[contents][chapter 1][chapter 2][chapter 3][chapter 4][chapter 5][chapter 6][references]


                                                                6.1  Summary of the findings

                                                                6.2  Conclusion  

                                                                6.3  Recommendations for further studies  

6.1  Summary of the findings

The statistical results of this research reveal that the readers' (244 students from Tamsui and Tamkang Universities) overall attitudes toward written Taiwanese are positive. That is, with mean scores 5.15 ((C + D + E + F + G) / 5) or 4.50 ((A + B + C + D + E + F + G) / 7) based on a seven-point semantic differential scale. Also, the results reveal that the readers evaluated the prepared 7 reading samples significantly different (except E vs. F). The ranking of the reading scores is: (lowest) B < A < D < EF < C < G (highest). This ranking reflects the preferences of the Mandarin and Han-characters educated college students with regard to the orthographies of written Taiwanese. The survey reveals that Roman script and Bopomo (ㄅㄆㄇ; National Phonetic Symbols for Mandarin in Taiwan) used in Taibun texts received more negative evaluations (lower scores) by the 244 readers; in other words, Roman script and Bopomo can reduce the readers' degree of acceptance; Han characters received the most positive evaluation. The survey indicates that readers will give higher ratings to those orthographies which are more "readable" to them. In other words, the ratings of seven reading samples are a reflection of readability to the 244 subjects. The readability could be affected by readers' language and orthography abilities. For detailed results and discussion, see section 5.1.

In addition to the orthography factor, the background of the readers could also affect their evaluations (see section 5.2). Based on the results of the investigations, there are 6 factors, which can affect readers' evaluations. There are: place of residence (Taipei vs. non-Taipei), major (Taiwanese and English vs. Mechanical Engineering vs. Chinese, Japanese, and Public Administration), mother tongue (Taiwanese vs. non-Taiwanese), language ability (Taiwanese vs. non-Taiwanese or we could say non-Mandarin-only vs. Mandarin-only), the individual's evaluation of her/his national identity (Taiwanese vs. non-Taiwanese), and assertions on Taiwan's preferred national status (independence vs. non-independence). There are 3 factors, which do not affect readers' evaluations. They are: gender, age, and political leanings. "Taiwanese" mentioned in this paragraph includes Hakfa and/or Holo languages.

Further, based on the findings above, a prediction equation, the Taibun equation, which can predict the reading scores of particular reading samples evaluated by different readers, was formulated as follows:

         Y' = 4.78 - 1.41 (A) - 2.92 (B) + 0.25 (C) - 0.61 (D) + 0.00 (EF) + 1.00 (G)

                -0.13 (Taipei) + 0.18 (TB-EN) - 0.23 (ME) + 0.23 (NTL) - 0.34 (M-only)

                + 0.15 (T-id) + 0.15 (TI)


A, B, C, D, EF, G refer to the reading sample

Taipei: Taipei as the place of residence

TB-TN: major in Taiwanese or English

ME: major in Mechanical Engineering

NTL: native Taiwanese languages (i.e., Taiwanese or Hakfa) as mother tongues

M-only: monolingual in Mandarin

T-id: Taiwanese identity (i.e., Taiwanese-only, or Taiwanese-Chinese)

TI: assertion of Taiwan independence


 All the independent variables must encoded either 1 (yes) or 0 (no) when applied to this Taibun equation. The value of Y' is based on a seven-point semantic differential scale, from lowest 1 to highest 7.

         Readers' reactions to the writers of written Taiwanese were also examined (see section 5.1.3). Generally speaking, readers didn't have any idea regarding the writers. However, if readers associated writers with particular expectations, Taibun writers were mostly regarded as male, with native political leanings, native religions, and native identity. In the dimension of age, readers didn't associate Taibun writers highly with any particular age group. In the gender dimension, 62% of the readers didn't assign a particular gender to the Taibun writers as a group, 31% assigned "male" to them, and 7% "female." In the dimension of political leanings, 63% of readers didn't associate the writers with particular parties, 24% associated them with DPP, 8% with KMT, 3% with TAIP, 2% with CNP, and 1% with GPT. In the religion dimension, 71% didn't think there was a connection between Taibun writers and religion. The others assigned Buddhism (10%),Taoism (11%), Christianity (5%), or Catholicity (2%) to the Taibun writers. In the dimension of national identity, 63% didn't associate the Taibun writers with any expected national identity. Of the rest of them, 20% associated the writers with independence, 13% with maintaining current status, and 4% with unification. Finally, the readers' understanding of the languages the Taibun writers are expressing was also tested. It reveals that language ability plays an important role on readers' understanding of Taibun writings. For instance, compared to other speakers, Hakka speakers were more likely to tell reading D (written in Hakfa) from other readings (in Holo Taiwanese).

In section 5.3.1, the results of chi-square tests reveal the relationships among three characteristics: ethnic identity, mother tongue, and language ability. The results of tests between ethnic identity and mother tongue show that: (1) the relationship between Holo people and Holooe speaker is interdependent. (2) Even though a person identifies herself/himself as ethnic Hakka, s/he may not regard Hakfa as her/his mother tongue. (3) Mandarin speakers may not identify themselves as Mainlanders. The results of tests between ethnic identity and language ability show that: (1) Holo/Hakka people are interdependent with M-Holooe/Hakfa-plus speakers. (2) Mainlanders and M-only speakers are not interdependent with each other. The results of tests between mother tongue and language ability show that: (1) people with Holooe/Hakfa as a mother tongue are more likely to possess M-Holooe/Hakfa-plus ability. (2) People with Mandarin as a mother tongue do not necessarily possess Mandarin-only ability.

