The Shibboleths of TEFL, or Sense and
Nonsense in Language Teaching
by Mark Lowe

A shibboleth is ‘an old idea, principle or phrase, that is no longer accepted by many people as important or appropriate to modern life’. (Advanced Learners Dictionary). The discourse of Teaching English as a Foreign Language is riddled with shibboleths that distort our thinking and disrupt our teaching methods. The aim of this paper is to expose the shibboleths, straighten out our thinking, and free our methods from obsolete and mischievous ideas.

Let us start with quotations from key language thinkers to provide some theoretical background – and ammunition - for the shibboleth hunt.

Wittgenstein (a key 20th century philosopher, and the pioneer of a philosophical view of language based on function and use rather than abstract system. He also pioneered the role of philosophy as ‘language therapy’, sorting out confusion in our thinking caused by muddles in language):

the meaning of a word is its use

Language is an instrument.

Speech… is part of the web of human life, interwoven with a multitude of acts, activities, reactions and responses

If a lion could talk, we would not understand him

Grammar is a free-floating array of rules for the use of language… It is not answerable to the nature of reality, to the structure of the mind or the ‘laws of thought’. It is autonomous.

Philosophy is a battle against the bewitchment of our intelligence by means of language

Philosophy is a fight against the fascination which our forms of expression exert on us.

We are struggling with language

The aim of philosophy is to let the fly out of the bottle

Searle (a leading American contemporary philosopher of language, with special interests in speech acts, the social role of language, and the neuro-physiological foundations of language):

When we think about language, much of our vocabulary is obsolete and our assumptions are false.

Many of the currently fashionable views about language and the mind are inconsistent with what we know about the world

Mental phenomena are caused by neuro-physiological processes in the brain.

In our skulls there is just the brain with all of its intricacy, and consciousness with all its colour and variety. The brain produces the conscious states that are occurring in you and me right now, and it has the capacity to produce many others that are not now occurring. But that is it. Where the mind is concerned, that is the end of the story. There are brute, blind neuro-physiological processes and there is consciousness, but there is nothing else. If we are looking for phenomena that are intrinsically intentional but inaccessible in principle to consciousness, there is nothing there: no rule following, no mental information processing, no unconscious inferences, no mental models… no language of thought, no LAD and no innate or universal grammar.

Halliday (leading applied linguist, a pioneer of functional language theory, and author of An Introduction to Functional Grammar, Learning How to Mean etc)

Language is a means to an end, rather than an end in itself.

Language has evolved to satisfy human needs, and the way it is organized is functional with respect to those needs – it is not arbitrary. A functional grammar is essentially a natural grammar, in the sense that everything in it can be expressed ultimately by reference to how language is used.

The fundamental components of meaning in language are functional. All language is organized around two main kinds of meaning: (a) the ideational or reflective, and (b) the interpersonal or active. The first enables us to understand the environment, and the second to act on each other. .

All the units of language – its clauses, phrases and words etc – are organic configurations of functions.

A language is interpreted as a system of meanings, accompanied by forms through which the meanings are realized.

Language is natural. It reflects experience, eg process = verb, and participant = noun.

Language is an evolved system, and not a designed one. There is congruence between language expressions and the facts in the world it reflects or relates to.

Linguistics is in the same condition today as Physics was in the 15th century

Let us summarise these ideas. The way of thinking about language that informs this article is based on insights from both philosophy and linguistics. It incorporates ideas from Searle’s Theory of Mind, Halliday’s functional linguistics, and Wittgenstein’s Philosophical Investigations (which takes a basically functional view of language). It is critical of Chomskyan psycholinguistics and it follows Wittgenstein’s own rejection of the ideas in his early Tractatus (which interpreted language as a logical system reflecting the ‘logical’ structure of reality). This view of language grounds our thinking in verified truth rather than in metaphysics, in science and not in myth.

So much for preliminaries. Now for the shibboleth hunt.

1. Acquisition and Learning

No word in our field is more laden with ideological baggage than acquisition. The definition of the word that has bewitched TEFL comes from Krashen and Dulay’s Language Two, in which acquisition is sharply differentiated from learning. ‘Acquisition’ is said to take place in the intuitive right hemisphere of the brain, while ‘learning’ is said to take place in the analytic, language-controlling left side of the brain. ‘Acquisition’ activates the universal grammar hard-wired into the LAD (Language Acquisition Device) of the brain, while learning is an entirely separate process, unrelated to language acquisition. In other words, we pick up language unconsciously, while we study (or learn) language consciously. Acquisition is natural, while learning is unnatural.

