A Closer Look at TBLT

The aim throughout these various posts is to explore how we should organise our teaching. I start from the assumption that nobody knows precisely how people learn English as a second language and there is no proven “best way” to help them in the task of learning it. Nevertheless, I suggest that we should take notice of the results of studies in SLA, and that a task-based approach to ELT is worth a close look.

The story so far

In April, I commented on the plenary Michael Hoey gave at the 2014 IATEFL conference where he suggested that his Lexical Priming theory provided compelling evidence that Michael Lewis’ Lexical Approach and Krashen’s Monitor Model were “true”. I made these counter-claims:

1. Michael Lewis’ Lexical Approach and Krashen’s Monitor Model are not true.
2. Krashen’s and Lewis’ models do not receive support from Hoey’s theory.
3. Hoey’s theory offends basic considerations of rational theory construction.

See the page *Krashen 1 for the full account.

There followed a discussion of Krashen’s Monitor theory where Krashen himself took part. See the pages *Krashen 2 to *Krashen 6 in the pages listed in red on the right of the screen.

I then discussed Hugh Dellar’s use of Lewis’ Lexical Approach in 2 posts, which can be found below. The gist of my argument was that Dellar fails to confront the limitations of Lewis’ approach, especially with regard to syllabus design. Subsequently, in Lexical Priming and the Competition Model, I answered questions which Nick had asked.

Finally, in an attempt to suggest an alternative to a lexical syllabus, I posted Task-Based Language Teaching, followed by A Closer Look at Task-Based Language Teaching. Part 1. In the first post, I summarised the 10 methodological principles which Long and Doughty claim should inform syllabus design. In the second post, I focused on Methodological Principle 8: “Respect Developmental Processes and “Learner Syllabuses”. I outlined the claim that those learning English as an L2 follow a route determined by the development of an interlanguage which is impervious to teaching. As Pienemann says “An L2 structure can be learnt from instruction only if the learner’s interlanguage is close to the point when this structure is acquired in a natural setting”. As a consequence, in Doughty and Long’s words, “The idea that what you teach is what they learn, and when you teach it is when they learn it, is not just simplistic, but wrong”. Those in the ELT profession are thus invited to revisit the question of what and how to teach, and, in a classroom setting, this means looking at syllabus design.

Syllabus Design

Breen (1987) classifies syllabuses into two basic types: “propositional” or “product” syllabuses and “process” syllabuses. Breen argued that product syllabuses concentrate on what is to be learned while process sylabuses concentrate on how the learning is to be done. Other academics developed Breen’s work and a distinction was made between analytic and synthetic syllabuses (Long & Crookes 1993), while White (1988) made a distinction between “Type A” and “Type B” syllabuses. First, let’s look at Long and Crookes synthetic and analytic syllabuses, since this is the distinction Doughty and Long make in their outline of TBLT.

Synthetic syllabuses chop language into discrete linguistic items which are presented separately and step by step on the assumption that acquisition is a process of gradual accumulation of parts until the whole structure of language is assembled and grasped. Synthetic refers to the learner’s task of re-synthesising the language which has been broken down into a number of small pieces so as to make learning easier. Lexical, structural, notional, topical, situational, and functional syllabuses are synthetic.

Analytic syllabuses offer the learner target language samples, which haven’t been controlled for structure or lexis. Analytic approaches are organized in terms of the purposes for which people are learning L2 and the kinds of L2 performance that are necessary to meet those purposes; they concentrate on what the learner does. The learner is assumed to have the ability to perceive regularities in the input and to induce rules. Procedural, process, and task syllabuses are examples of analytic syllabuses.

An alternative view is offered by White (1988), who distinguishes between Type A and Type B syllabuses.

Type A syllabuses focus on WHAT is to be learned. The syllabus designer divides the content into small pieces, and determines objectives. The syllabus is external to the learner, determined by authority. The teacher is the decision maker, and assessment of success and failure is done in terms of achievement or mastery.

