Task-Based Language Teaching

I’ve recently suggested that one of the weaknesses of Michael Lewis’ “Lexical Approach” is that it provides no clear account of how it should be implemented in a syllabus. As a kind of antidote, I here present a very brief summary of an article by Doughty and Long . The article outlines how their “Task-Based Language Teaching” (TBLT) approach might be used for distance learning, but I’ve ignored that part of the article and concentrated on the methodological principles which inform TBLT. I should explain that this post was triggered by watching Mike Griffin interview Scott Thornbury at the KOTESOL conference last week. Scott mentioned Mike Long’s plenary where Long talked about the 10 methodological principles which lie behind TBLT, and which Scott reckons Dogme is also faithful to. I agree that Dogme does broadly follow TBLT principles, most importantly perhaps in being process-driven and in respecting “Learner Syllabuses”. So, below, the summary.

Doughty and Long start by saying that Task-Based Language Teaching constitutes a coherent, theoretically motivated approach to all six components of the design, implementation, and evaluation of a genuinely task-based language teaching program: (a) needs and means analysis, (b) syllabus design, (c) materials design, (d) methodology and pedagogy, (e) testing, and (f) evaluation. I should say that I haven’t gone into all six: needs analysis, for example, which is absolutely vital to TBLT, is not discussed. A distinction is made in TBLT between methodological principles (MPs) and pedagogic procedures. Methodological principles (MPs) are putatively universally desirable instructional design features, motivated by theory and research findings in SLA, educational psychology, and elsewhere. Whereas these MPs are language teaching universals, pedagogic procedures comprise the potentially infinite range of local options for realizing the principles at the classroom level. By way of illustration, Doughty and Long cite MP7: “Provide negative feedback,” which has the status of a methodological principle in TBLT. But how that feedback is best provided in any particular classroom is a matter of local circumstance. Options range from overt and explicit procedures (e.g., use of a rule or explanation delivered in oral, manual, or written mode, in the L1 or L2, or repetition of the correct response), through less intrusive ones (e.g., teacher “clarification requests”) to covert and implicit ones (e.g., manipulation of input frequency to increase perceptual salience, or the use of corrective recasts, of which students, and even teachers themselves, may sometimes barely be aware).

MP1: Use Task, Not Text, as the Unit of Analysis

Doughty and Long make clear that the focus in TBLT lessons is on task completion, not study of a decontextualized linguistic structure or a list of vocabulary items, or a text. Spoken or written texts, they insist, are static records of someone else’s task accomplishment. Building lessons around texts means studying “language as object”, not learning language as a living entity through using it and experiencing its use during task completion. Learners need to learn how to do a task themselves. Doughty and Long contrast two classroom activities: learning to make a particular kind of social, business, or emergency medical telephone call through acting one out, as in a role play and/or making a real one to given specifications, on the one hand, and on the other, in a text-based activity where the learners listen to or read a “dead” script of someone else’s effort.

One problem which arises when task is selected as the unit of analysis is sequencing of course material. This problem is hardly even addressed in most materials, being left to “some intuition-based and question-begging notion of linguistic complexity (e.g., teach the “simplest” structures first)”. In contrast Doughty and Long believe that “the ultimate solution, which is an important component of TBLT, will lie in the development of series of pedagogic tasks sequenced in terms of (inherent, unchanging, and objectively measurable) task complexity, with task difficulty (which varies for specific learners according to such factors as their L2 proficiency) modifiable as needed by alterations to task conditions (the circumstances under which the tasks are carried out). By working through the series of pedagogic tasks, learners can build up the abilities needed eventually to perform the target tasks identified by the learner needs analysis at the levels required”.

MP2: Promote Learning by Doing

The basic idea is that practical hands-on experience with real-world tasks brings abstract concepts and theories to life and makes them more understandable. New knowledge is better integrated into long-term memory and more easily retrieved if tied to real-world events and activities. TBLT is an example of learning by doing at several levels. It aims to equip learners to meet their present or future real-world communicative needs, as identified through a task-based learner needs analysis, the first step in course design. Then, inside the classroom, instead of studying the new language as object in order to use it to communicate at some later date, students learn language through doing pedagogic tasks. Pedagogic tasks combine language learning and action at various levels. Almost all pedagogic tasks have a hands-on, problem-solving quality designed to arouse learners’ interest and hold their attention.

