Humanising Language Teaching
Year 5; Issue 4; July 03

Skehan's proposals for task design in Task Based Learning/Teaching

Seth Lindstromberg, Hilderstone College,UK

In his book A Cognitive Approach to Language Learning (1998), Peter Skehan closely examines the issue of TBT/TBL in light of:

• various bodies of theory (i.e., Universal Grammar Theory, Learning Theory, Second Language Acquisition Theory…)
• research results.

I am devoting this article to a part summary of A Cognitive Approach to Language Learning for the reason that—although the proposals it offers should raise discussion of TBT/TBL to a new level of sophistication—this book seems to have attracted little of the attention I believe it deserves. And so what follows is my attempt at a preçis I hope will tempt people to read the book and make up their own minds about the implementability of what he advocates.

1. Skehan believes the following—

1. L2 learners already have a system of procedural language knowledgeàtheir mother tongue.
2. Older L2 learners typically have much greater cognitive abilities and knowledge of the world than children do when they learn their L1.
3. Use of the mother language may promote L2 learning.
4. Unlike L1 learners, older L2 learners will in general not make adequate progress solely through exposure to the target language, even if the exposure is meaningful.

2. Skehan accepts the distinction between mental representations of language (what is stored and how) and mental processes (how we access representations in 'real time'). Further, he believes that…

• Processes depend on representations.
• But representations can change as a result of processing; in other words, processing can cause changes in interlanguage.

3. Skehan's process approach to thinking about tasks derives from the belief that processing of language (if not also representation of language) is bi-modal—that is, it may be lexical or syntactical (see especially Pinker 1999). For example, when speaking we may combine words according to rules (syntactical processing) or we may just utter one pre- fabricated phrase after another (lexical processing). We are most likely to do the latter when we (1) are concentrating on the message we want to get across and (1) are under some kind of pressure.

4. Phases in the development of processing of language:

lexicalization => syntacticalization => re-lexicalization

This three stage evolutionary sequence goes on all the time with respect to different phrases.

The gist is that learners may first learn a phrase as an unanalyzed chunk. This is lexicalization. Later they may see how the phrase is structured; that is, they may see its internal grammar. This is syntacticalization. Even later, learners may once again process the phrase as an unanalyzed chunk since communication is faster if we are not always assembling/interpreting phrases word by word. This is re-lexicalization.

Skehan believes that all three stages are to be encouraged. In fluency tasks, learners tend to process language 'lexically'. For example, in speaking they may use key words and unanalyzed phrases they have learned—but they may make grammar mistakes, including ones they could self-correct if they were focusing on meaning. He is particularly interested in tasks which encourage syntacticalization—that is, with tasks likely to lead not so much to improvements in fluency but rather, especially, to readjustment of learners' interlanguage.

This is not to say that Skehan scorns fluency tasks; it's just that he thinks that not enough has been said about tasks which have potential to develop the complexity of interlanguage).

5. Believe it or not, SLA theorists and researchers seem to have discovered the importance of attention and noticing only a decade or so ago. In other words, it is only recently that much consideration has been given to the fact that change in interlanguage can happen only if learners are (1) paying attention to linguistic form and (2) that they actually notice particular formal features. The new wisdom is that tasks which create the conditions for attention and noticing are good.

6. Use of target structures in a task can be…

• natural

  That is, the structures can be used but can also be easily avoided without hindering the achievement of a desirable outcome. For example, we might devise a task that involves learners in talking about the weather tomorrow hoping that they will use the structures 'will + verb' and 'may + verb'. But they may carry out this task perfectly well using other structures. That is, there may be no loss whatsoever in efficiency or information transfer resultant from avoidance of the target structures.

• useful, or beneficial,

  In this case, the target structures make it possible to do the task especially efficiently even though a desirable outcome can be achieved even if these structures are avoided.

• essential

  In this case, a desirable outcome cannot be achieved except through use of the target structures. This is the kind of activity which designers of PPP lessons love to have in the 'Production' stage. Some proponents of blended TBL also favour tasks of this 'use-of-target-structure-is-essential' kind. But for many structures, such tasks are incredibly difficult to design. (Dave Willis argues that tasks of this sort, in addition, render impossible communication that is authentic.)

7. Skehan says that tasks may promote the development of fluency, accuracy or complexity but that one and the same task will not do all three things, or even two owing to the fact that learners have limited attentional resources. He notes, for instance, that research suggests that complexity comes only at the cost of (decreased) fluency and accuracy.

To say much the same thing in more detail--there is some evidence that…if students concentrate on the meanings they want to convey, they find it easier to speak fluently but they tend to make mistakes in grammar, pronunciation and wording--mistakes which, in other circumstances, they could self-correct. If, on the other hand, they concentrate on getting their grammar, pronunciation and wording right, their fluency is reduced. So--among other things--if learners have a diet of nothing but fluency tasks, it becomes probable that…

• their accuracy in speaking and writing will develop relatively slowly if at all
• their acquisition of new grammatical structures will develop relatively slowly if at all.

Conversely, if learners do nothing but accuracy exercises, it becomes likely that it will be their fluency that fails to develop satisfactorily.

A note on complexity--This is sometime measured by the number of clauses per c-unit. Skehan (and others) suggest that task designers who wish to help develop learners' interlanguage, should take complexity into account.

8. Drawing on the work of others, Skehan observes that tasks can be difficult in terms of—

• Linguistic features

• cognitive features—e.g.,
  --the number of elements and/or characters figuring in the task
  --the stability of the data (i.e., are they given new data during the task or not?)
  --the concreteness or abstractness of the task
  --the degree of familiarity of the task.

• situational features—e.g., learners may be working under the pressure of a time limit, or they may be speaking 'in the spotlight'.

Skehan believes that task design and task sequencing must take (7) and (8) into account in a detailed way.

9. Skehan's information processing approach to tasks (p. 129 ff.) is, in brief—

• “Choose a range of target structures.”
  “It is futile to fix on a particular structure and expect it to be learned. On the other hand, there is a need for systematicity in language development…”

• “Choose tasks which meet the utility criterion.”
  “Learners (and native speakers) are adept at evading the use of any particular structure…” So, a probablistic approach to aims selection.

• “Select and sequence tasks to achieve balanced goal development.”
  e.g., “[over] focus on meaning will make it less likely that continuing interlanguage growth will occur.” Tasks should address fluency, accuracy and complexity…but not all at the same time!

• “Maximize the chances of focus on form through attentional manipulation.”
  Desirable: “…a focus on form in the context of meaningful language use.”
  “At initial stages of task use, conditions need to be established to maximize the chances of noticing.”
  “Ensure that transacting tasks does not consume all attentional resources.”
  “There must be an opportunity for reflection and awareness….”

• “Use cycles of accountability.”
  “The key to making progress here would seem to be to draw learners into consciously engaging in cycles of evaluation.”

References/Further reading

Pinker, Steven. 1999. Words and Rules: The ingredients of language. Phoenix.
Skehan, Peter. 1998. A Cognitive Approach to Language Learning. Cambridge U Press.