Abstract
There has been talk in TBL of the dangers of giving students target language before or during the pre-task because students may use the subsequent task to practice target forms and not focus on communication (Ellis, 2003, p. 246). Textbooks are often considered culprits in this predetermination of language forms (Willis, 1990), as model dialogs lock students into particular grammatical forms and restrict student vocabulary, thus reducing the communicative value of a task. This research addresses the concern of supplying learners with target language forms during the pre-task phase. Suggested phrases from the textbook were introduced before the task, but students were encouraged to also use their own ideas in task completion. Whether students deviated from or remained bound by the suggested forms and vocabulary during task completion was analyzed. Thirty-six student performances on a task were analyzed. Preliminary results indicate students use textbook language as a scaffold, employ unique vocabulary not included in the textbook, and do not vary grammatical forms.

Keywords: Task-based learning, focus on form, target forms, pre-task language focus

Introduction
Task-based researchers have long discouraged teachers from sharing target language during the pre-task phase of a task-based lesson. Summarizing the concerns of many, Ellis writes:

Despite these calls for caution, there is little or no literature that deals with the issue directly through research. Thus this study was conceived as a preliminary exploration of how learners use the language focused on before a task during task performance.

Literature
There are two aspects to the danger of focusing on language forms before task performance.
1. The language, as predetermined by the teacher, may not match native speaker models.
2. The students may look upon the rest of the task as a production phase, where they are expected to use the pre-specified forms previously introduced.

Regarding point one, Edwards (Undated) asked native speakers to perform a guessing task, where they discussed the possible function of unusual kitchen tools. Example language she predicted the speakers might use included modal verbs such as might be or could be. However, during the task the native speakers didn't use modal verbs. Instead their language included hedges such as It almost looks like a and It looks as though you can actually. Edwards concludes that in introducing target language forms before a task, those forms may not be typical of native speaker interaction.

The second point can be seen as contrasting TBL, where student focus is supposed to be on communication, and PPP, where students are expected to practice target forms. Criticism of PPP has become widespread and Skehan offers a concise summary of its shortcomings.

The point Skehan (1996) is making is that PPP, which selects certain forms which are first practiced through drilling before students are expected to produce the forms, ignores the fact that learners don't learn this way. Instead learner acquisition is dependent on internal factors, such as the current state of their internal L2 framework and the level of language pattern they are able to notice. Tasks, initially proposed as an alternative to PPP, are considered an improvement because rather than predetermining and drilling language forms, they offer a means of allowing learners to use the language currently available to them for genuine communication. During the course of this genuine communication learners experience a need to convey their intended meaning then after the task learners are encouraged to expand their language proficiency through consciousness raising (Skehan, 1996).

My context
While proponents of TBL make eloquent arguments against pre-specifying language forms, it is still necessary to adapt the theory of TBL to specific teaching contexts. In my own context at a private women's junior college in Nagano, Japan, there are two factors which potentially interfere with the implementation of a pure task-based lesson.
1. The syllabus for the course must be based on a textbook, and I am only a minor player in the textbook selection process.
2. A paper test on the contents of the course must be administered at the end of the semester.

Regarding the first point, if textbook selection were up to individual teachers, it may be possible to select a task-based text which would make employing tasks in the classroom easier. However, the textbook is chosen by a group of teachers, and is used across the curriculum. Regarding the second point, since the material in the textbook forms the basis of the end of semester exam, it is important to ensure that the content in the textbook is covered in class.

As a believer in task-based learning and its effectiveness in helping learners to improve their communicative ability, rectifying the disparity between the literature and the demands of my own context was difficult, particularly if "Tasks are genuinely not intended as practice activities, not even 'free' practice, of language presented earlier in the lesson or course" (Edwards, undated). Since my students would be tested on the material in their books, I felt it necessary to incorporate that language into the tasks, but was also interested in encouraging students to use original language as well. Thus this research was conducted to determine to what extent introducing students to textbook language before performing a task would affect task performance.

Methods
As this study was intended as a preliminary inquiry, students' written work was collected and analyzed. Completed written tasks from the top two streams of three first-year classes were chosen for this research. Since full tasks are often long and involved the activities in this research are more accurately pre tasks. Students were asked to write their responses on a worksheet then they shared their answers in small groups of five or six. While one member of the group spoke, the other members wrote the speaker's response on their own sheets. For a sample worksheet, please see Appendix 1. The task was performed twice and both performances were analyzed for this research.

