Low Proficiency Group’s Inappropriate Use of Cohesive Devices

most in the four subcategories of lexical cohesive devices in the two genres, they made most errors in repetition as well. It was found that low proficiency writers’

repetition caused redundancy and failed to use pronouns appropriately. The failure to use pronouns may be an indication of writers’ low ability level at the stage of the writing development. Therefore, the number of repetition in the low proficiency

104

group could not serve as an index of a good writing. Based on the above-mentioned observation, the conclusion can be drawn that there was no great variation in terms of error types in the low proficiency writers’ narrative and expository writing.

In addition to the qualitative analysis of error types in the two genres, the total number of inappropriate use of cohesive devices in the two genres was compared.

The overall error rates in the two genres are 17% in narration and 26% in exposition.

The results show that low proficiency writers made much more errors in exposition than in narration. To further explore whether errors of grammatical cohesive devices caused the variation in the two genres, the error rates of the low proficiency group in grammatical cohesive devices were are 13% in narration and 16% in exposition respectively. That is, 13% grammatical cohesive devices contributed to the inappropriate use of grammatical cohesive devices in narration while 16%

grammatical cohesive devices contributed to the inappropriate use of grammatical cohesive devices in exposition. The findings seem to indicate that low proficiency groups made more inappropriate use of grammatical cohesive devices in exposition.

As for the subcategories of the grammatical cohesive devices, as can be seen from Table 87, more errors were found in the category of pronominal reference in exposition. The calculation of the error rate of the use of pronominal reference validated that in the low proficiency group’s narrative writing, 6% pronominal reference contributed to inappropriate use of pronominal cohesive devices while 28%

pronominal reference contributed to inappropriate use of pronominal cohesive devices in expository writing. Qualitatively, by the analysis of the errors in pronominal reference, the low proficiency group’s errors in exposition were unclear referents. That is, they would use the singular pronoun to refer to the plural referent.

Consequently, more inappropriate pronominal reference was found in exposition than in narration. The result seemed to show that low proficiency level writers were less

105 lexical cohesive devices, we will find that in both genres, the percentage of the error rates of the lexical cohesive devices were both 22.8%. From the findings, we can infer that low proficiency level’s inappropriate use of lexical cohesive devices did not vary greatly in the two genres.

Table 87

Low Proficiency Group’s Inappropriate Use of Grammatical Cohesive Devices

Low Reference Substitution

Pronominal Definite Comparative Nominal Verbal Clausal

N 14 12 4 0 0 0

E 39 15 7 1 1 0

Note. L= low proficiency level writers; N=narration; E=exposition.

Table 88

Low Proficiency Group’s Inappropriate Use of Grammatical Cohesive Devices

Low Ellipsis Conjunction

Nominal Verbal Clausal Additive Adversative Causal Temporal

N 0 0 0 4 3 6 3

E 0 0 0 3 7 4 3

Note. N=narration; E=exposition

Table 89

Low Proficiency Level Writers’ Inappropriate Use of Lexical Cohesive Devices

Essay Lexical Total

Repetition Synonym Antonym Superordinate

N 48 4 0 5 103

E 42 2 7 1 131

Note. N=narration; E=exposition

106

CHAPTER FIVE CONCLUSIONS

In this chapter, the research findings of the current study are summarized first.

Then, pedagogical implications will be presented for ESL instructors in order to benefit students’ writing. Furthermore, the limitations of the study will be pointed out and related suggestions for future research will also be provided.

