國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
CHAPTER TWO LITERATURE REVIEW
Human beings are not born with writing competence (Martlew, 1983). Whether native or foreign language users, people develop their writing competence via learning (Kent, 2002). To know more about writing, this chapter consists of four sections. The first section defines writing, reviews the change and development of writing models and refers to the implementation of writing pedagogy in Taiwan. The second section discusses the grammatical errors in writing works and lists out common grammatical errors and different correction ways. The third section defines the concept of the teacher feedback, analyzes the types of teacher feedback and at last presents the related studies on teacher feedback by different characteristics in the fourth section.
The Concept of Writing
Writing is defined as a complex process involving various skills such as handwriting, spelling words, contextualizing the conventions, planning ideas and segmenting information properly (Martlew, 1983). Instead of incorporating these skills and applying the words, conventions and their ideas into writing, writers have to go through a series of procedures for composing a piece of work.
Hayes and Flower (1980) propose the three main steps to composing written passages to be planning, generating and reviewing (cited in Martlew, 1983). Planning is the process of creating new ideas, arranging related ideas sequentially and setting goals for the writer to achieve while composing text. Generating includes producing words and selecting the proper ones in the authentic language context and organizing sentences. Reviewing refers to reading the text again, editing the paragraphs, revising the errors that were found and to rechecking the cohesion of context. To compose the written works, writers need to utilize the three-staged writing process aptly (Hyland,
13
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
2003). However, the inexperienced writers seem to be incapable of managing the process with ease (Stoke, 2007), because they cannot generate writing ideas clearly, cannot separate the writing information appropriately and cannot revise their statement to meet the cohesion in their written works (Shaughnessy, 1979). As a result, writing pedagogy becomes important to help them improve their writing competence.
Furthermore, writing in a foreign language involves the knowledge of foreign language. Writers need to be familiar with the foreign linguistic knowledge such as knowing about foreign words, syntactic patterns and cohesive devices that link up the key sentences and main ideas in contexts (Hyland, 2003). Beside word and syntax knowledge, writers need to convert their thoughts into the target language with appropriate grammar rules (Anderson, 1985). This means, to write in a foreign language, writers need not only go through the three-staged writing procedure with various writing skills, but also know the foreign language well. These two tasks are challenging for foreign language writers (Hung, 2007), and thus, writing in a foreign language depends on the instruction to know where to and how to improve the writing skills (Purves, 1987). Writing instruction helps writers deal with the foreign linguistic knowledge, be skillful at the writing procedure and compose articles matching their intellectual abilities (Leki, 2005); for these reasons, writing instruction is essential to learners.
A system of instruction teaching students how to write (Zamel, 1987) and improving writing skills efficiently (Leki, 2005) is defined as writing pedagogy.
According to Badger and White (2000), the writing pedagogy consists of two main teaching models: the product and the process models.
14
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Writing Models
In the past 30 years, product and process approaches have been taken as two influential models for writing instruction in the EFL classroom (Badger & White, 2000). The following section defines these two models and then discusses their strengths and weaknesses respectively.
Product Model. In this model, writing is seen as using the linguistic knowledge accurately, including the vocabulary, syntax and cohesive devices (Pincas, 1982b).
Learning to apply the linguistic knowledge appropriately, students develop their writing skills in four steps: familiarization, controlled writing, guided writing and free writing (Pincas, 1982a).
During the first stage, familiarization, students become familiar with particular traits of the writing format, for example, the format is suggested as three parts,
introduction, body and conclusion (Williams, 1998). Learners are taught with the fixed formats at the start. In the second stage, controlled writing, students practice writing with the instructor’s support. The instructor may offer partial information about the writing topic, such as the first paragraph, and the learners continue writing in the same genre. The third stage is guided writing. It means students gain increasing writing freedom with teachers’ guidance. Teachers provide the instruction before writing and learners try to compose articles with the guide. In the last stage, the free writing stage, students express themselves in the writing passages without instructions from teachers.
In the first three stages, students practice writing skills under teachers’ instruction, learn the designed context and try to practice new words. That is, learners construct sentences in set-up patterns and arrange cohesive devices with their instructors’
support. Moving on to next stages, writers become more skillful and gain increasing freedom to write; for instance, writers start to generate their own writing ideas and
15
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
develop their own voice (Pincas, 1982b).