Data was collected over the course of the 2004 semester but one particularly successful task will be analyzed here. In this task students first completed the picture task, where they chose a picture in their textbooks and explained why the person in the picture was studying English. Next students completed the personal task, where they explained why they studied English. A total of 36 students in two classes participated. Their completed worksheets were collected and analyzed to determine whether students used textbook language or unique language.

Results
Table 1 summarizes the results of student task performance.

Table 1: Results of student task performance

{1} One student unreadable

In one sense the concerns of the TBL literature were verified, in that no students varied their grammatical forms from the examples in the textbook. They used one of the two sample forms below in all instances.
1. I study English to…
2. I study English because…

However, when student vocabulary use was analyzed, there tended to be considerable variation, with a majority of students using original vocabulary in the picture task, and a more even split of textbook and original vocabulary in the personal task. Interestingly, students who used textbook language for the picture task tended to use original language in the personal task and vice-versa, indicating students were comfortable changing between using textbook and original language.

Discussion
The data is encouraging in that it seems to indicate students both used original language and manipulated textbook forms, meeting institutional expectations and realizing my desire to practice genuine communication during class. The use of original language and genuine communication, a prerequisite for successful task implementation, is encouraging. Also encouraging is the fact that students are manipulating and being exposed to the textbook forms which will be tested at the end of the semester. Furthermore, it was evident from student answers that they enjoyed the activity and at least some students strove to be original and interesting. Some sample student language is included below.

Sample picture task language:
1. to make ikayaki abroad
2. to be in the movies
3. to having family meal

Sample personal task language:
1. I study English because I like Western music
2. I study English because I wants to speak English!!! and I like English
3. to go junior college

Contrast these with the textbook vocabulary students were exposed to before the task (Richards, 1998):
1. to travel abroad
2. to pass an exam
3. to get a better job
4. to make new friends
5. because it's required
6. because I like studying languages
7. because English is an international language

It was also interesting to note that even when using textbook language forms, errors tended to creep into student writing. This may indicate that even though students are using textbook forms, they are "passing through fixed developmental sequences…which have to include often quite lengthy stages of nontargetlike use of forms…" (Long & Crookes, 1992) which are thought to be a prerequisite for language learning.

Conclusion
This study was intended as a preliminary investigation into what TBL literature considers the danger of supplying students with sample language before a task. The need for introducing such language arose when attempting to adapt TBL theory to institutional requirements. While the sample of students for this research is small and only one task has been included, the data seem to indicate that the low proficiency learners in my classes tended to be influenced by sample grammatical forms but could and did use unique vocabulary to complete the task. Before broader conclusions can be drawn, it will be necessary to analyze student speech during task performance and to expand the scope of research beyond my own local context.

While the task described in this research was successful in that students varied their language beyond that available in their textbooks, this is one of the few such successful tasks employed during a semester of study, indicating a number of factors must be successfully balanced in order to ensure a variety of language use. Unsuccessful tasks included:
1. Choose a piece of clothing or an accessory you want to buy and explain why you want to buy it.
2. Describe your ideal boyfriend.
3. Describe your ideal vacation.

In the above tasks student language tended to adhere strictly to the sample language in the textbook and when language did vary, the vocabulary used was often inappropriate and the intended meaning obscure. Several factors which inhibited student creativity and which should be kept in mind when designing similar tasks in the future include:
1. Clarity of task directions
2. Example language and task explanation complexity
3. Difficulty of textbook
4. Difficulty of concept

It is hoped that additional research into the issue of introducing sample language forms before task performance will confirm the observations of this preliminary investigation. If it does, then the attitude of TBL researchers toward pre-task introduction of sample language can hopefully shift from concern over potentially inhibiting learners to thinking of pre-task language as a tool to help liberate learners to be confident they have a scaffold to ensure successful task completion when their interlanguage framework fails them. Such a shift would hopefully ease the difficulty of implementing TBL in second and third ring countries as teachers could be confident they were meeting institutional requirements while at the same time facilitating the development of genuine communication skills among students.

References
Edwards, C. (Undated). Activity Cycle 2: What counts as a task? Retrieved on April 28, 2006 from www.delphi.bham.ac.uk.

Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University Press.

Richards, J. (1998). Springboard 1 Oxford: Oxford University Press.

Willis, J. (1990). The lexical syllabus. London: Collins COBUILD.

Willis, J. (1996). A framework for task-based learning. Longman: Essex.

Appendix 1: A sample complete task worksheet