The Cognitive Approach

Awareness of the rules

Cognitive theory assumes that responses are also the result of insight and intentional patterning.

Insight can be directed to (a) the concepts behind language i.e. to traditional grammar.
It can also be directed to (b) language as an operation - sets of communicative functions.

A variety of activities practised in new situations will allow assimilation of what has already been learnt or partly learnt. It will also create further situations for which existing language resources are inadequate and must accordingly be modified or extended - "accommodation". This ensures an awareness and a continuing supply of learning goals as well as aiding the motivation of the learner.

Cognitive theory therefore acknowledges the role of mistakes. See Dakin's Novish lesson in which he sets deliberate traps in The Language Laboratory and Language Learning by Julian Dakin published by Longman 1973.


How much cognitive theory do English language teachers need to know?

Trainers of English language teachers can achieve practical coverage of cognitive learning theory by reviewing the history of language teaching, especially the period in the mid 20th century when "meaningful drills" were being advocated and the shortcomings of "meaningless drills" were being highlighted. Although drilling and rote learning became subject to considerable prejudice in some educational circles in the late 20th century, no language learner will proceed very far without recognition of language structure and nobody will succeed in learning much without practice and repetition. Knowledge of the "types of drill" which the accomplished language teacher or informed computer learning program can employ provide a full toolkit for anybody responsible for learning and teaching. A fuller examination of drills is therefore contained below.

Another ploy often used by teacher trainers is to put trainee teachers into the situations encountered by language learners. This is often done through demonstrations where new languages (of which trainees have no knowledge) are presented. As homework (especially on MA Courses where "Reflective Learning" features as a component) students are often required to learn new languages (and alphabets!) to a basic level. It is hoped that they will be reminded of the problems, especially the conceptual ones. Often, there is not enough time to do this on short teacher training courses. However, there is a famous chapter which trainees can read where the experience of learning a new language is simulated. This is Julian Dakin's introduction to "Novish" [ a fictitious language designed especially to simulate conditions experienced in real language learning situations ]. The chapter appears in "The Language Laboratory and Language Learning". This is possibly the best book ever written on language learning - the reference in the title to the language laboratory reflects a technology in fashion in the 1960s and 1970s and does not detract from the book's main treatise on language learning. The chapter on Novish is also reproduced in The Edinburgh Course of Applied Linguistics [ a four volume publication which may be easier to find than Dakin's book in second hand bookshops].

What are the principle drawbacks of mechanical or controlled drills and the ways of overcoming them?

The aim of language practice drills is to train learners to talk and to help them master the basic structural patterns of the target language. As a method of language practice, drills are difficult to reconcile when the language becomes "meaningless".

The drawbacks of meaningless drills:

  1. lack of context
  2. failure to offer learner an element of discrimination or choice
  3. failure to give rise to naturalistic speech
  4. they fail grammatically in many instances.

Lack of context results from behaviourist principle of focussing uniquely on form: the one-step-at-a-time approach which attempts to forestall mistakes. Unique focus on form may succeed in the controlled environment, but the benefits of structural learning may not be transferred into the real environment. Drills attempting to forestall mistakes show only positive instances of what can be done. Negative instances are not given. The meaning conveyed by an utterance (e.g. I'm not going) is a matter of the function of the sentence as a whole in the larger context in which it occurs. A sentence does more than communicate information. It performs a role both in relation to other utterances that have been produced and as part of the interactive process involving the participants.

Without this wider context, drills run the risk of overgeneralisation. They may cause as opposed to correct mistakes. The absence of an element of choice within a drill undermines the semantico-grammatical category of communicative function from which conceptual meaning is derived, thus inhibiting the learning process. When the only changes are vocabulary items controlled by prompts i.e. when drills embody invariant structural patterns, the given structures may just as well be represented by the sounds TUM and TE. [ Julian Dakin 1973 ]

Meaningful drills

In order to qualify as "meaningful", a drill must provide:

  1. A context for the utterance it contains - without context, there is a risk of over-generalization. [ As put by D.A. Wilkins ] "The meaning conveyed by an utterance is a matter of the function of the sentence (utterance) as a whole in the larger context in which it occurs.
  2. It should give rise to naturalistic language
  3. It should allow the learner some element of choice or discrimination.

Arguments for drills

Track record and variety of exercise-types

There have been many successful courses which have been largely dependent on drills. An early example was the Minimal Language Acquisition Programme, designed by Charles Fries and Richard Lado. A later example was "Streamline Departures" [Oxford 1979], a UK English course book with a remarkable long shelf life, though the orginal method recommended in the Teacher's Book depended on many of the following drill-types.

