1. What is a task?
In this context researchers make a distinction between exercises/activities and tasks. Ellis also refers to a distinction between task-as-workplan and task-as-process (Breen, 1989). The former is the materials which make up the lesson plan, including the instructions. He mentions the Heart Transplant Task where you give students information about four people and have to discuss and decide who is most deserving of a heart transplant. (This is a version of a balloon debate.) The task-as-process is the activity which occurs when students perform the task. I'm not sure how useful this distinction is myself for teachers, but the essence of a task is, according to Ellis, as follows:
a. The primary focus is on meaning, not linguistic form.
b. An information gap is needed, creating a need to communicate.
c. Learners should rely largely on their own linguistic resources; the point is not to teach new language, although students may use input from the task materials.
d. There is a clearly defined outcome to the task. The prime aim is not to use language, but to achieve a goal (e.g. the Heart Transplant task).
(My note: Bill VanPatten, the well known American researcher, who favours tasks with his own communicative approach, makes the point that a classic information gap activity may or may not be a task, since he argues that tasks should resemble real life situations to make them more motivational, less like a language exercise. Willis and Willis (2007) make the same argument, as Ellis points out.
In addition, as Ellis points out later, Widdowson (2003) criticises the loose definition of tasks. I would add that, in the end, does it matter if a classroom activity is strictly-speaking a task? Students can learn very well from interactive activities which are only tasks in the loosest sense.)
2. Origins of TBLT
TBLT is a development of CLT (Communicative Language Teaching) which grew up in the 1970s and can be seen as a significant area within CLT. This leads to two types of TBLT.
2.1 Task-Based Language Teaching
Here, tasks form the basis of the whole syllabus. Teaching is planned entirely with tasks in mind. Attention to form occurs within tasks, it is not the basis of the syllabus (in contrast with traditional structural, grammar-based syllabuses which many of us use.) This presupposes a view of language acquisition whereby learning is largely incidental or unconscious (think of Krashen or VanPatten).
2.1 Task-Supported Language Teaching
In this case tasks are used to supplement a more traditional structural course and there is a greater focus on the form of the language. This presupposes a strong intentional, skill-acquisition, or explicit basis for acquisition (think of DeKeyser). Learning may also take place unconsciously during a task of course.
(My note: in my own teaching I used tasks in this second way, as a means of providing interaction, practice and maybe experimentation with previously learned language. A simple example for near beginners is getting them to do an opinion poll about favourite subjects; an advanced example (via Penny Ur) would be the zoo reorganisation task where you give groups a plan of a zoo, explain a number of problems to solve and get the students to reorganise the placement of the animals.)
3. Theoretical underpinnings
One theoretical basis of TBLT and TSLT is to be found in the work of Michael Long (1996). His Interaction Hypothesis states is that interaction helps make input comprehensible and encourages students to modify their own output. In other words, when learners 'negotiate meaning' acquisition is facilitated when learners experience a communication problem. In addition, working within the skill-acquisition paradigm, DeKeyser (1998) stressed that learners need to use explicitly taught language features, e.g. a new tense, in real operating conditions, to develop automaticity.
(My note: in lay person's terms, when students have to listen and extemporise this helps them recall words and structures they know (retrieval), develop fluency and even formulate rules in their heads.)
4. Traditional v progressive?
Ellis points out that TSLT and TBLT can be distinguished in terms of the educational philosophies which underpin them. TSLT leans towards the traditional view of education as transmitting knowledge. Language is taught bit by bit and pieced together. TBLT accords more closely with discovery learning - a learner-driven pedagogy whereby students pick up the language through experience.
Long (2015) puts TBLT alongside philosophies which emphasise "guided individual freedom to learn, emancipation, learner-centredness,egalitarian teacher-student relationships, participatory democracy and the natural human inclination to behave cooperatively".
(My note: I wouldn't go along with this entirely. To me PPP might be labelled as 'traditional', but TSLT could be seen as a combination of traditional and more progressive. It depends on what the task is and the way a task is administered.)
5. Empirical evidence
5.1 Unproven hypotheses?
In response to criticism by Swan (2005) that TBLT relies on unproven hypotheses, supported by a lack of wide-ranging evidence, However, Ellis cites research which shows that 'online learning' takes place while students are doing tasks, e.g. Mackey & Goo (2007), Loewen (2005) and Mackey, Gass & McDonough (2000). In addition, Ellis refers to Pienemann's Teachability Hypothesis (1985) which argues that the traditional PPP approach is ineffective because learners may not be developmentally ready to learn new structures when they are taught. Ellis takes from this that TBLT has the advantage of allowing students to work at their own level of readiness.
Other research has focused on which types of task design and implementation make a difference, as what aspects of language are developed during tasks,e.g. accuracy and fluency. Reviews of such research can be found in Ellis (2003) and Robinson (2011). Skehan's (1998) Limited Attention Capacity Hypothesis and Robinson's (2001) Cognitive Hypothesis draw on models of working memory to show that students find it hard to focus simultaneously on form and meaning, tending to prioritise one over the other. During tasks students find it hard to produce language which is both complex and accurate.
Just to add that a recent study by Bryfonski and McKay (2017) looked at 52 studies of TBLT and found good statistical support for the approach. Of note, perhaps, is that only some of these studies were carried out in secondary school settings.
5.2 Comparing TBLT with PPP
Ellis acknowledges that there have been few studies comparing these two approaches. Shintani (2011, 2015) compared TBLT with input-based tasks with PPP on the acquisition of two grammatical structures with young Japanese learners of English. Both types of teaching were effective, but the TBLT group did better overall. Ellis points out that more studies are needed. In short, there is scant specific support for TBLT over PPP.
Researchers take different views on the extent to which teachers should include explicit teaching during tasks. Willis (1996) advised teachers to stand back and let students proceed independently. Ellis believes that some focus on form is useful during tasks
(My note: imagine how hard it is to design ethical longitudinal studies in this area while controlling for all the variables such as teacher influence and beliefs. Secondly, implementing tasks is a subtle business; the amount of intervention and attention to form will vary according to the class, mood of the class, nature of the task, amount of input provided with the task and so on. )
6. What does this mean fore the teacher?
Ellis provides a number of points of specific guidance:
- Don't be frightened to focus students' attention on form ('time-outs') - while they are doing a task, while trying to avoid the communicative flow.
- Focus on input-based tasks with beginners.
- Use input-based tasks as a preparation for an output-based task.
- Sequence tasks according to their difficulty and your intuition as to what will be the right level for your class.
In my experience Task-Supported Language Teaching can easily be built in to almost any syllabus, and certainly within GCSE and A-level in England and Wales. The younger the student, the more tightly controlled the task needs to be. The simplest information-gap task or guessing game can be viewed as a task if you want it to be - it has an aim beyond the use of language and usually focuses on meaning rather than form. At advanced level you can set up tasks and largely let students get on, answering queries when they come up, or jollying pairs or groups along when they run out of steam. Full-blown TBLT would be impossible to implement in England and, by the way, runs counter to the latest advice from the TSC (2016) which favours a skill-acquisition approach to language learning.
For three examples of task-based lessons see here.
My main reference is given in the introduction. You could try following up the abbreviated references in the text, but few will be open access.