All this is myth. There is no universal grammar, no LAD, no language hard-wired into the brain, no unconscious learning, and no ‘acquisition’ in the sense claimed by Krashen and his followers.

Of course, we do pick up languages through use, through enjoying songs, through conversation, through reading books etc. However, the true explanation for this process is not Krashen’s acquisition theory, but the normal process of learning and mastering new things. We learn what we are interested in and what matters to us, whether languages or music or how to play games or science or whatever. This ‘picking up’ is no different in principle from learning. We explore this idea in more detail later (see especially the section on Inductive and Deductive reasoning)

Halliday recommends that we drop the loaded terms ‘acquisition’ and ‘learning’ and adopt the neutral term ‘language mastery’ instead. Our profession would do well to follow this excellent and timely advice.

A methodology based on the idea that language acquisition takes place through letting the intuitive right hemisphere of the brain do all the work is a mischievous illusion. ‘Acquisition’ in this sense is our first shibboleth.

2. Deep and Surface Structures

There are ‘hard’ and ‘soft’ versions of this theory. In the hard Chomskyan version, deep structures are hard-wired into the brain. They are manifestations of the universal grammar that is thought to be common to all mankind. They generate surface structures – a process that can be depicted through tree diagrams. Deep structures are still language structures – they are part of the whole system which makes language possible. The soft version, on the other hand, sees ‘deep structures’ as part of our general mental universe (ie including features other than language), rather like Freudian subconscious drives. For instance, the surface structure of Hamlet deals with power and jealousy, while the deep (and hidden) structure is concerned with the Oedipus complex (in one celebrated reading). The soft version sees so-called deep structure as psychological rather than linguistic. This is the version that electrified a million cocktail parties a generation ago. It is an interesting, useful and valid idea.

The thinking behind this paper leads us to the conclusion that whereas ‘soft’ deep structures are a valid concept, hard Chomskyan deep structures are chimeras. They are our second shibboleth.

3. Innate Grammar

Interested readers are referred to an earlier MET article of mine on this topic for a more detailed study. Briefly, grammar is not innate, but learned. There is no reliable evidence to support the innate grammar hypothesis, and such evidence as is available can be explained more convincingly by evolutionary and functional hypotheses. Innate grammar is our third shibboleth.

4. Humanistic Approaches

These include The Silent Way, Suggestopaedia, Counselling Learning, the so-called Natural Approach, and related fringe methods. Again there are hard and soft versions of the theory behind these approaches. The hard version of the theory holds that such methods work by encouraging right-hemisphere acquisition to take place, while suppressing left-hemisphere ‘monitor’ learning. This version rests on a Chomskyan view of the structure of the brain and the basis of language: deep structures, hard-wired innate grammar, the LAD and the Monitor Model, etc. The soft view simply holds that people learn more effectively if they are comfortable, warm, well fed, and generally happy and appreciated: there is no elaborate supporting theory behind this view. If the ideas behind this paper are correct, it follows that the hard version is false, while the validity of the soft version can be proved or disproved by straightforward empirical research. Is the soft version true? Up to a point… But there are other views. Consider the words of my old headmaster. “Why did the Roman Empire fall”, he used to ask us, “too many baths… people had too many baths. They got soft”. And we all understood that what applied to the rulers of the Roman Empire applied equally to boys trying to learn their Latin declensions. Life is earnest, life is real...discipline, commitment, hard work.… And consider also these words of Baroness Thatcher: “if you want people to work, keep them cold; and if you want them to work hard, keep them hungry too.” If there is any truth in what the headmaster and the baroness said (and who would dare to contradict them?), the soft version is true only up to a point.

To conclude: while there is partial truth in the soft version of humanistic language learning, there is no scientific evidence to support hard-core humanistic approaches - they rest on a false view of language and the mind, and they are therefore built on sand. Hard-core humanistic language teaching methods are our fourth shibboleth.

5. Inductive and Deductive Learning

The problems here stem from confusion about how the brain works and about what actually happens when people think and learn. Let us recap basics. Deductive reasoning means starting from a rule, and deducing a conclusion from it. For instance, if the third person singular of the present simple has an s on the end, and a sentence requires a third person singular present simple verb, then that verb will have an s on the end. Or if a sentence requires a second conditional with the verb ‘to be’ in the third person, it must be ‘was’, and ‘were’ is wrong (wait a moment, please). We start from the rule, and work out the answer from the rule. Inductive learning, on the other hand, starts from the data, and then works out the rule from the data. We notice started, finished, lived, died – so we work out that the past tense adds ‘ed’ - inductive. (Hang on a minute..). TEFL is generally in favour of inductive rather than deductive reasoning as a means to help us master languages. Deductive reasoning = grammar-based learning, and that is old-fashioned and reactionary and bad. Inductive reasoning = learner-centered education, .discovery learning, tasks etc – and this is progressive and modern and good.