The main concern of this model is the structure of language (Badger & White, 2000). In other words, good writing means correct writing (Williams, 1998). Teachers emphasize the correct usage on the sentence level, focus on the exercise of grammar drilling and expect that students can learn from the error correction provided by teachers (Applebee, 1986). Since teachers focus on training students to write accurately and expect them to follow the writing format; this model is considered teacher-centered and product-oriented (Huang, 2009). Students can learn from
teachers’ samples, duplicate the format and add their thought into the writing work by their competence. In the product model, teachers provide the text form and language input, acting as a presenter demonstrating the accurate format while learners are the followers to learn what teachers provided (Pincas, 1982b). With teachers’ instruction, writing ability is developed and the students’ written works are monitored by teachers and under control as products (Huang, 2009).
The advantage of this model is that students can learn to write via teachers’
prescription (Zamel, 1985). Under this condition, this model mitigates the cognitive burden of learners and helps them work on the smaller and manageable parts, focusing on language structures, receiving error correction, and revising their written work according to teachers’ feedback first (Camhi & Ebsworth, 2008). While inexperienced writers concentrated on one issue at a time, their learning load is eased off. Students concentrate learning on language structures and after they can handle the grammar rules, students work on the content and add new ideas when these issues are cognitively affordable to learners. That is to say, this model supports learners with error correction feedback and helps them first work on the manageable part, the sentence level (Camhi & Ebsworth, 2008).
16
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
However, two disadvantages are proposed (Applebee, 1986; Hyland, 2000;
Sokolik, 2003; William, 1998; Zamel, 1985; 1987) as below. The first disadvantage of the product model is that students learned limited understanding of writing (Zamel, 1985). Because the product model focuses on the correct usage of grammar, students’
writing work turns in mainly for the purpose of correction (William, 1998). Students do not receive the feedback on the content at the beginning of learning writing so they gain limited understanding of writing (Zamel, 1985; Sokolik, 2003). This model is criticized for the overt emphasis on the sentence-level errors instead of improving the content (Applebee, 1986). Secondly, since teachers mainly provide the corrections and grades (Zamel, 1987), students keep receiving the error correction pointing out
weaknesses of their writing work. Gradually, in students’ perceptions, the error correction is connected to their negative feeling (Hyland, 2000). The negative feeling lessens learners’ confidence and interest toward writing (Zamel, 1987) because they always receive the corrections instead of the encouraging comments (Sokolik, 2003).
Process Model. In this model, writing is considered as a process of discovery (Zamel, 1985). Badger and White (2000) define writing as a procedure to apply
linguistic skills and student writing is regarded as the exercise for sharpening linguistic skills in this model. As the writing works are composed, the writing competence is developed unconsciously (William, 1998).
A 3-staged writing procedure is proposed as follow: pre-writing, while-writing and post-writing (Winterowd & Murray, 1985). First, in the pre-writing stage, writers planned ideas for the specific purpose and the audience; then, in the while-writing stage, the planned ideas are written into words and the articles are composed as drafts.
Writers pause, review and revise the content if it does not correspond to the main purpose. At last, in the post-writing stage, writers do the final review and fix the errors
17
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
on the sentence-level.
The process model focuses more about the process of writing than the outcome of writing (Sokolik, 2003). This model emphasizes on the function of communication and fixes the grammatical errors in the last writing stage (Camhi & Ebsworth, 2008).
When writers are able to express their ideas clearly and to communicate with the readers, the writing work is considered successful and acceptable. During the while-writing stage, the revision is made if the content does not match the main purpose (Winterowd & Murray, 1985). In this stage, both the teacher and the peers act as the audience to offer feedback to the writers (Kroll, 2001). While the peers provide their feedback as the audience, they offer their background knowledge, e.g. the writing ideas or their feeling after reading. With various backgrounds and different viewpoints, students bring diverse information into the classroom (Kroll, 2011). Because of the feedbacks, multiple drafts are needed for the global changes, such as the changes on the content or on the organization (William, 1998). In the revising process, students develop their writing skills (Badger & White, 2000) and the student texts are
considered works in progress (Zamel, 1985). In the other words, the content and the student revisions are emphasized in this model (Applebee, 1986).
