1 Introduction
   引言

Writing has been neglected in college English teaching in China. Both teachers and students have not paid enough attention to English writing. There is no special English writing class designed for non-English majors in many universities due to the limited teaching time. Writing is incorporated into the teaching of intensive reading. As a result, many college students are still unable to write a coherent English paragraph after having learned English for many years. So there is great need to help college students improve their English writing competence.
在中國，寫作在英語教學中一直被忽視。老師和學生都沒有對英語寫作給予應有的關注。由于教學時間的限制，許多大學沒都有為非英語專業學生開設專門的英語寫作課程。寫作是英語精讀教學的不可分割的一部分。結果，許多大學生在已經學習英語多年后仍然不能夠寫出一段連貫的英語段落。因此，幫助大學生提高英語寫作能力勢在必行。
　　
Being aware of this problem, some researchers and English teachers try various approaches, such as process approach, task-based approach, portfolio approach etc. But these are put forward in the second language context and may not be suitable for Chinese context. Then some pioneers began to explore writing teaching methods that fit Chinese context. In 2001, Wang Chuming put forward ‘Length Approach’. But this method is helpful to the learners of advanced levels rather than the learners of low and intermediate levels.
　　
With the introduction of Swain’s Output Hypothesis into China, there emerge the solutions to this big nut. Some prescribe the way of integrating writing activities into intensive reading classes (Zhao Zhihui, 2003; Cai Mingde , 2003; Liu Dongli, 2007); others claim the way of integrating reading activities into writing classes (Deng Chunxu,2004; Liu Ye, 2006; Chen Qinyun, 2008). Both of them, in the writer’s mind, have some problems. The former think that integrating writing activities into intensive reading class is enough for improving writing ability, neglecting the importance of offering special writing course. The later claim that writing instruction should be carried out by reading the samples and thus they realize that writing instruction is necessary for writing improvement. But writing instruction is carried out independently of reading instruction. This method neglects the connection between reading instruction and writing instruction, which may lesson the significant value of writing instruction. Bearing these shortcomings in mind, the writer put forward Reading-Based Writing Instruction Model in hope that this method can help learners of different levels develop their writing proficiency.

　
2 Theoretical Background
   理論背景

Reading-Based Writing Instruction Model is based on two important but contrasting hypotheses in the field of second language acquisition.

2.1The Input Hypothesis
   輸入假設

The Input hypothesis, proposed first by Krashen, claims that there is only one necessary and sufficient condition for second language acquisition―comprehensible input. According to Krashen, if second language learners are exposed to multitude of comprehensible input, they will acquire the second language. There is no doubt that second language acquisition can’t take place unless learners access second language input. 2.2The Output Hypothesis
Since the input hypothesis was put forward, some researchers have been against it. Among them, the most influential one is Merrill Swain, who put forward the output hypothesis as opposed to Krashen’s input hypothesis.

2.2The output  Hypothesis
輸出假說

In a seminal paper, Swain (1985) argued that comprehensible input was not sufficient for successful SLA, but that opportunities to produce comprehensible output were also necessary. Swain claims that the role of learner production of comprehensible output is independent in many ways of the role of comprehensible input. Producing the L2 pushes learners to make their output more coherent and appropriate. In addition, producing the L2 may force the learner to pay attention to the means of expression needed in order to successfully convey his or her own intended meaning. Swain acknowledged the role of comprehensible input in SLA, but argued that comprehensible output was also necessary.
　　
Since then, Swain began to pay attention to the roles and functions of output. In 1995, she proposed three functions of output: noticing function; hypothesis testing function; metalinguistic function. Noticing function claims that while trying to produce the target language, learners may notice that they do not know how to say (or write) precisely what they wish to convey. This awareness triggers cognitive processes that have been implicated in second language learning. In this process, learners generate new linguistic knowledge for them, or consolidate their current existing knowledge (Swain & Lapkin, 1995). The hypothesis testing function refers to the fact that output can be used as a way of trying out new language forms and structures just to see what works and what does not. The metalinguistic function claims that using language to reflect on language produced by others or the self mediates second language learning.
　　
The appearance of the output hypothesis shifts the focus of research in the second language acquisition. Studies concerning output sprang out both abroad and at home. There are a lot of carefully designed experimental studies which have examined the effect of noticing function of output(e.g. Qi, 1998; Swain & Lapkin 1995; Izumi, 2000，2002; Izumi et al 2000). 3Brief Introduction of2.3 2.3Reading-Based Writing Instruction Model
　　
Reading-Based Writing Instruction Model is a method designed by the writer according to Krashen’s Input Hypothesis, Swain’s Output Hypothesis. The essence of this model is that the improvement of writing entails writing instruction and that writing instruction should be carried out based on reading instruction. In other words, the content of writing class, the reading materials used in writing class and the writing topic assigned to students are determined by what students have learned in reading class. This model aims to help students improve their writing competence through the interaction between writing instruction and reading instruction.