This cut-and-dried distinction between deductive and inductive learning distorts our understanding of the way the human brain actually operates. Karl Popper, in his The Logic of Scientific Discovery, proposes a bold and integrated inductive/deductive theory of discovery which has convincing explanatory power. Suppose we are trying to learn the second conditional. We start with a fairly vague idea of how it works: we have heard people using the word ‘if’ and the past tense – we are not completely clueless. We are a bit woolly about the details, but we start from a simple theory (deductive). We then find examples in use and compare them to the theory (inductive). We refine our theory (deductive). We then notice that people tend to say ‘if I were you’ as well as ‘if I was you’ (inductive). We devise a more elaborate theory to account for this, and take on board the notion of subjunctive mood expressed by past forms (deductive). We then listen to people using all the conditional forms in a piece of extended discourse (inductive). We fit the details into a broader framework that explains the whole conditional system (deductive). The same underlying process happens when we gradually assimilate the many irregular forms of the past as well as the regular ones, and explain some of them (ie we see mini-regularities as well as the major one) by reference to German and other roots, and we go on to understand the many functions of the past form other than expressing simple past time (deductive and inductive together). . .

Karl Popper developed his theory to explain how the brain makes scientific discoveries. But his theory can also explain a much wider set of phenomena. It explains how the brain actually works when it is learning something new or solving problems. The brain conducts a sort of dialectic between deductive and inductive reasoning in a lengthy process of refinement, testing against experience, further refinement, further testing etc etc. If this idea is correct, it ought to have a profound influence on the way we teach. We should employ both deductive and inductive reasoning together. We should not favour inductive reasoning at the expense of deductive reasoning.

The shibboleth here is the notion that only inductive reasoning is useful in language learning.

6. The Language Acquisition Device, or LAD

This myth has already been exposed, and further discussion is unnecessary here. The LAD is our sixth shibboleth.

7. The Natural Order

Is there a ‘natural order of acquisition’? If so, how do we explain it? Krashenite theory holds that the so-called natural order of acquisition reflects the way the Language Acquisition Device functions. We learn a language in the order dictated by the workings of the LAD: first isolated nouns and verbs, then pronouns, then x, then y, and finally z (eg third person s in the present simple tense) – or whatever. The LAD is like an old LP gramophone record. You stick it on the machine, and it plays in the order dictated by the grooves. The natural order is held to apply to L2 learning as well as to L1 learning. Quite detailed empirical research is cited in support of this theory.

However, the evidence used in this research is open to criticism. Firstly, it analyses only grammar: it takes no account of vocabulary, or pronunciation. Secondly, it ignores meaning and functional explanations such as ‘use’, ‘importance’, ‘interest’ etc. Third, the evidence is limited to few languages – mainly English and Spanish. The evidence thus provides only a severely limited account of language development. Alternative explanations of the order of language development are given in such language development studies as Halliday’s classic Learning How to Mean. The so-called ‘natural’ order of first language development can be explained through measurable and scientific categories such as easy/difficult, interesting/boring, simple/complex, important/peripheral, and – in first language development studies – by whether the child’s brain is yet able to handle the concepts behind the language. Children first learn what is simple, important to them, interesting – and within their mental grasp. Adults learn what is important and relevant to their needs too, but they can also master quite complex forms at an early stage..

This way of explaining the order in which adults develop a command of languages accounts for the many stories of language mastery that do not conform to the so-called ‘natural order’. Think of air traffic controllers, of musicians, of waiters, of priests learning the Latin Mass – and of ladies of the night. All have their own special language needs, and all master the language they need, ignoring much of the rest.

The true explanation of the so-called ‘natural order’ is functional need, not the LAD. Any language teaching method or material that attempts to teach language in the order dictated by Chomskyan theory ignores reality and is based on illusion. The Chomskyan ‘natural order’ is our seventh shibboleth.

8. Right-hemisphere and left-hemisphere learning

Let us begin this section by reminding ourselves of the theory. Psychologists and neurologists tell us, more or less, that the right hemisphere of the brain is the seat of patterns, intuition, art, music, dreaming, love and the universe of feeling etc. The left hemisphere of the brain, on the other hand, is the seat of analysis, logic, language, economics and science, etc.