MP3: Elaborate Input

Doughty and Long insist that both genuine and simplified texts are psycholinguistically inappropriate for learners. Genuine (popularly know as “authentic”) texts, originally written by and for native speakers, are usually too complex for all but very advanced learners. As a result, they typically require explicit metalinguistic study to render them comprehensible, which leads, in turn, to the study of language as object rather than development of a functional ability to use language. The traditional language teaching alternative, simplified texts, are unnatural and unrealistic in their tendency to be self-contained, with little or none of the usual implicitness, open-endedness, and intertextuality that characterizes authentic discourse. Also, while s implified texts are (in most cases) easier to understand than genuine texts, the improved comprehensibility comes at the cost of much of their value for language learning. How are learners to acquire items that have been removed from the input, and how are they to learn real NS use of new items if presented with something far less and unrepresentative?

The alternative to genuine and simplified texts recommended is elaborated input. “Elaboration is the term given to the myriad ways NSs modify discourse (i.e., in language use for non-native speakers [NNSs] to make meaning comprehensible, as revealed by studies of foreigner talk discourse). Most of the modifications occur during negotiation for meaning (i.e., when NS and NNS are focused on achieving communication while working cooperatively on a task). They include partial and complete, exact and semantic, self- and other-repetition; confirmation checks, comprehension checks, and clarification requests; rearrangement of utterances so that order of events and order of mention are iconic; paraphrase; lexical switch; decomposition; a preference for intonation and yes/no questions over WH questions; use of redundancy of various kinds; and many other “scaffolding” devices”. Elaborated input can be provided in advance (e.g., in the pre-scripted materials sources for pedagogic tasks), but also occurs naturally in teacher speech and in learner-learner discourse, as long as participants are focused on task completion and, therefore, on communication.

MP4: Provide Rich Input

Linguistically simplified input, which goes hand in hand with synthetic (especially structural, or grammatical) syllabuses, also tends to be impoverished input. Controlling grammar, vocabulary and sentence length results, intentionally and by definition, in a more limited source of target-language use upon which learners must rely in order to learn the code. The often tiny samples are worked and reworked in class, whether practiced until rote-memorized, milked meta-linguistically, or both, and learners are expected to learn the full language on the basis of access to such limited data. Elaborated texts go a long way towards remedying the situation. They alone are insufficient, however. Adult foreign language learners require not just linguistically complex input, but rich input (i.e., realistic samples of discourse use surrounding NS and NS-NNS accomplishment of target tasks).

This will usually mean task-specific and domain-specific target-language use not typically found in commercially published language teaching materials, not even those allegedly designed for language-for-specific-purposes programs. Commercial materials writers and publishers generally aim for the least context-, domain-, and task-specific texts possible, in order to boost the potential market for a book. This is the opposite of what is needed, especially if advanced, functional proficiency is the goal. Numerous studies have shown large discrepancies between the models presented in “general” textbooks and genuine NS use on real tasks in particular domains, even when those domains are relatively ordinary and “non-technical”. Learners need (a) elaborated texts, (b) plenty of them, (c) texts derived from a far greater range of target tasks and discourse domains than is currently typical in commercial language teaching materials, and most important of all, (d) texts motivated by tasks of the specific kinds a needs analysis has shown to be relevant. The examples will usually need to be based upon “field work” of various sorts by course designers (e.g., in situ audio or video recordings of NSs performing target tasks, and the gathering of authentic written documents relevant to those target tasks). Rich input, in sum, is not just a matter of linguistic complexity, but of quality, quantity, variety, genuineness, and relevance.

MP5: Encourage Inductive (“Chunk”) Learning

“If adult foreign language learners are to sound like natives, they need to be exposed to realistic (genuine or elaborated) samples of target language use, for example , as input components of pedagogic tasks, and then helped to incorporate, store and retrieve whole chunks of that input as whole chunks“.

I suggest reading the article to get the authors’ point here, but basically they propose that adults will abstract the language chunks that they need during the course of learning to perform the task at hand. They cite various studies on implicit learning of complex systems (Berry, 1997; Berry & Dienes, 1993; and Stadler & Frensch, 1998) where subjects are given input values and are told to arrive at particular output values by attempting to manage the system through trial and error. They are provided no information whatsoever concerning the underlying structure of the system, but are usually given feedback as to the effect of their input to the system. Some examples of complex systems often cited are the management of a sugar factory (variables are workers and amount of production), city traffic flow management (variables are bus schedules and parking lot fees), and interaction with a computer “personality” (where the computer person’s mood is dependent upon the input from the subject). The basic and consistent finding of this research is that subjects become highly skilled at managing complex systems long before they are able to explain the rules underlying those systems. Given enough time, they can be made to verbalize the rules that guide their own performance, but the ability to express the rules always develops after that, and, crucially, is not necessary for the improvements in performance evidenced in doing the tasks (see Doughty, 2003, for further discussion). “To date most published materials promote the explicit analysis of foreign languages. This approach results in declarative knowledge when what is needed is the development of language ability that is deployable during spontaneous interaction”.

