000)

26.2% (87) 33.4% (68) 12.5% (12) 21.8% (7)

LF 效 (35)

11.3% (23) 21.3% (16) 6.2% (6) 3.1% (1)

孝 (24)

14.1% (36) 19.8% (25) 8.3% (8) 9.3% (3)

/ɕiou/ HF 修 (319)

34.1% (57) 38.4% (15) 36.4% (35) 21.8% (7)

– – – – –

LF 休 (29)

26.6% (53) 33.8% (24) 21.8% (21) 25.0% (8)

羞 (17)

34.1% (55) 30.3% (10) 40.6% (39) 18.7% (6)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Table 32 (continued)

According to Table 32, in most characters, [s]% increases when the formality decreases. Also, the [s]% of many of the characters including (西,息,小,修,羞,心,信, 星,行,性,向,像, and 享.) are over 20%. Most of these characters are of high frequency and are in close syllable. These findings indicate that character frequency and syllable structure do affect the use of [s].

4.6 Subjects’ Stereotypical Impressions toward Hakka Dialect

In this section, subjects’ attitude toward Hakka dialect were collected through the use of 10 pairs of attitudinal attributes on 7-point semantic differential scales.

4.6.1 Subjects’ stereotypical impressions toward Hakka dialect

Since Hakka dialect is the mother tongue of the subjects. It was expected that the subjects would hold favorable impression toward their mother tongue. Table 33 /ɕin/ HF 心(2598)

26.9% (71) 30.3% (41) 25.0% (24) 18.7% (6)

信 (563)

22.3% (43) 28.1% (18) 17.7% (17) 25.0% (8)

LF 辛 (5)

19.3% (37) 22.2% (14) 20.8% (20) 9.3% (3)

星 (123)

20.8% (36) 20.0% (9) 23.9% (23) 12.5% (4)

/ɕi ŋ/ HF 行 (945)

24.0% (49) 35.5% (27) 16.6% (16) 18.7% (6)

性 (217)

53.7% (129) 62.5% (70) 54.1% (52) 21.8% (7)

LF 幸 (49)

18.0% (30) 13.1% (5) 21.8% (21) 12.5% (4)

形 (83)

14.8% (22) 10.0% (2) 14.5% (14) 18.7% (6)

/ɕiɛn/ HF 先(2500)

8.3% (14) 12.8%(5) 4.1% (4) 15.6% (5)

現 (277)

19.5% (66) 27.7% (58) 5.2% (5) 9.3% (3)

LF 險 (54)

1.9% (3) 3.7% (1) 1.0% (1) 3.1% (1)

獻 (21)

10.6% (24) 13.4% (13) 11.4% (11) 0% (0)

/ɕia ŋ/ HF 向 (3978)

25.1% (57) 34.3% (34) 17.7% (17) 18.7% (6)

像(4412)

46.2% (110) 38.5% (91) 15.6% (15) 12.5% (4)

LF 享 (55)

26.9% (49) 37.0% (20) 28.1% (27) 6.2% (2)

詳 (27)

11.2% (16) 21.4% (3) 11.4% (11) 6.2% (2)

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

presents results of data analyses of the stereotypical impressions toward Hakka dialect offered by subjects a whole.

Table 33. Stereotypical mpressions toward Hakka dialect by subjects as a whole

According to Table 33, it is noticed that the subjects’ general stereotypical impression toward Hakka people is fairly positive (with a mean of 5.41). Among the attributes, the subjects score particular high on and the Hakka people’s being warm, likeable, and agreeable.

4.6.2 Subjects’ stereotypical impressions toward Hakka dialect by social/ geographical factors

It was predicted that the subjects are related with their social/geographical backgrounds, attitudes toward the Hakka dialect, with Chungli subjects, subjects of older age, male subjects, and subjects from higher education levels holding more positive attititudes than the other four correspondent groups. Table 34 presents the

Attitudinal attributes

Mean

1. warm / cold 6.00

2. smart / stupid 5.27 3. noble / vulgar 5.00

4. rich / poor 4.96

5. likable / disgust 5.89 6. confident / inferior 5.65 7. agreeable / negative 5.96 8. strict / relax 5.37 9. generous / stingy 4.65 10. important / unconcerned 5.44

MEAN 5.41

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

related data.

Table 34. Stereotypical impressions toward Hakka dialect by social/geographical factors (for education level, HE= higher education , LE= lower education)

In further examinations, first, it is found that Chungli group and the older group consistently show more positive attitude toward Hakka people , across the 10

traits.Next, in seven out of the ten traits, it is females, not males, that show more positive attitude toward Hakka people, with cross gender gaps much smaller than those in cross-geographical and cross-age comparisons. As for attitudes offered by the two education groups, subjects from higher education level evaluate Hakka dialect higher in six out of the ten traits, with the gap between two education groups being smaller than those by age and geographical area.

According to Table 34, it is verified that Chungli group, the older group, and the higher education group do hold more positive attitude toward Hakka people than

Social/Geographical factors

Attitudinal attributes

Geographical Area Age Gender Education Level Mean

Taoyuan Chungli Old Young Male Female

^HE

LE

1.