Type B syllabuses focus on HOW the L2 is to be learned. They involve no artificial pre-selection or arrangement of items and allow objectives to be determined by a process of negotiation between teacher and learners after they meet, as a course evolves. The syllabus is thus internal to the learner, negotiated between learners and teacher as joint decision makers, emphasizes the process of learning rather than the subject matter, and assesses accomplishment in relationship to learner’s criteria of success.


Evident here is the clear bias shown by Breen, Long and Crookes, and White in favour of the process / analytical / Type B syllabus. We might well challenge one or more of the assumptions underlying this bias, but let’s at least accept that SLA research findings concerning interlanguage development seriously call into question the fundamental premise of product / synthetic / Type A syllabuses. Note that this includes lexical syllabuses which focus on “what” to the exclusion of proper considerations of how lexical chunks (for example) are woven into an EFL / ESL course. To be clear: syllabuses (and the coursebooks used to implement them) which, one way or another, present and practice bits of language in a sequential way on the assumption that these bits will be learned in the order that they’re presented – or learned at all – are based on a fundamentally erroneous view of SLA.

We may note that there’s also a conflict between Breen’s and White’s view and that of Long and Crookes. While Breen and White argue for an approach where the learner is central to decision-making, Long and Crookes (and subsequently Doughty and Long) are more concerned with matters which can easily be handled within a teacher-directed approach. That’s to say that there remains an important educational issue to be discussed – the issue, dramatically put, of education for liberation. Education, which ‘domesticates’ in Paulo Friere’s term, aims to deny people the right to ‘name their world’. Freire (1993) refers to this process as the ‘banking concept of education’. He describes it as follows: “Education thus becomes the act of depositing, in which students are the depositories and the teacher the depositor. Instead of communicating, the teacher uses communiqués and makes deposits, which the students patiently receive, memorise and repeat. This is the ‘banking concept of education’, in which the scope of action allowed to the students extends only as far as receiving, filing and storing the deposits” (Freire, 1993). Let’s park that consideration for now, and hope that Rose Bard will pick it up.

So let’s suppose for a moment that we’re persuaded of the folly of pursuing a grammar-based syllabus, or any other “product” syllabus. What then? An analytic syllabus relies on the learner’s putative ability to “perceive regularities in the input and to induce rules”, thus allowing the teacher to concentrate on providing the input and opportunities for learners to process it. But we know, from the results of research already cited in the 2 previous posts on TBLT, that input and practice aren’t enough: most learners, especially those in classroom environments, need to give explicit attention to formal aspects of the language if they are not to remain forever at some unsatisfactory intermediate level of proficiency.

Recall that Krashen says that SLA, regardless of age or environment, is essentially a matter of implicit learning; that Lewis mumbles his agreement; and that Hoey gives this dubious judgement his seal of approval. Recall also that Doughty and Long, basing themselves on the results of SLA research, beg to disagree. For Doughty and Long, explicit learning is important, although they think that the distinction between ”focus on form” and the more traditional “focus on forms” is important. So, during an otherwise meaning-focused lesson, learners’ attention is briefly shifted to linguistic code features, in context, to induce “noticing”, when students experience problems as they work on communicative tasks. As we saw in the TBLT post, examples of focus-on-form techniques include (a) input flood; (b) input elaboration; (c) input enhancement; (d) corrective feedback on error, such as recasting; and (e) input processing. It’s evident from a reading of Long’s successive work on TBLT that in classroom-based learning, where face-to-face interaction is the norm, Long regards recasting as the most powerful and effective pedagogical procedure.

Task complexity and difficulty

In Part 3, I’ll try to put together a task-based syllabus, but before doing so, we need to take a closer look at the core unit: the task. Doughty and Long see TBLT as comprising a series of pedagogic tasks sequenced in terms of task complexity and task difficulty, but it’s not always clear how these two constructs can be used to sequence tasks. A general finding in SLA research is that, as Robinson puts it “easier tasks tend to result in more fluent speech, since cognitive and processing demands are low. More complex tasks force learners to attend to the language used on task, resulting in less fluent but more complex and accurate production”. So far, so easy; but when we come to cognitive complexity, things are less clear. Some dimensions of cognitive complexity which have been proposed include:

a) Planning time : Tasks with planning time are easier than tasks without planning time.
b) Single versus dual task : Tasks making only one demand, such as describing a route marked on a map to another person, are easier than tasks with two demands, such as thinking up the route and describing it at the same time.
c) Prior knowledge : Tasks in a domain the learner has prior knowledge of are easier than tasks in a domain the learner has no prior knowledge of.
d) Number of elements: Tasks involve few elements are easier than tasks involving many elements)”.