The merit of this model is providing opportunities for learners to develop the metalinguistic ability while composing. Metalinguistic ability refers to the awareness of the writers have toward the knowledge of language (Gombert, 1992). Camhi and Ebsworth (2008) noted the metalinguistic principles are put into practice via analysis and exploration. In this model, students analyze the peers’ writing work with teachers and offer feedback. At the same time, they receive the feedback from others and improve their writing work. They explore their writing skills in the revising process, and help learning process become more efficient. Thus, in the process model, students
18
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
enhance their metalinguistic abilities while receiving feedback from teachers and peers (Kroll, 2011).
The process model acknowledges the contribution of what learners bring into the classrooms, such as the background knowledge and metalinguistic ability; this model considers the knowledge besides the linguistics as beneficial and realizes the
importance of the revision process, and thus, positive outcomes are showed on writing fluency and comprehensibility (Camhi & Ebsworth, 2008).
On the opposite end, the weakness of this model is regarding all writings as being produced by one uniform set of procedure as pre-writing, while-writing and
post-writing. Zamel (1985) pointed out that the process model did not offer fixed writing forms but expected learners to learn from experience and revisions. Badger and White (2000) even argued that any corrective feedback was pointless and any overt concern to linguistic form was needless (Camhi, 2004). That is, it has been criticized of offering inadequate input to learners and neglecting the value of writing accuracy and rhetorical clarity.
In conclusion, in the product model, students learn with teachers’ prescriptions and error corrections (Zamel, 1985), emphasize the writing outcome (Pincas, 1982b) and care about the accuracy of sentence structures. While in the process model, writers develop writing skills by receiving feedback on content (Kroll, 2001), emphasize the revision process (William, 1998) and fix the grammatical errors on the sentence level in the last writing stage (Winterowd & Murray, 1985). Therefore, these two models differentiated from what they emphasized and the main concerns were affected by the writing procedures.
19
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Writing Pedagogy in Taiwan
In Taiwan, writing pedagogy was implemented more often from junior high schools to colleges than in primary schools (Hung, 2007). Unlike the primary school students, the students from junior high schools to universities know more vocabulary and have authentic purposes for writing, such as preparing for entrance exams, writing reports, and practicing writing skills in the composition classes (Hammill & Larsen, 1988). Thus, teachers implement writing pedagogy more often from junior high to university level than in the primary grades (Hung, 2007).
From junior high schools to college, most teachers admitted the benefit of the process model, but when carrying out the writing classes, teachers put the product model into practice more often in classrooms (Huang, 2009). Most teachers
acknowledged the main features of the process model such as peer feedback and the revision process (Kuo, 2004); they also considered teachers should play different roles in the writing class, such as the roles of the audience and the instructor (Wu, 2006).
However, in the in-class observation, the product model activities, played the main role such as grammatical drilling practice, discussion of the writing format and error correction on the sentence-level (Hsu, 2005; Kuo, 2004). In other words, teachers agreed upon the value of the process model but conducted more activities related to the product model (Huang, 2009).
Taiwanese teachers chose to carry out the product model which offered more samples and focused more on grammatical error correction for two possible reasons below (Chen, 2001; Kuo, 2004; Wu, 2006; Huang, 2009; Ferris, 1995b):
First, both instructors and learners in Taiwan considered that error correction should be done and had an expectation for it (Chen, 2001; Kuo, 2004; Wu, 2006;
Huang, 2009). Chen (2001) noted that most Taiwanese teachers considered that error
20
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
correction could help students write better. About seventy percent of Taiwanese teachers believed that the grammatical errors should be corrected (Kuo, 2004; Wu, 2006) and students took a positive view toward error correction (Huang, 2009; Leki, 1990).
Second, if teachers do not focus on grammatical errors, the ignorance toward errors leads students to undervalue the importance of accuracy and have them lose the opportunity to develop their error-detection skills (Ferris, 1995b). To prevent students from typical grammar errors, Taiwanese teachers tend to point out the grammatical errors more often than working on the content (Kuo, 2004).
Because of these two main reasons, grammatical instruction and error correction received attention in writing pedagogy in Taiwan. Though teachers agreed with the advantages of the process model (Huang, 2009), they still modified their teaching into a grammar-focused way and altered their feedback to meet students’ expectations (Kuo, 2004; Wu, 2006).
Teachers who choose to carry out the process model and apply the recursive revision process might face confused parents asking the reason for not pointing out the grammatical errors in every draft. Since students cannot gain good grades on writing exams, the teachers are also questioning the effect of the process model. As a result, how to deal with grammatical errors becomes a crucial issue.