5.1. Summary of the Quantitative Analysis

The purpose of the present study is to examine Taiwanese college students’

employment of cohesive devices in writing in terms of distinct proficiency levels and different genres quantitatively and qualitatively. To be more specific, we attempt to examine whether proficiency level will influence the employment of cohesive devices in narration and exposition respectively and whether genres will affect high or low proficiency level writers’ employment of cohesive devices. In addition to the quantitative analysis of the frequency counts, the qualitative analysis is indispensable to identify the inappropriate use of cohesive devices from the perspective of proficiency difference and genre difference. A total of 60 college students, recruited from senior English majors and freshman non-English majors with equal numbers from National Taiwan Normal University, took part in the study. The participants were requested to compose narrative and expository writing. After the scoring section, the collected essays were further divided into high-rated and low-rated groups, representing high and low proficiency level writers respectively. Later, the total number of cohesive devices was calculated in high and low level writers’ writing and was analyzed with the modified version of Halliday and Hasan’s (1976) taxonomy of cohesion in English and with t-test. Additionally, the inappropriate use of cohesive devices was identified in high and low proficiency level writers’ essays in order to conduct the qualitative analysis of the participant’s errors. The major findings of the

107

study are summarized in the following.

There were significant differences in high and low proficiency level writers’

overall number of cohesive devices both in narration and in exposition, which revealed that proficiency level was the predominant factor that influenced the employment of cohesive devices. Assuming from this, the employment of the numerous cohesive devices seemed to serve as an index for the better writing quality (Ferris, 1994; Liu & Braine, 2005; McCulley, 1985; Normant, 1982; Witte & Faigley, 1981). In other words, high proficiency level writers have much competence in the perception of the grammatical sentence structures, discourse cohesion, organization of ideas, and semantic meaning of words than low proficiency level writers do.

In the comparison between high and low proficiency groups’ grammatical cohesive devices, there were significant differences between the two groups in narration and in exposition respectively. In addition, significant differences were found in high and low proficiency level writers’ use of reference and conjunction in the narration, especially in the category of pronominal reference, definite article, additive and temporal conjunction, while no significant differences were found in the use of the subcategories of grammatical cohesive devices in exposition. These findings are not only in agreement with the previous Jin’s (2001) study but probably suggest that high proficiency writers were better at narrating stories by employing more pronominal reference that revolved around the leading character and the development of the story. Additionally, high proficiency writers were good at exerting various conjunctive cohesive elements to achieve the global discourse coherence while low proficiency writers were less capable of applying the discourse transition words to make the text more globally coherent.

With regard to the comparison of lexical cohesive devices, high proficiency writers used significantly more lexical cohesive devices than low proficiency writers

108

in the exposition, but not in the narration. The results appear to suggest that the expository writing may be a more difficult genre for low proficiency writers to fully express their arguments by applying words with different semantic relationship.

Moreover, writers of different proficiency levels used repetition most in their writing, which may indicate that repetition of the key idea is not only indispensable but also easier for writers to employ in the text. More importantly, it is obvious that high proficiency writers distinguished from low proficiency writers by the employment of the various synonymous words whether in narration or exposition, which further supports the idea that proficiency is the dominant factor that determines the difference in the use of cohesive devices (Chen, 2003; Crowhurst, 1987; Jin, 2001).

Concerning the factor of genres, there were no significant differences in both high and low proficiency level writers’ overall number of cohesive devices between the narrative and expository essays. In spite of this, from the observation of the means, both high and low proficiency level students used more cohesive devices in the narration than in the exposition. In conclusion, for writers of different proficiency levels, no genre difference was found in the total number of cohesive devices and more cohesive devices were found in narration than in exposition. These findings are consistent with the results of the previous studies despite the fact that different subjects and genres were used (Jin, 2001; Norment, 1982, 2002).

With regard to the grammatical cohesive devices in the two genres, high proficiency writers significantly used more grammatical cohesive devices in narration while no significant difference was found in low proficiency writers’ writing between the two genres. In addition, both groups used more pronominal reference in narration than in exposition. The result perhaps reveals that the discourse type may influence the writers’ employment of referential cohesive devices because in narrative writing, the elaboration of characters is indispensable and the development of the story mainly

109

revolves around the characters. Therefore, as the story enfolds, it is necessary to use personal pronouns, either from the first or the third person perspective, in the narrative text. On the other hand, there was also a significant difference in the low proficiency level’s use of conjunction, namely, adversative.