Robinson takes things a step further with his Cognition Hypothesis.

The Cognition Hypothesis

Robinson’s “Cognition Hypothesis” claims that pedagogic tasks should be sequenced for learners in an order of increasing cognitive complexity on the grounds that this promotes L2 development and improvements in the ability to perform target tasks in the L2. Robinson suggests that cognitive complexity has three elements:

1) “Task Complexity” (cognitive factors affecting the cognitive challenge of the task)
2) “Task Condition” (interactive factors)
3) “Task Difficulty” (factors contributing to variations in the learners’ ability to perform a task).

In relation to Task Complexity, Robinson makes an important distinction between two subcategories. On the one hand, resource-directing variables direct learner attention and effort at conceptualization of ways in which the L2 system works; and on the other hand, resource-dispersing variables direct learner attention and effort to performance and procedural demands, without focusing on any particular aspect of language code.

The Cognition Hypothesis results in interesting claims for speech production. On resource-directing dimensions there will be increased accuracy, complexity and less fluency on complex versus simpler tasks. On resource-dispersing dimensions, complex tasks will elicit less accurate, complex and fluent production when compared to simpler versions.

In further work trying to tease out the dimensions of cognitive complexity, Robinson (2001) described a framework for making sequencing decisions in task-based syllabuses. He reports results of a study of the relationship between task complexity, difficulty, and production, which found that

1) Increasing the cognitive complexity of a direction-giving map task significantly affects speaker information-giver production (more lexical variety on a complex version and greater fluency on a simple version) and hearer information-receiver interaction.
2) Cognitive complexity significantly affects learner perceptions of difficulty (e.g. a complex version is rated significantly more stressful than a simple version).
3) Task role significantly affects ratings of difficulty
4) Task sequencing (simple to complex versus the reverse sequence) doesn’t affects ratings of difficulty, but it does affect the accuracy and fluency of speaker production.

The Trade-off Hypothesis

We can compare Robinson’s Cognition Hypothesis with Skehan’s “Trade-off” Hypothesis which claims that since attentional capacity is limited, attending to one performance area may take attention away from others. Under certain conditions, raised performance in one area may be at the expense of performance in other areas. Task difficulty will be associated with lowered performance in some areas, and complexity and accuracy compete. Skehan suggests that the syllabus designer should choose tasks and task conditions to promote the sequence Complexity — Accuracy — Fluency, such that new language is used, then control is gained over this language so that error is reduced, and then fluency-lexicalised language is achieved. Skehan further argues that we haven’t solved the problem of task difficulty, so we can’t rely upon sequencing tasks in a difficulty order. Skehan thus has a less theoretically-motivated view of task complexity and difficulty, which relies more on a fairly simple assumption of pervasive limited capacity. We may note in conclusion that Skehan stresses the importance of the start and finish stages of tasks:

* Planning: Train learners to get better at pre-task activities which anticipate problems, especially lexical, and rehearse effectively
* The post-task stage: “The aim is to nurture, consolidate, and complexify new language which emerges through performance. Use language whose salience has just been realised and practise it, build upon it, integrate it, recycle it” (Skehan, 2007).

My aim has been simply to indicate some of the issues involved when attempting to implement a particular approach to ELT. Obviously, the specification and sequencing of tasks is at the heart of any task-based syllabus, and in my next post, I’ll outline one such syllabus. I’ll also try to include what Swan said to Widdowson.


Breen, M. (1987) Learner contributions to task design. In C. Candlin and D. Murphy (eds.), Language Learning Tasks. Englewood Cliffs, N.J.: Prentice Hall. 23-46.

