In his opening plenary to the conference Valorizing practice: grounded histories of language learning and teaching (Bremen 13-15 November 2019 see programme and abstracts), Widdowson starts with the observation that the problem is not that old men forget, but that they remember at tedious length, and so he intends to be concise on his talk on “New starts and different kinds of failure.” As a veteran of both British Council and Council of Europe language education campaigns, Widdowson aims to analyse what he consider two misconceived ventures.
He offers this TS Eliot quote which he sees as applicable to pedagogy as well as language and literature:
Trying to use words, and every attempt
Is a wholly new start, and a different kind of failure.
Among language teaching methods catalogued by Richards and Rodgers, on one hand, and Tony Howatt on the other, Widdowson focuses on two prominent new starts in language education which attracted institional recognition and promotion, and were established as orthodoxies difficult to resist. He cites
1. the communicative approach (CLT)
2. the authenticity doctrine
The Communicative approach: new start #1
He reminds us of Hymes’ (1972) 4-fold concept of communicative competence, which involves making a judgment about a sample of language and deciding how far it is
1. possible (conforms to encoding rules),
CLT was presented as a radical break from structuralist language teaching. The structural approach focused on the possible (the first of Hymes’ strictures), while in CL the focus shifts to the appropriate, that is, focusing on meaning rather than form: “Enough of form, enough of the possible. Up with the appropriate, and focus on meaning.” For Widdowson, however, this is not really new at all. Structuralism was already focused on meaning, as the basic activity of “situational demonstration” shows: “This is a book, this book versus that book.” This is not normal use of language, since the emphasis is on formal, or encoded semantic meaning, not contextual meaning. But for CLT proponents: if you didn’t focus on pragmatic meaning you were not focusing on meaning.
For Widdowson, what is possible in the code is related to what is appropriate in context, but in CLT the dependency is reversed. In the structural approach, texts were created in order to be appropriate to the code (semantic meaning) whereas in CLT, the dependency was reversed, and the code had to be appropriate to the context (pragmatic meaning). In CLT we start with context and find code to fit the context, thus focusing on a particular kind of meaning, contextual meaning. This means that this new start corresponds to a shift to the reverse dependency (not context invented to service code, but rather code used as appropriate to context)
If we consider the application of CLT by the Council of Europe via the Threshold level programme (van Ek 1975), we find very detailed specifications of which form is to be used as appropriate to which communicative function. For these context-code mappings, first communicative function is determined, then linguistic form (e.g., forms for inviting, advising others). This approach remains focus on form, but in a different relation to context. What are these specifications based on? Widdowson asks. It is undoubtedly a pragmatic approach to language use, but it is not based on actual observations of speakers, rather but idealised expressions based on individual intuition, and so reminiscent of Chomsky’s ideal speaker-hearers
The Authenticity doctrine: new start #2
The authenticity doctrine represents a move from the ideal to the real. Widdowson cites the role of Sinclair in the emergence of corpus linguistics leading to radical change in ways of describing language
The categories and methods we use to describe English are not appropriate to the new material. We shall need to overhaul our descriptive systems (Sinclair 1985)
We are teaching English in ignorance of a vast amount of basic fact. This is not our fault, but it should not inhibit the absorption of new material (Sinclair 1985)
Here we have a shift in Hymes’ framework: Sinclair assumes that what is appropriate to teach IS what is actually performed. His first precept says “present real examples only” Widdowson notes that an extract from the corpus carries the imprimatur of authenticity. He cites another corpus linguist, Ute Römer, who asks rhetorically: Should we teach our pupils authentic English? What can we do to improve EFL teaching materials? The expected answer, according to Widdowson, is “make them real.”
To summarise the story so far, the first new pedagogical start (communicative approach) shifts the focus from the possible to the appropriate; the second focuses on the actually performed. But in one crucial respect there has been no shift at all. What is understood by “real” is “contextually appropriate in native-speaking communities.” Teachers try to get learners to conform to these norms and fail, even in the most favourable and privileged circumstances, let alone in those parts of the world where these new starts are also promoted, thereby undermining teachers without providing something they can actually use.