  1. Substitution drills merely require the learner to substitute in the previous response the word provided or embedded in the next prompt. The stimulus to which the response is trained is therefore the prompt taken in conjunction with the previous response. The prompts signal the internal changes and the series of responses set the pattern. For the teacher who sees the need for isolation and practice of mechanical production of sentences to improve learners' command of structure or pronunciation.
  2. Mutation drills require systematic changes in the form of words provided in the prompt before a substitution is made. They may therefore be useful in practising inflection of verbs or nouns, agreements between such constituents in the sentence as subject and verb, adjective and noun (in French & Spanish) and case endings.
  3. Transformation drills may embody the changes outlined above but also require at least the option of a change in word order, the addition or deletion of grammatical constituents and may exact the alternation of grammatical pairs. They can accordingly practise changes from affirmative to negative, changes in voice from active to passive, changes in mood, from indicative to interrogative to imperative to subjunctive and changes in sentence-type from simple to compound or complex. A further use of Transformation Drills is in the process of word derivation.
  4. Application relationships (relationships of reference) prompted by pictures, sound effects or knowledge of the world.
  5. Collocation relationships between vocabulary items in a sentence (involving any or all of its constituents) prompted by cue words or whole sentences. The relationship is exclusively verbal and responses depend on a knowledge of lexical inter-dependencies.
  6. Implication relationships between sentences prompted by whole sentences and requiring the substitution of synonyms, hyponyms, antonyms, converse terms or consequences in place of their antecedents. S a word or words R its/their counterpart.
  7. Consequence, Hypnonmy and Antonym Drills - S: This is a wonderful book. R: Good, I'd like to read it. S: This is a fantastic record. S: Good, I'd like to hear it. R: Felicity is a very nice girl S: Good I'd like to meet her.
  8. Synonymy Drills

The role of repetition - a principle of both behaviourist and cognitive theories of learning

Regardless of preferences for behaviourist or cognitive, most teachers would find a place for repetition (for purposes of practice & consolidation), comparison (differentiation through minimal pairs or paired grammatical forms). Depending on their nature and scope, drills may EITHER elicit sequences of unrelated sentences from the learner OR build up something which begins to look like connected spoken prose. Given sufficient definition of aims and the avoidance of monotony, artificiality and inefficiency, drills must surely contribute to language learning by virtue of their many useful applications.

The use of drills at different levels of language proficiency:

  1. Drills are likely to be useful at elementary level or in the "practice phase" of a lesson where limitation of the learning goal is desirable.
  2. Drills are likely to be useful at the intermediate level where practice, revison and checking of learning is particularly important.
  3. Drills are likely to be useful at the advanced level to diagnose and iron out a particular difficulty.

Drills may be tried with the whole class or used on an individual basis, perhaps for remedial purposes. The limitations of drills are clearly matched by useful possibilities.

Julian Dakin's ingenious use of drills in introducing us to Novish [an imaginary language]

a practical demonstration of language learning supported by Cognitive theory: a problem-solving approach.

Dakin's introduction to Novish (The Language Laboratory & Language Learning Longman 1973) is hardly a programme which invites "the minimum number of mistakes consonant with, and conducive to, learning new rules." Indeed, he readily admits that he was going deliberately out of his way to trap us.

The Novish structures, which contain the conceptual difficulties, perform such basic functions as identification and verification of class (N.B. whether a thing grows or not is of great social & cultural importance to native Novish speakers).

The behaviourist could not realistically avoid the "sademane" / "sadegru" distinction through selection or careful grading.

Dakin forces us into traps by including problem-solving in the drills he presents in his programme for learning Novish. Our mistakes very often derive from lack of conceptual awareness and failure to grasp important semantic criteria. However, Novish children make the same mistakes! The conversion of "rule" to speaking habit is likely to be a slower more conscious process in the case of L2 learners!

How important is it to understand the underlying rule for each step?

Under Skinner's model of language behaviour formed through the application of "habits", consciousness of underlying rules is not of any importance.

Chomsky's conception of language as "rule-governed" would imply that we must at the very least allow our students to induce the rules.

Carroll defines "rule" as "simply a formal, usually vebal, statement of the conditions under which something is expected to occur or not to occur under certain sanctions." He adds that it is a construct in some sense independent of actual behaviour.

Carroll illustrates this claim by citing the fact that people can speak a language without any conscious knowledge or application of the rules that underlie their language.

The importance of semantics conceptual awareness as a structurally-based basic language programme unfolds

Novish Frame 2 six different objects. "Sademane" is apparently used to define them. Insufficient knowledge of Novish to allow many L2 analogies, so we are tempted to measure each new item using L1 concepts as a gauge.

Novish Frame 3 introduces a refinement. Correct form is "Sademanena gal". However, the underlying rule is less important since a Novish speaker would probably understand our meaning if we said "Sademanena gal".

Novish Frame 6 introduces the use of "Sademane" in a question. It is noticeable that "Sademane" or "Sademanena" is replaced by "Sadestil" when verification is given.

At this stage, we think we know what is being verified just as we think we know what is being asked, but we are already on dangerous ground.