　　
It follows the unique procedure of “input1→output1→input2→output2”. “input1” refers to intensive reading instruction, where students can get plenty of comprehensible input through reading, while writing instruction includes output1, input2 and output2”. “output1” refers to the writing activity designed shortly after the intensive reading instruction, in which students are asked to write a short passage related to the text learnt in intensive reading class in topics or in genre; “input2” refers to the process in writing instruction, in which students acquire writing strategies systematically by analyzing reading materials and reanalyzing the selections from the text; in“output2”, students are asked to evaluate their products in output1 by using what they learned in input2 and revise them after class. In this model, output1, input2 and output2 are closely related to input1. It reflects the nature of reading-writing relationship: both dependence and independence between writing and reading. The following part will discuss the effectiveness of this model from theoretical perspective.
　　
3 Discussion
  探討　

Compared with the methods employed previously, Reading-Based Writing Instruction Model has more advantages.
　　
Firstly, this model follows the unique procedure of “input1+output1+input2 +output2”. And in this procedure, “output1+input2 +output2”(writing instruction) centers on “input1”(reading instruction)―the content of writing class, the reading materials used in writing class and the writing topic assigned to students depend on what students have learned in reading class. These are the unique features of this model. Such method is in accordance with the interaction of input, attention and output.
　　
According to Krashen, second language acquisition takes place only if the learners received a large number of comprehensible input. Sufficient comprehensible input is essential ingredient for second language acquisition (Krashen, 1985). In this model, input1, reading class, guarantees that learners are exposed to sufficient comprehensible input, a prerequisite in language learning. There, they may have learned many phrases and sentence patterns and had a sense of how to make sentences coherent and how to write. But the comprehensible input in reading class is far from enough. There is always the situation that learners often can’t get what they have been taught.
　　
Notice The Gap Principle, offered by Schmidt and Froda in 1986, can give a satisfying explanation to this phenomenon. It states that “only when they attend to the input, learners can learn it and the input became intake, which is “the subset of all input that actually gets assigned to our long-term memory store” (Brown, 2003). So providing enough input may not guarantee students can use the target language effectively. Then students are asked to write a short passage around topics related to the text learnt in reading class―output1. In the process of writing, they may try to use the words, phrases, sentence patterns or writing strategies learnt in reading class. This not only helps learners consolidate what they have learned, but also helps them notice some problems in learning. Swain points out that through output, the learner can check them whether they learned all the comprehensible input.

　　
Noticing the gap may stimulate learners to focus their attention on the following input. So after writing a short passage, students are asked to analyze a lot of reading materials in writing class to acquire relevant writing rules or strategies―input2. If, for example, the students don’t know how to write an contrast and comparison exposition after learning the text of this genre, there is likelihood that they will pay great attention to the input offered in the writhing class. The treatment of ‘output +input’ has been proved as the most effective way to acquire language. In 2002, Izumi have found that if learners were exposed to relevant input immediately after their production experience, the heightened sense of problematicity would lead them to pay closer attention to what was identified to be a problematic area in their interlanguage.
　　
According to Schmidt and Frota, for noticed input to become intake, learners have to carry out a comparison of what they have observed in the input and what they themselves are typically producing on the basis of their current interlanguage system. So, after acquiring the writing rules, students are asked to use what they learned to analyze their products and then revise them―output2. In this way, there is very likely that the input in the writing class can become intake and students can learn how to write a coherent and appropriate passage.
　　
To sum up, the procedure used in this model is designed in accordance with advanced language learning and teaching theories. And the interaction of writing instruction with reading instruction can help students to maximize the benefits from the input in both reading class and writing class.
　　
Secondly,this model gives writing an instructional emphasis, which is often overlooked in other methods aiming to develop learners’ writing proficiency. In current China’s English teaching, special writing course is not taken into consideration. Many educators think reading and writing are closely related with each other so that reading instruction is sufficient to accomplish the goals of both reading and writing. Unlike the traditional one, this model attaches considerable importance to the special writing course. The essence of it is that offering special writing course is imperative for the improvement of writing ability.
　　
This can be justified by Anderson’s ACT model. According to Anderson, skills are initially learned as a body of declarative knowledge which is then transformed into procedural knowledge through practice, and then automatized through further practice. “Learning a language, like any other type of skill learning, involves the development of procedures that transform declarative knowledge into a form that makes for easy and efficient performance”(Rod Ellis, 1994). It is clear that language learning begins with declarative knowledge which slowly becomes proceduralized. As far as English learning is concerned, the declarative knowledge is the knowledge about English language and its culture, which constitutes the learners’ cognitive structure. It is prerequisite for learning language skills. The lack of such declarative knowledge hinders the development of it. Therefore, in teaching English as foreign language, we should follow a cognitive way of moving from knowing to doing. The Reading-Based Writing Instruction model follows this principle. It emphasizes the act of offering special writing course, where students are provided with systematic writing theory and skills. Such kind of knowledge provides direction and guidance for learners in composing. When asked to write, students have more confidence and know how to deal with it. Such feelings of mastery and control over the learning activity stimulate learners’ drives to learn it.