In terms of the total number of lexical cohesive devices, for both groups, there was no significant difference as regards genre differences. Also, repetition was the most frequently-used lexical cohesive devices regarding genre differences. On the other hand, writers of different levels were found to use more antonym in exposition than in narration, which perhaps shows that genres affect the use of antonym because in the expository writing, clarifying or explaining the topic from different sides to support a thesis statement is essential and indispensable. By the employment of the antonym, writers can strengthen the arguments by offering a broader perspective for readers.

5.2 Summary of the Qualitative Analysis

In the section, the summary of the qualitative analysis regarding (1) high and low proficiency groups’ inappropriate use of cohesive devices in narration, (2) high and low proficiency groups’ inappropriate use of cohesive devices in exposition, (3) high proficiency groups’ inappropriate use of cohesive devices in the two genres and (4) low proficiency groups’ inappropriate use of cohesive devices in the two genres, will be presented as what follows.

5.2.1 High and Low Proficiency Groups’ Inappropriate Use of Cohesive Devices in Narrative Writing

In the qualitative analysis of different proficiency writers’ inappropriate use of cohesive devices in the narrative writing, low proficiency level writers had made much more errors in cohesive devices than high proficiency level writers did (103 vs.

28). In terms of grammatical cohesive devices, reference, the major problem with

110

both high and low proficiency level writers was the use of definite or demonstrative reference. Both high and low proficiency writers made mistakes because of the addition or lack of definite articles in the text. Different from high proficiency writers, low proficiency writers also had problems with how to use the deictic demonstrative reference to refer to a present or past-time referent. The second common error is the use of pronominal reference, low proficiency writers had a mismatch between the pronoun and the referent. Moreover, they also suffered from poor sentence structures so that they failed to supply a needed pronominal reference in the text. As for comparative reference, unlike high proficiency writers, low proficiency writers were weaker in producing the correct comparative structure and even failed to judge when to use the structure in the text.

In addition to reference, both high and low proficiency groups had errors in the use of conjunction. First, both groups had fragments and failed to use conjunction in their texts. Moreover, low proficiency writers tended to produce simple sentences and also failed to recognize the inter-sentential relationship so that they failed to supply a needed conjunctive tie in the discourse. What’s more, even though low proficiency group writers used conjunction in their text, they sometimes suffered from poor sentence structures and produced fragments. Further, they had no idea of the correct expression of each conjunction and failed to choose the correct form in the text.

In the aspect of lexical cohesive devices, high proficiency writers made few errors while low proficiency writers made a tremendous number of errors, especially in the category of repetition. Simply stated, although both high and low proficiency groups utilized repetition most, the quantitative number of repetition in the low proficiency writers’ samples could not serve as an index for the quality of a good writing because low proficiency writers’ repetition led to the redundancy and non-idiomatic expression in the text.

111

Observing from the overall errors made by both groups, the major problem was the word choice. Although writers employed the lexical cohesive devices to form different kinds of ties, they failed to choose the proper word in the context. As a result, the wrong choice of word may cause readers’ misinterpretation of the text. In addition, despite the fact that both groups employed various numbers of lexical cohesive devices in the text, their minor mistakes resulting from basic grammars (e.g., non-native English, misspellings, tense, plural inflections and sentence structures) made their cohesive devices imperfect and inappropriate. In comparison with high and low proficiency writers’ performance, and from the observation of the total number of the inappropriate use of cohesive devices, low proficiency writers had more local or global mistakes in terms of grammatical and lexical cohesive devices than high proficiency writers. Further, a large proportion of low proficiency groups’ errors were involved with the basic grammatical sentence structures, lexical spellings or plural inflections, which seemed to indicate that the low proficiency writers’ poor command of English may be detrimental to the cohesion in the text.

5.2.2 High and Low Proficiency Groups’ Inappropriate Use of Cohesive Devices in Expository Writing

Regarding high and low proficiency level writers’ inappropriate use of cohesive devices in the expository writing, low proficiency writers still made more errors than high proficiency writers did. In terms of the grammatical cohesive devices, the major problems in both groups’ samples were unclear referents of the pronominal reference.