MP6: Focus on Form

This is Long’s most well-known contribution to pedagogy. Given that research has shown that a focus on meaning alone is insufficient to achieve full native-like competence, (after as much as 12 years of classroom immersion, Canadian French immersion students’ productive skills remain far from native-like, particularly with respect to grammatical competence (Lapkin, Hart, & Swain, 1991)), we can conclude that comprehensible L2 input is necessary, but not sufficient. “A focus on meaning, moreover, can be improved upon, in terms of both rate and ultimate attainment, by periodic attention to language as object (Long, 1988). This is best achieved not by a return to discrete-point grammar teaching, or focus on forms, where students spend much of their time working on isolated linguistic structures in a sequence predetermined externally and imposed on them by a syllabus designer or textbook writer, in conflict with the learner’s internal syllabus. Rather, during an otherwise meaning-focused lesson, and using a variety of pedagogic procedures, learners’ attention is briefly shifted to linguistic code features, in context, to induce “noticing” (Schmidt, 1990, and elsewhere), when students experience problems as they work on communicative tasks (i.e., in a sequence determined by their own internal syllabuses, current processing capacity, and learnability constraints). This is called focus on form (Doughty & Williams, 1998a; Long, 1988, 1991, 1997, 2000a; Long & Robinson, 1998)”.

Examples of focus-on-form techniques, ranging from less to more explicit, include (a) input flood, where texts are saturated with L2 models; (b) input elaboration, as described in MP2; (c) input enhancement, where learner attention is drawn to the target through visual highlighting or auditory stress; (d) corrective feedback on error, such as recasting; and (e) input processing, where learners are given practice in using L2 rather than L1 cues. The most difficult practical aspect of focus on form is that, to be psycholinguistically relevant, it should be employed only when a learner need arises, thus presenting a difficulty for the novice teacher, who may not have relevant materials to provide. Where face-to-face interaction is the norm, as in L2 classrooms, recasting is an obvious potential pedagogical procedure. Once an L2 problem has been diagnosed for a learner, then pedagogical procedures may be decided upon and materials developed for use when the need next arises.

MP7: Provide Negative Feedback

While argument persists as to the necessity of negative evidence in language learning, recent work on both traditional explicit teacher “error correction” and implicit negative feedback in the form of corrective recasts (see, e.g., DeKeyser, 1993; Long, 2004) suggests strongly that negative feedback can be facilitative, at the very least, with certain classes of L2 structures.

Since the value of negative feedback lies in drawing learner attention to some problematic aspect of their interlanguage (i.e., inducing “noticing,” Schmidt, 2001), then the timing of that feedback is critical. Where corrective recasts are concerned, the information must be provided within some as-yet-little-understood cognitive processing window (for instance, but not necessarily, in working memory), such that learners can make some sort of comparison between the information provided in the feedback and their own preceding utterance (Doughty, 2001a). Recasts are proposed as an ideal (but not the only) form of negative feedback in TBLT for some classes of grammatical and lexical problems, at least, because they are not intrusive on the processing of meaning during task accomplishment and do not depend upon metalinguistic discussion of a language problem. Recasts are pervasive in child-adult discourse and in L2 classroom discourse. The psycholinguistic mechanism by which they are believed to work depends upon the juxtaposition of the learner utterance and the recast. It is claimed that learners have sufficient working memory to hold both utterances, thereby enabling the comparison to take place.

MP8: Respect Developmental Processes and “Learner Syllabuses”

Doughty and Long point to the strong evidence for various kinds of developmental sequences and stages in interlanguage development, such as the well known four-stage sequence for ESL negation (Pica, 1983; Schumann, 1979), the six-stage sequence for English relative clauses (Doughty, 1991; Eckman, Bell, & Nelson, 1988; Gass, 1982), and sequences in many other grammatical domains in a variety of L2s (Johnston, 1985, 1997). The sequences are impervious to instruction, in the sense that it is impossible to alter stage order or to make learners skip stages altogether (e.g., R. Ellis, 1989; Lightbown, 1983). Acquisition sequences do not reflect instructional sequences, and teachability is constrained by learnability (Pienemann, 1984). 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.