Freire, P. (1993) Pedagogy of the city. New York, Seabury Press.

Robinson, P. (2001) Task complexity, task difficulty, and task production: exploring interactions in a componential framework. Applied Linguistics, 22 (1): 27-57.

Skehan, P. (2007) Tradeoff and Cognition: Two hypotheses regarding attention during task-based performance. Presentation given at the Second International Conference on TBLT, University of Hawaii, Sept. 20th – 22nd 2007.

White, R.V. (1988) The ELT Curriculum, Design, Innovation and Management. Oxford: Basil Blackwell.


2 thoughts on “A Closer Look at TBLT

  1. Hi Geoff,
    This series of posts is particularly useful because it refers to such a wealth of studies in language acquisition. I have always found it strange how the synthetic syllabus has managed to endure in spite of all the evidence to suggest that linguistic complexity has no bearing on the order of acquisition. Whether the organising principle is grammatical or lexical, any kind of linear syllabus presupposes that all students will learn the same language at the same time, and it’s widely accepted that this is not the case. And if the natural order hypothesis is inconclusive, or if the order of acquisition varies from learner to learner, then any syllabus that tries to sequence language input seems rather pointless.
    In my view, a task-based/project-based approach to programme design, where the main focus or the main product is something other than language, is the only way that a syllabus can be truly communicative. As well as the points made above, a syllabus that uses language as an organising principle can be further criticised because it does not prioritise communicative competence. I have previously argued that this obsession with the linear syllabus is what stopped Communicative Language Teaching from getting off the ground in the first place:
    In fact, I’ve started to wonder whether there is any point in trying to sequence a syllabus at all. You can talk about task complexity, but if your organising principle is something other than language, a group that has been placed according to language level is likely to be very mixed-ability in terms of their ability to complete the actual task. I’m not saying this is a bad thing – I’m just saying that it makes it difficult to see a logical order in which to present these tasks. But maybe the order in which the tasks are presented is not so important.
    My own approach to syllabus design is, I think, moving towards something that these posts of yours seem to be advocating, Geoff. By this I mean the rejection of a linear approach to language input and a rather more ‘welcoming’ (some might say chaotic) approach to language focus within lessons. Is that what you’re advocating? Links below to some posts of mine, if you’re interested.


  2. Hi Steve,

    Thanks very much for this, and for the links to your own attempts to find a principled way of organising classroom-based ELT. I agree 100% with your view. I’m going to outline a task-based syllabus in the next post, but I’ll make a few comments here.

    First, needs analysis really should inform any course, so each syllabus will be different. Second, I’d adapt Breen’s process model to TBLT so that, on the basis of the data from a needs analysis, the teacher kicks off with “Task 1”, after which there’s a feedback and planning session where the teacher and learners together discuss what happened and plan Task 2. After Task 2 (which might last 4 to 10 hours) there’s another feedback and planning session and on we go.

    I think the different ways which Long, Breen, White, Robinson, and Skehan, for example, treat sequencing tasks suggests that it’s not safe to suppose that you can pre-plan a 100 hour course which might comprise of a sequence of 10 to 30 tasks. Rather, one needs a bank of tasks, each one with its own materials, which teacher and learners can choose from and then mould to their own uses. In practice, one might well be able to increase the complexity of the tasks as the course progresses.

    To those who say that such a syllabus is unrealistic, I’d say that while it certainly requires more effort than using one of the big EFL /ESL coursebooks, it doesn’t actually mean that much more work for the teacher. The crucial bit is getting a bank of tasks and their associated materials, together. There’s a ton of great material available these days without going near a coursebook; it’s just a question of selecting, assembling and organising all this material, which certainly does require an initial investment of time. If an institution supports teachers in implementing this kind of teaching, then it ceases to be “unrealistic”, and IMHO, everybody wins.

    I think that this kind of TBLT is in line with the Dogme approach and also chimes with the commitment to a learner-based approach so splendidly championed by Ruth Bard.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s