　　
Apart from this, the emphasis of writing instruction accords with the notions put forward and testified by Shanahan that reading and writing are not only dependent but also independent. In 1984, Shanahan conducted an exploratory multivariate analysis of the relationship between learning to read and learning to write at an elementary school level. He concluded that “at any given point of development, reading and writing consist of both dependent and independent abilities”. Shanahan points out that writing instruction is not sufficient to teach reading, nor is reading instruction sufficient to teach writing. He further emphasizes that both reading and writing require instructional emphasis.
　　
Therefore, reading instruction alone can’t satisfy the need of improving both reading and writing proficiency. Writing instruction is necessary and indispensable in English teaching.
　　
Thirdly, the advantage of this model also lies in the fact that writing instruction in this model is carried out by analyzing reading materials. This reflects the other side of the nature of reading-writing relationship―their dependence. To be a good writer, it helps to be a good reader. Although the exercise of reading involves understanding why a text has been written in such a way rather than do actual writing, “understanding the structure of a text, for example, makes you a good reader, and helps to make you a good writer too”(Johnson, 2001).
　　
4 Conclusion
   結論
　　
From what have been mentioned above, it is easy to conclude that Reading-Based Writing Instruction Model is an effective way to help students improve their writing ability. First, this model follows the unique procedure of ‘writing a short passage (output1) shortly after the reading class(input1)―acquiring writing rules in writing class by analyzing the reading materials or selections from the text in reading class(input1)―analyzing by using the writing rules just learnt the selections from the texts in reading class ―evaluating and revising the first output by using the writing rules just learnt (output2)’. In this procedure, the content of writing class, the reading materials used in writing class and the writing topic assigned to students are based on what students have learned in reading class. Writing instruction is closely related to reading instruction. This is the essence of this model. Besides, this method gives writing special instructional emphasis, which is often neglected in College English teaching in China. Third, students acquire writing strategies by analyzing multitude of reading materials.
　　
References：
[1]Brown, H. D. Principles of Language Learning and Teaching [M]. Beijing: Foreign Language Teaching and Research Press, 2003.
[2]Cook Vivian. Second Language Learning and Language Teaching[M]. Beijing: Foreign Language Teaching and Research Press, 1991.
[3]Ellis Rod. The Study of Second Language Acquisition[M]. Shanghai: Shanghai Foreign Language Press, 1994.
[4]Izumi, S. Output, Input Enhancement, and the Noticing Hypothesis: An Experimental Study on ESL Relativization [J]. Studies in Second Language Acquisition 24(2002): 541-577.
[5]Johnson Keith. An Introduction to Foreign Language Learning and Teaching[M]. Beijing: Foreign Language Teaching and Research Press, 2001.
[6]Krashen, S. The Input Hypothesis: Issues and Implications [M]. London: Longman, 1985.
[7]Qi, Donald S. and Lapkin, Sharon. Exploring the Role of Noticing in a Three-stage Second Language Writing Task [J]. Journal of Second Language Writing 10(2001): 277-303.
[8]Shanahan, T. Nature of the Reading-Writing Relation: An Exploratory Multivariate Analysis[J]. Journal of Educational Psychology 76(1984): 466-477.
[9]Swain, M. Three Functions of Output in Second Language Learning[A]. Guy Cook, Barbara Seildlhofer. Principle and Practice in Applied Linguistics[C]. Oxford: Oxford University Press, 1995, 125-144.
[10]范燁.有關注意在二語習得中的作用研究綜述[J].外語界2009(2),56-65.