In section 5.3.2, the results of factor analysis on the 40 questions (items 15-54 in the questionnaire) indicate that 11 factors were extracted. They are the dimensions: (1) national identity, (2) language attitudes, (3) orthography, (4) Taiwanization (本土化), (5) political majority, (6) Guoyu (國語), (7) linguistic heritage, (8) multi-lingualism, (9) vernacular education, (10) Taiwanese identity, and (11) language of education. 

6.2  Conclusion          [back to top]

There are three fundamental writing schemes of Taibun in the contemporary issues of written Taiwanese. They are Han character-only, Han-Roman mixed, and Roman script-only. The results of the investigation reveal that the college students surveyed have positive attitudes toward overall Taibun (regardless of different orthography). As for which orthography is preferred, the results reveal that the college students tend to prefer Han-only more than Han-Roman and Roman-only. The results reflect the preferences of the Mandarin and Han character-educated college students with regard to the written Taiwanese. Since all students in Taiwan have been taught the Mandarin and Han characters through the national education system since 1945, it implies the potential difficulty of promoting Roman script in the Han character dominated society.

Usually, many factors are involved in the choice and shift of orthography. From the perspective of social demand, the factor that the increasing[1] use of spoken and written Mandarin by Taiwanese people has reduced the demand for a new orthography. In other words, people may not feel the necessity of learning a new orthographic tool since they have already acquired writing skill in modern standard Chinese. Even so, the readers' positive attitudes toward Taibun indicate that it is still possible for Taibun to be accepted in addition to the existed Mandarin writing. Thus, what findings of the survey may contribute to the promotion of Taibun? According to the results of the survey, there are seven factors that could affect readers' evaluations on Taibun. They are orthographic design, place of residence, major, mother tongue, language ability, national identity, and national status. Since place of residence and school major are not controllable factors (because there always have been people living in different places and with different majors), a Taibun promoter may pay attention to the other factors, which can be divided into three domains: 

(1) Orthographic domain, which refer to the designs of orthography. Usually, good orthographic designs do not absolutely guarantee that they will be accepted by public. On the other hand, the acceptance of orthographies by people does not necessarily mean that the orthographies were well designed. In this survey, although Roman script was rated lower than Han character, the economy and easy learning of Roman script make Romanization still worth consideration. The fact that most of the current Taibun publications are published in the Han-Roman mixed scheme instead of Han-only, points out that readers may tend to prefer Roman script after they are skilled in Taiwanese Romanization. In other words, if the current Bopomo, which is taught through the national education system in Taiwan, can be replaced by Romanization, the circumstance of using Romanization will increase the possibility of promoting Romanized Taibun. The Roman script might be in competition with Han character, or even replace Han character if Romaniztion is taught with Han character at the same time when students enter elementary school.     

(2) Language domain, which includes the factors of mother tongue and language ability. The survey reveals that people who are able to speak native Taiwanese languages are more likely to give higher ratings to Taibun. This fact points out that the promotion of Taibun should focus on the particular groups who frequently use or are able to use Holooe or Hakfa. Moreover, Taibun should be promoted to Taiwanese public as soon as possible, before people entirely shift to monolingual Mandarin Chinese.

(3) Political domain, which covers the factors of national identity and national status. Usually, political transitions can affect the language situation. In the case of Taiwan, the current ambiguous national status and diversity of national identity reflect people's uncertain determinations on the issue of written Taiwanese. On the other hand, people's uncertain determinations on the Taibun issue also reflect the political controversy on national issues of Taiwan. Although Taiwan is still under the rule of the KMT, there are still some chances that Taibun could be adopted as an official written language under some conditions. For instance, if the government is dominated by a native political party (and if the members of the native party have a strong will to promote Taibun), then Taibun could become an official written language in Taiwan.

In short, whether or not Taibun will be accepted and successfully promoted to a national status highly depends on people's orthography demands and their attitudes toward written Taiwanese. Moreover, their language ability and national identity also will play an important role while they are making the determinations.

6.3  Recommendations for further studies          [back to top]

 While this thesis has uncovered many of the factors determining Taiwanese readers' reactions to various orthographies, it should be look upon as preliminary. Subsequent research is necessary to test the conclusions drawn here. There are three primary concerns that need to be addressed:

 Because of the limits of time and cost, the investigations on attitudes toward written Taiwanese of this thesis focus on college students. Some factors such as educational level and social class were excluded from the study. Therefore, those factors may be considered in further studies.

 The original purpose of the study was to predict readers' reactions to Taibun articles occurring in newspapers or magazines. Therefore, ability in Taibun writing is not required for the readers in the test. Other research may be proposed to survey the people who are skilled in Romanization, such as the Church people who are able to use the traditional Romanized Peh-oe-ji, and Taibun writers who publish Taibun works. They might have different preferences from the readers in this study.

 In this study, readings mostly or partly in Roman scripts were more negatively evaluated by the readers. One of the factors might be that their reading efficiencies were reduced because of the fact that most of the readers were not skilled in Taiwanese Romanization. Therefore, it might be interesting to consider how different orthographies affect readers' reading efficiency. For instance, how long does it take for readers to go through paragraphs in Han-only, and in Roman-only? What proportion of Han characters to Roman scripts results in the best reading efficiency?

 [back to top]

 [1] See chapter 3 for references of language use and shift in Taiwan.