21
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Grammatical Errors and Corrections
After discussing the concept of writing and the discussion of writing models, this section defines grammatical errors, common grammatical errors in EFL writing and different ways toward correcting.
According to Weinreich (1953), errors are considered as the phenomenon of interlingual identifications, and thus, are defined as interlanguage (IL). It is the misuse of the phonemic, grammatical or semantic features between two languages (Lennon, 1991). That is, errors happen in phonemic, grammatical or semantic ways.
For example, when errors happened in grammatical aspects, they were considered as errors in relation to the syntax, word structures and semantic rules (Huddleston &
Pullum, 2002), and often were the inaccurate language production that reflected the language competence of the learners (Corder, 1967). If a student repeated the same error, this means that he or she does not understand one certain grammar rule, and this repeated error then caused the concern of fossilization (Han, 2001).
The concept of fossilization emerged from repeating errors (Han, 2001), and is defined as a lasting mental and behavioral status representing linguistic segments and grammatical conceptions (Selinker, 1972). The incorrect application happens
habitually in productive competence, especially speaking and writing; in a long term, even when it seems eliminated, it still returns incidentally (Selinker, 1972).
Furthermore, Selinker (1972) noted that if errors are not corrected on time, learners might internalize the incorrect forms. The incorrect usage becomes a rhetoric habit and will be reproduced unconsciously again and again.
The main concern of fossilization is to greatly hinder the comprehension of readers and block the written communication (Liu & Qi, 2010). To reduce this phenomenon, Li (2009) proposed one solution: to offer the accurate language input.
22
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Offering the accurate input refers to pointing out the students’ grammatical errors and reinforcing their awareness of errors. While errors are pointed out repeatedly in written works, learners can more easily see the fact that they are duplicating the similar errors again and again. For this reason, teachers tend to point out the grammatical errors in students’ writing.
Common Grammatical Errors in EFL Writing
Students made different types of grammatical errors at different ages or competence levels (Chiang, 1992; Ferris & Roberts, 2001). In this part, the writing errors from university, senior high, junior high and primary school students are listed out, and then, those grammatical errors are compared in order to show the most common error in different contexts.
In a university context, Ferris and Roberts (2001) proposed that the most common error in university students’ writing was in the verb category. The verb
category refers to all errors related to verb tense, form and subject-verb agreement, e.g.
He go swimming yesterday; the cake is eat by the boy and she want the flowers. Since different errors stand for particular domains of language knowledge (Corder, 1967), the most common error means that the university students lack the knowledge of verb utilization the most.
In a senior high school context, Huang (2009) proposed that the most frequent error was misspelling. Misspelling contained the misspellings of past tense verbs, past participle verbs, plural nouns, blanks for missing words and Chinese characters, e.g.
John want hiking yesterday. In Huang’s (2009) study, the error ratio of misspelling, ranking the greatest number, was 3.14 words per 100 words; this number almost doubled the second error type, tense. Furthermore, the misspelling errors were not decreased after the teacher feedback treatment; on the contrary, it increased to 3.66
23
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
words; it showed that the teacher feedback did little to help with students’ spelling abilities.
From the junior high school contexts, Huang (1994) and Chiou (2005) noticed that the errors related to the future perfect construction are the most frequent errors;
the future perfect construction is used to express what will occur by a particular time in the future, for example, we’ll have finished the assignment by tomorrow. According to Huang (1994), the possible reason for the most common error is due to the
interference of L1 and the suggested solution is to clarify the relationship between the notional time and linguistic time to learners.
As to the primary school context, Hung (2007) analyzed the common errors related to students’ writing work in a primary school and categorized the errors according to James’ (1980) classification, such as the substance, the text and the discourse level. According to Hung (2007), the grammatical errors found in the text level had the greatest proportion among all the errors; furthermore, the grammatical errors were analyzed again and divided into seven morphological features, e.g.
As to the primary school context, Hung (2007) analyzed the common errors related to students’ writing work in a primary school and categorized the errors according to James’ (1980) classification, such as the substance, the text and the discourse level. According to Hung (2007), the grammatical errors found in the text level had the greatest proportion among all the errors; furthermore, the grammatical errors were analyzed again and divided into seven morphological features, e.g.