Some language-acquisition theorists have used this model to develop a method of teaching that attempts to exploit right-hemisphere intuition and eliminate left-hemisphere analysis in order to teach language. In another words - to shut down the logical and inhibiting ‘monitoring’ left hemisphere of the brain, in order to allow free rein to the intuitive, uninhibited and creative right hemisphere. Krashen’s so-called Natural Approach is one example of such methods. Lie back on the carpet, listen to baroque music, eat chocolate cake – you’ll learn English (or French or even Chinese) double quick. Is this true?

Alas, no. This is simply not the way the brain works.

Relevant research has been carried out on the brains of musicians - research which is highly relevant to our concerns as language teachers. If you sit and listen to – say – Tristan and Isolde, in a darkened room, accompanied by an extremely beautiful girl, sitting on satin cushions, with a glass of Imperial Tokai in your hand, you will luxuriate in the tumultuous and erotic sounds with the right hemisphere of your brain. However, if you are one of the musicians actually playing or singing the music – and especially if you are the conductor who directs the whole performance, then the left hemisphere is very active indeed. Performing musicians use the left and right hemispheres together, each hemisphere helping the other. And the better you perform, the more you use the left hemisphere.

Most mental activity – including successful learning - involves the left and right hemispheres of the brain working together. So scientific research tells us, and such is our own experience. We start to learn a new language properly when our emotions are engaged and we want to communicate with real people, make friends, read the signs on shops, etc. That is the essential precondition (right hemisphere). But we make sense of the language data and remember it so that we can actually use it by analyzing it and seeing the regular patterns (left hemisphere). If this is so, it follows that language teaching methods should also combine left and right hemisphere approaches into an integrated whole. We can’t learn languages – any more than we can learn to perform Tristan – by sitting in darkened rooms and letting it all wash over us, even with the help of a very beautiful girl. Pity…...

Right-hemisphere learning theory is a last flicker from the dying embers of 1960s Californian hippy culture. It is time to move on.

Our eighth shibboleth is the idea that you can learn a language using exclusively the right (or the left) hemisphere of the brain. We must use both together.

9. Task-Based Learning

There are hard and soft versions of TBL, too. The hard version says that you must on no account teach a language form before performing a task: the language to be studied should evolve naturally out of the task. The soft version says that it is OK to teach the language first, provided you choose a task which generates that language naturally – and have a feedback session at the end.

The dogma seems to rest on the notion that the old Presentation/Practice/Production paradigm is out of date and inconsistent with sound learning principles. It is difficult to see any justification for this notion. The PPP paradigm holds for most things that we learn. If you are learning to ski, you practise your snow-ploughs before you go on the dangerous slopes. If you are studying Beethoven’s op.110 piano sonata, you work on your arpeggios first. If you are studying mathematics, you practise doing equations before you tackle a complex problem. Is there any reason why it should be different with language-learning? Are not the underlying mental processes similar in all these kinds of learning? Is language learning special? I know of no convincing evidence to suggest that it is.

Pragmatic TBL-based textbook series like Cutting Edge take a flexible and pragmatic attitude to this controversial question. In the elementary to intermediate stages, the language comes first and the task second. In the more advanced stages, it is the other way round. This strategy reflects the fact that at the early stages, the students do not have enough language to perform a task without some language guidance, whereas at the later stages, they do. The course is principled and undogmatic with respect to this question, and it is all the more effective as a result. All dogma is to be deplored, and TBL dogma is no exception.

Our ninth shibboleth is the dogma that tasks must always precede language work.

10. Unconscious learning

Confusion reigns here too. As we have already seen, the idea that you can lie back and dream, and thereby achieve language mastery, is an illusion. This is not how the human brain works. There is no such thing as ‘unconscious learning’ in that sense. However, ‘unconscious mental processes’ do exist and they do have a part to play in learning. Many scientists report how the solution to difficult problems sometimes comes to them after sleep. Artists and writers report that solutions to their creative puzzles often emerge only after sleep. Confronted with a complex problem, the brain churns away during sleep and sorts the data into a coherent pattern, thus revealing possible answers to the conscious mind next morning. However, this process only operates when the brain has data to work on: it does not operate without such data. To conclude: unconscious learning as a separate and autonomous activity is a myth, but the brain can and does sort data already stored in the memory while the conscious mind is resting. Or to put it in another way, isolated unconscious learning is a chimera, but unconscious sorting of material that has previously been consciously worked on is a reality.

This conclusion provides another argument in favour of a balanced method of teaching, combining communication and analysis – and against methods that rely exclusively on alleged ‘unconscious acquisition’. Unconscious learning divorced from cognitive analysis of input is our tenth shibboleth.