From this point, conceptual awareness of the distinction which Novish speakers make between things that grow and things that don't becomes increasingly important.

Novish Frame 9. The learner will quickly recognise "sadegru" as a second word he will sometimes have to use (as opposed to "sademane") in giving confirmation. Dakin has deliberately selected nouns which will lead to a false distinction: objects v people.

Whether it is justifiable for a teacher to lead his students into a trap and then to mystify them with "Ye sadegru opl" is a question in its own right.

At what stage should a teacher make learners aware of the rules rather than trying to trap them?

Clearly the proper distinction is one of some importance and any mystification should certainly not be prolonged beyond the point where Ss recognise that they have something new to learn.

To depend on "mim-mem" techniques to somehow unconsciously teach this distinction is clearly ludicrous.

It is widely recognised that learning language purely by imitation and repetition is uneconomical and that if each new speech pattern had to be learnt by imitation the task would be endless.

The catalogue of things which grow and don't grow is enormous and the structure under consideration is of fundamental importance and seems likely to allow further creation by analogy. Therefore in frames 9-12 the underlying rule must be realised.

The dangers of over-generalizing when forming new rules

Novish Frames 13 & 14 at first sight seem to be analogous to language concepts with which we are familiar. "Nu sadegru poi, sadestil tavl!" would appear to mean "No, it isn't a boy, it's a table.

Little do we suspect that the first phrase indicates that the table "doesn't grow like a boy".

We cannot develop a sound basis for further analogy until we have encountered steps 15 & 16.

Novish Frames 15 & 16. Here we learn that in comparing and contrasting different objects or people Novish speakers are vitally concerned with difference or even similarity of class as well as difference in identity.

The words "Ye" or "Nu" are applied essentially to class likenesses and differences and not to precise definition of what an object or person is or is not.

I can imagine many potential misunderstandings in situations where English speakers might use or take "Ye" to indicate a particular identity when what a Novish speaker understands is common membership of a certain group:

Q: Ki ku sademane? A: Ye sadegru ku, sadegru Margaret Thatcher!

Can language learning proceed without conceptual awareness and knowledge of culture?

Students should be given the chance to share the concepts of their target language. To deny them of what they are ready for, is to overlook what Chomsky recognised as the "creative aspect" of language use. Such a denial would serve to discourage creation by analogy, to kill the spirit of enquiry and to isolate the learner from a knowledge of the utterances which represent his achievements.

Classroom techniques: practical problems in (cognitive) learning and teaching:

  1. Could a particular class understand rules of the complexity of Dakin's for Novish and if they couldn't, what should the teacher do?

  2. To what extent can the teacher organise the examples for things so that the class can infer from them the "rule" without explicit explanation?

  3. How can the teacher be sure that a class or a particular learner has actually inferred correctly?

Parallels between Dakin's rules for Novish and the rules which elementary learners of English need to know

Not all rules met in elementary English classes are so complex as those of Novish. Many things in English are much easier to work out from examples than this, and so might not need such "rules".

There are still a number of things that do appear to require explicit explanation, such as "mass" and "unit" nouns, the contrast between Present Perfect and Past, etc.

Did all behaviourists imagine that language learning could proceed without formulations of rules?

Sophisticated behaviourists like Fries [ in Language Learning ] did not suppose that the mind was a mechanism of habits, and no more. Fries merely argued that, given that it was very sophisticated and subtle, the human mind was capable of inferring underlying rules if the examples were well-chosen. Fries thought that the best way to infer underlying rules was through practice (of the pattern drill type) supported by judicious explanation of rules at times. Read Fries' own introduction to English Pattern Practices.


  1. The Language Laboratory and Language Learning by Julian Dakin (Longman 1973)
  2. Teaching Oral English by Donn Byrne (Longman)
  3. Best "meaningful drills" ever published: Kernel Lessons Plus Laboratory Drills/Tapescript Longman Group Ltd (c) Eurozentren 1974
    These are set at the higher intermediate level. Superb use of situational context: e.g. Unit 10: law court as setting in which to practise Third Conditional forms.
  4. Most comprehensive series of mainly "meaningful" drills: Streamline Departures Speechwork (elementary level);
    Streamline Connections Speechwork (pre-intermediate to intermediate);
    Streamline Destinations Speechwork (intermediate to higher intermediate).
    These materials, published in the UK by Oxford Univesity Press in the 1980s, were widely used over a period of almost twenty years.
  5. Key Figures in the history of drills include: a) Harold E Palmer, b) Charles Fries and Richard Lado "English Pattern Practice" c) Comenius.
  6. History of theory - a) Skinner and Watson [See: B.F. Skinner's Verbal Behaviour 1957]
    b) See intro to Correct Your English B Mendelssohn & J.W. Palmer Longman 1940
  7. A History of ELT (second edition) - 1400 to the present, by A.P.R.Howatt with H.G.Widdowson (OUP)