親切的/冷漠的 5.78 6.20 6.78 5.26 6.15 6.26 6.18 5.76 6.00

2.聰明的/愚蠢的 4.92 5.60 6.28 4.33 5.23 5.66 5.12 5.30 5.27

3.高尚的/低俗的 4.71 5.20 6.00 4.06 5.00 5.33 4.87 5.15 5.00

4.富裕的/貧窮的 4.35 5.20 5.35 4.26 4.92 5.00 4.87 4.69 4.96

5.喜歡的/厭惡的 5.57 6.20 6.57 5.26 6.23 6.00 6.00 5.76 5.89

6.自信的/自卑的 5.42 5.86 6.64 4.73 6.00 5.73 5.75 5.53 5.65

7.認同的/否定的 5.57 6.33 6.71 5.26 6.30 6.06 6.12 5.76 5.96

8.嚴謹的/隨便的 5.35 5.40 6.07 4.73 5.07 6.00 5.25 5.53 5.37

9.慷慨的/計較的 4.28 5.00 5.42 3.93 4.76 4.86 4.87 4.38 4.65

10.重要的/不在乎的 5.21 5.66 5.92 5.00 5.38 5.86 5.43 5.46 5.44

Mean 5.11 5.66 6.17 4.68 5.50 5.67 5.44 5.33 5.41

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Taoyuan group (5.66 vs. 5.11), the younger group (6.17 vs. 4.68), and the lower education group (5.44 vs. 5.33), with the gap between the two age groups being the largest, followed by that between the two geographical groups, followed by that between the two education groups. As for the two gender groups, out of the researcher’s expectation, it is the female group overrides the male group (5.67 vs.

5.50).

4.6.3 [s]% by stereotypical impressions toward Hakka dialect, formality, and social/geographical factors

The subjects’ stereotypical impressions toward Hakka dialect was anticipated to be reflected their use of [s] for Mandarin /ɕ/. Table 35 depicts the distributions of [s]%, in general as well as in different styles.

Table 35. [s]% by stereotypical impressions toward Hakka people, formality, and social/ geographical factors

Social/Geographical Factors

Stereotypical Impression

[s]%

In general Conversation Reading

All subjects as a whole 5.41 19.65% 26.15% 12.60%

Geographical Area

Taoyuan 5.11 7.08% 14.45% 3.41%

Chungli 5.66 26.48% 37.08% 21.19%

Age Old 6.17 29.56% 41.87% 23.40%

Young 4.68 5.51% 11.49% 2.52%

Gender Male 5.50 13.98% 23.10% 9.43%

Female 5.67 19.66% 28.63% 15.18%

Education Level

Higher Education

5.44 16.25% 24.56% 12.1%

Lower Education

5.33 18.34% 28.12% 13.46%

‧ 國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

According to Table 35, a general pattern of the use of [s] is found: when

geographical area, age, and gender are taken into considerations, the subjects’ use of [s] increases with their stereotypical impression toward Hakka dialect. To be specific, Chungli group, the older group, and the female group, which give higher evaluations toward Hakka dialect, actually score higher in [s]%, and the gap of [s]% between each pair of the correspondent groups seems to coincide with the gap of the two

correspondent groups’ stereotypical impression toward Hakka dialect. In other words, the larger the stereotypical impression gap is, the larger the discrepancy of [s]%.

As for education level, unlike the other three non-linguistic factors it is found that the higher education group, although holding better education of Hakka people,

在文檔中客語母語者使用國音/ɕ/的狀況:社會語言學分析 - 政大學術集成 (頁 59-0)

General distributions of [s]% by the advancement

Chapter 4 Data Analyses

4.3 Distributions of [s]% by the Advancement of the Preceding Vowels

4.3.1 General distributions of [s]% by the advancement

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.3 Distributions of [s]% by the Advancement of the Preceding Vowels

4.3.1 General distributions of [s]% by the advancement of the preceding vowels

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.3.2 Distributions of [s]% by the advancement of the preceding vowels and formality

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.3.3 Distributions of [s]% by the advancement of the preceding vowels and social/geographical factors

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.3.4 Distributions of [s]% by the advancement of the preceding vowels, formality, and social/geographical factors

‧

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.3.5 Distributions of [s]% by advancement of the preceding vowels and the interactions of the social/geographical factors

‧

Table 20. Distributions of [s]% by the advancement of the preceding vowels and the interactions of social/geographical factors (APV= Advancement of Preceding Vowels; numbers in parentheses are frequencies)

APV

Social/Geographical factors

Geographical Area Age Gender Education Level

Taoyuan Chungli Old Young Male Female High Low

‧

Social/geographical factors

‧

Social/geographical factors

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

‧

APV=

Advancement of Preceding Vowels;

Social/geographical

‧

Social/geographical factors

‧

Social/geographical

‧

Social/geographical

‧

Social/geographical facotors

‧

Social/geographical factors

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.4 Distributions of [s]% by the Preceding Nasals

4.4.1 General distributions of [s]% by the preceding nasals

4.4.2 Distributions of [s]% by preceding nasals and formality

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.4.3 Distributions of [s]% by the preceding nasals and social/geographical factors

‧ 國

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學

立政治大學