Equally well attested are the beneficial effects of instruction in such areas as accelerating passage through the sequences and extending the scope of application of grammatical rules (Pienemann, 1989), in dealing with areas of the L2 grammar supposedly unlearnable from positive evidence alone (White, 1991), and in generally improving accuracy, rate of learning, and level of ultimate attainment (Doughty, 2003; Long, 1988). The question, then, is how to harmonize instruction with the learner’s internal syllabus, with so-called “natural” developmental processes. TBLT does this in a variety of ways, first and foremost by employing an analytic, not synthetic, syllabus, thereby avoiding futile attempts to impose an external linguistic syllabus on learners (e.g., the third conditional because it is the third Wednesday in November), and instead, providing input that is at least roughly tuned to learners’ current processing capacity by virtue of having been negotiated by them during collaborative work on pedagogic tasks. The learner syllabus is also respected through use of (by definition, reactive) focus on form and a preference for recasts where the results are comparable with more overt forms of “error correction” , as their use implies learner direction to at least some classroom communication. In other words, not only in that course content is determined by student needs, but also in this psycholinguistic sense, TBLT is radically learner-centered. Universal developmental processes and the learner’s internal syllabus are clearly and consciously allowed to guide and mediate instruction.

MP9: Promote Co-Operative/Collaborative Learning

Research findings in both child L1A (Ochs & Schieffelin, 1979) and child and adult L2A (Gass, 2003; Hatch, 1978; Long, 1983) reveal a facilitative role in language development for collaborative, “scaffolded” discourse across utterances and speakers. Research in general education (e.g., Barnes, 1976; Holt, 1993; Webb, 1991) has documented the positive effects of co-operative, collaborative group work on attainment. Research on cooperative learning and small group work in second language learning provides similar findings (Jacobs, 1998; Liang, Mohan, & Early, 1998; Long & Porter, 1985; Oxford, 1997; Pica et al, 1996).

MP10: Individualize Instruction

Work by numerous scholars in general education and in foreign language classrooms has long shown the benefits of tailoring instruction to cater to individual differences in goals, interests, motivation, cognitive style, and learning strategies (Altman & James, 1980; Harlow, 1987; Logan, 1973; Sawyer & Ranta, 2001; Wesche, 1981). Improvements in the measurement of these and other individual difference variables, such as language learning aptitude and short-term memory (see, e.g., Ehrman & Leaver, 2001; N. Ellis, 2001; Grigorenko, Sternberg, & Ehrman, 2000; Miyake & Friedman, 2001), further justify the individualization of instruction in any language teaching program. In TBLT, individualization occurs in the selection of syllabus content, in respect for individual internal syllabuses, and in modifications of the pace at which and manner in which instruction is delivered, as suggested by diagnostic information gathered on individual differences.


I hope this has given those who only got Mike Long’s quick run through TBLT in his KOTESOL plenary a bit of additional information. I hope it also indicates the areas which any well-developed approach to ELT should cover. Whatever one think about Doughty and Long’s account, one must, I think, acknowledge that it is well-argued, thorough, and well-supported by research findings. It is also, of course, very critical of synthetic, product-orientated syllabuses and the materials they use, making it clear that such methods and materials fly in the face of research findings.

Please refer to the original article (link in red at start of this post) for references.

12 thoughts on “Task-Based Language Teaching

  1. Hi Geoff, thanks for the excellent summary of the article.

    I’m just wondering whether the link is working though. It seems to be highlighted in your post, but when I click it, it doesn’t seem to point anywhere (other links are working fine).



    1. Hello Geoff,

      Thank you for putting this out there and sharing it with us. I’m currently designing a training course for teachers in Korea, so this helps me look at various facets to our roles as language teachers. I’m grateful to Willy Cardoso for leading me here via his post https://authenticteaching.wordpress.com/2015/03/31/initial-teacher-training-do-they-still-teach-ppp/

      Like David, I am curious about that article you linked in your first paragraph. I just clicked “an article by Doughty and Long” and it still seems to be a dud.

      All the best,


      1. Hi Josette,

        Thanks for your comments. Sorry about the dud link. I tried to fix it, thought I had, but ….

        Here’s the link: http://www.llt.msu.edu/vol7num3/doughty/default.html I should add that these 10 principles are discussed at more length, as are all aspects of Long’s approach to TBLT, in Long’s excellent 2015 “Second Language Acquisition and TBLT”, Wiley Blackwell. I highly recommend this book for anybody putting together TBLT courses or training programmes.


  2. I’m curious why these 10 and why not Ellis’ 10? Both are supported by evidence, aren’t they? I’m also curious why the two men don’t have identical lists if the evidence is as clear as they seem to indicate.


  3. Hi Russ,

    I’m afraid I don’t know what Ellis’ 10 methodological principles are.

    The 2 men don’t have identical lists because they have their own views: these are principles, not hypotheses.

    You seem to expect rather too much of supporting evidence. There’s a lot of supporting evidence for the predictive power of astrology, but the evidence doesn’t persuade me. Does it you? What do you take the role of evidence to be in assessing different proposals?


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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s