Moreover, low proficiency level writers tended to employ the pronominal reference throughout the text and failed to use the noun phrase to signal a new topic in the discourse. Also, they suffered from poor sentence structures even though they used the correct pronominal reference. More importantly, low proficiency writers produced shift of pronouns in their samples.

112

As for definite or demonstrative articles, low proficiency writers had more errors involving lack or addition of the definite articles than high proficiency ones. Next, for the use of comparative reference, compared with high proficiency writers, low proficiency ones not only had more errors but they also had suffered from writing the correct comparative structures. In respect of substitution and ellipsis, high proficiency writers had used more substitution and ellipsis than low proficiency ones, but their employment of substitution and ellipsis suffered from immature sentence structures.

Unlike high proficiency writers, members of low proficiency levels did not use ellipsis in the writing, but only one instance of substitution, which proved that ellipsis not only creates difficulty in learning what structural omissions are permissible, but also does not seem to be readily used even by proficient learners (Scarcella & Brunak, 1981).

As for conjunction, high proficiency writers only had two instances of conjunctive errors, lack of proper conjunction and misuse of the word, even, as a conjunction. In contrast with high proficiency writers, low proficiency writers suffered from run-on sentences. There were also fragments involved with the use of because and although. Furthermore, low proficiency writers had more global errors than high proficiency writers. They not only failed to supply a proper conjunctive element in the discourse but also failed to recognize the relationship of different contexts (e.g., causal relationship or additive relationship). Besides, because of their poor English, they couldn’t supply a correct form of the conjunction.

In the aspect of lexical cohesive devices in exposition, low proficiency writers still had more instances of mistakes in plural inflection of the word, misspellings and wrong part of speech than high proficiency writers did, which suggests that most of the low proficiency writers’ problems were local in nature. Also, in the category of lexical cohesive devices, both high and low proficiency writers had the most errors in

113

repetition, and they had similar errors in the word choice. Besides, the greater number of repetition in the low proficiency groups’ writing may cause the redundancy in the text and the repeated noun phrases also revealed the low proficiency writes’ failure to use pronouns in the text, which may indicate the writers’ low ability level because the use of pronouns is an indication of the higher writing level at the stage of the developing writing ability. To sum up, the low proficiency group greatly suffered from the basic vocabulary use such as spellings, countable nouns, part of speech, not to mention the higher level skills in word choices.

5.2.3 High Proficiency Group’s Inappropriate Use of Cohesive Devices in Two Genres

The comparison of the error types in high proficiency writers’ narrative and expository writing revealed that there was no difference in the high proficiency group’s errors in the two genres. With regard to grammatical cohesive devices, the high proficiency group had errors in lack/addition of definite articles, pronominal reference and the use of conjunction. As to lexical cohesive devices, the high

proficiency groups mainly had problems in the word choice and other minor ones in misspellings, plural inflection, part of speech and sentence structures in the two genres.

On the other hand, through the comparison of the error rates in the two genres, we can find that high proficiency level writers had slightly more number of inappropriate use of cohesive devices in exposition. Also, in the category of pronominal reference, high proficiency level writers were found to have more instances of inappropriate use of cohesive devices in the expository writing. Secondly, few instances of substitution and ellipsis were used in high proficiency level’s writing and there were only two instances of substitution errors regarding wrong sentence structures and lack of substitution in the expository writing. Thirdly, as for

114

conjunctions in the two genres, the common problems were the lack of conjunctions and the fragments in the participants’ writing.

Next, for the lexical cohesive devices, the overall error rate involving lexical cohesive devices in the two genres did not vary greatly as well. To sum up, high proficiency level writers had more instances of errors in the expository writing and the major problems of their lexical cohesive devices were the word choice or usage in the discourse.

5.2.4 Low Proficiency Group’s Inappropriate Use of Cohesive Devices in Two Genres

The comparison of the error types in low proficiency writers’ narrative and

The comparison of the error types in low proficiency writers’ narrative and