This is a difficult and complex topic. For fuller discussions of the issues involved, readers are referred to John Searle’s The Rediscovery of the Mind and Consciousness and Language, to Wittgenstein’s Philosophical Investigations and to Halliday’s An Introduction to Functional Grammar. .

11. Communicative Language Learning

So where does all this leave our fundamental ideas about language teaching and language learning? Is communicative language learning a shibboleth too?

Communicative Language Learning that relies exclusively on communication tasks without language awareness is not effective. A method that concentrates on language analysis without communicative language practice and tasks is equally ineffective. An eclectic method that combines communication and language awareness study is consistent with what we know about the functioning of the brain and how human beings learn, and it is effective. Communicative Language Learning divorced from cognitive language awareness study is our last shibboleth.

12. Conclusion

Because we are bewitched by our language, we language teachers are pulled in two opposite directions. In one direction lie the assumptions behind much of our TEFL meta-language, which was developed when Chomsky’s ideas dominated the theoretical scene. These assumptions lead us to think we ought to use humanistic methods, to activate the LAD, to promote right-hemisphere language acquisition, and to spend our time devising communicative activities for our students - and at the same time to eschew grammar, analysis, translation - and indeed anything reminiscent of the bad old days of the grammar-translation method. In the other direction lies our normal everyday practice, in which we make only occasional use of humanistic methods, in which we analyse grammar, devise carefully structured communication activities, and even translate at times. Modern scientific theory (and the choices of our market – our students and their employers and parents) – lead us to use such eclectic methods and to avoid the methods implied by our meta-language. So we are caught in a dilemma. Our meta-language pulls us towards beautiful humanistic methods. But modern science and our market pull us towards more cognitive and communicative methods. We feel we ought to be more humanistic, but we actually use more conventional methods. We are in a cleft stick. . . . We are like the inhabitants of Samuel Butler’s Erewhon. Mr Nosnibor and his wife and daughter visited ordinary banks and used ordinary money on weekdays, but they attended Musical Banks and used special Musical Money on Sundays and holidays. They believed that the Musical Banks were the really important ones, because the Musical Banks reminded everyone of profounder truths and deeper meanings than were preached in the ordinary banks. Everyone thought that they really ought to go to Musical Banks every day, but most people fell by the wayside and reverted to the ordinary banks during the week, feeling ashamed and guilty. .

We TEFL teachers are like Mr Nosnibor and his family. Our Musical Banks are humanistic methods and the beguiling metaphysical myths that underlie them: we pay homage to them in conferences and journal articles – but seldom in ordinary life. Our real banks are our ordinary practice and our ordinary textbooks and the scientific, empirical research that underlies them. We feel we ought to be doing all kinds of wonderful and imaginative humanistic things, but we actually do routine Headway-type analytical things. We feel inadequate as a result. We are on the horns of a dilemma.

The conclusions of this paper can resolve our dilemma. They move us from the 15 th to the 21 st century. They let the fly out of the bottle. They lead us to stop hankering after Musical Banks and to be content with ordinary banks. With the help of these conclusions, we are no longer led astray by seductive metaphysical fantasies: we tread the strait and narrow path of modern science. We are free.

Let us end with a brief summary of the main conclusions of this paper. Humanistic and extreme communicative approaches, and their attendant materials, are based on a false understanding of language and language development: they therefore rest on shibboleths. The eclectic communicative / language awareness approach of the Celta, Delta and similar teacher-training systems, and their attendant materials, are based on a broadly valid understanding of language and language development: they therefore rest on scientific truth, not on shibboleths.

Armed with these conclusions, we are no longer bewitched by the shibboleths of TEFL


Halliday: An Introduction to Functional Grammar (second edition, Arnold, 1996)
Learning How to Mean (Edward Arnold, 1977)
Popper The Logic of Scientific Discovery, (Routledge, 1956)
Searle Speech Acts, (CUP 1969)
The Rediscovery of the Mind (MIT Press 1994)
Consciousness and Language, (CUP 2003)
Wittgenstein: Philosophical Investigations (Blackwell 1953)

(Note. A version of this article was first published in Modern English Teacher, January 2005)


Mark Lowe studied philosophy at Cambridge University. He has maintained his involvement with the subject, and is particularly interested today in the uses of philosophy in sorting out real-world problems. He likes to play the piano in his free time

He is currently Director of Studies at International House, Tbilisi, Georgia.

To the original article

Back to the articles index

Tips & Newsletter Sign up —  Current Tip —  Past Tips 
Train with us — Online Development Courses — Lesson Plan Index 
Phonology —  English-To-Go Lesson  Articles Books
 Links —  Contact — Advertising — Web Hosting — Front page

Copyright 2000-2014© Developing