Where do the new starts come from and why do they occur? Widdowson cites a paper by Tony Howatt (1982) Language teaching must start afresh!, echoing Viëtor (1882) Der Sprachunterricht muss unkehren! But Widdowson contests Howatt’s translation: umkehren means to turn back, retrace one’s steps, and perhaps find out where you went wrong. In the German expression, there is continuity, turning back to “the road not taken” to follow a new direction, or as the French have it “reculer pour mieux sauter.” In contrast, the new starts already discussed are not historically grounded but rather abrupt discontinuities. There is no examination of where methods in the past might have gone astray – to find out whether and why they failed – and there is no attempt even to evaluate any failure (was it universal, cataclysmic failure?)
These starts came from shifts in fashion in linguistic description and the assumption that these shifts should influence pedagogy. Sinclair was “a genius of linguistic description” but also assumed that pedagogical practice overhaul also required. His justification was the following:
the precepts centre on data and arise from observations about the nature of language. They are not concerned with psychological or pedagogical approaches to language teaching (Sinclair 1997)
Widdowson interprets the Authenticity doctrine as a shift towards not just the possible but also the appropriate and the performed in conformity with NS norms. He argues, however, that reality (authenticity) is a property of NS usage. The assumption that is should be directly transferrable to the FL classroom was unexamined. As with CLT, the precepts are presented as universally applicable and context-free and carry with them the persuasive imprimatur of authority (of linguistics and native speakers, who are viewed as custodians of right and proper language). For both the British Council and the Council of Europe, this pedagogical approach is seen as globally transferrable – any “local stubborness and backwardness” should be “resisted and removed.”
However, Widdowson argues, it is local contexts which set conditions on the validity of an approach. This is where we fail when we complain “If these “wretched teachers didn’t do what they were supposed to do, it was because they were clinging to old-fashioned methods.” But teaching needs to engage with local contexts. The crucial context is not the context of the NS language user, it’s the context of the language learner in the classroom. This is the reality that pedagogy needs to engage with: What of the context of the classroom? What kind of language is appropriate here? The structuralists were right: it has to be language that activates the learning process, whether it conforms to language appropriateness or not.
This brings us to feasibility, which has been neglected. This is the key feature of pedagogy: the language presented to learners has to be such that they can access and process it. A corpus can provide samples, but these are only useful to learners if they can process and authenticate them for themselves. Learners need language that is appropriate because it is feasible. The neglect of feasibility comes about because the new starts focus on language teaching (objectives) rather than learners and learning process (subordinated to that goal). Learning is taken to be the reflex of teaching – learning is the progressive achievement of prescribed teaching objectives. Other learning is seen as an impediment to progress. But learner initative can be seen as an attempt to make input more feasible.
A monolingual approach means other languages should be regarded with suspicion, but FL learning is always bilingual or multilingual: learners’ own L1 is always present and is always naturally involved in the process of L2 learning. Where the learners are busy multilingually learning, the teacher is busy monolingually teaching. Learners relate learning to what they know, assimilating the new into the familiar. Learning errors are evidence of deforeignising FL learning. If we give primacy to the learning process over teaching, then teaching is relegated to an ancillary activity, subordinate to learning (and not other way round). Here we have radical implications.
FL learning involves being taught language, stable and defined language, a distinct foreign entity. This is not teaching how language is used to communicate, but what form communication is supposed to take in the NS community. But if learning is always a development from previous experience, an association of the second to the first language, then it is simply the extension of what learners already know about language by applying it to new linguistic data. Learners are in effect learning different realisations of language, rather than learning A different language. To use a modern term, learning is languaging. This suggests the possibility of yet another new start.
New start 3?
Focus attention not on teaching learners to conform to possible, appropriate, performed in NS community, but rather give primacy to what is feasible in language learning process
Not acquisition of L2 competence assumed to be objective of teaching
Instead, develop more general lingual capability
- learning process vs teaching objective
- learn language, not teach a language
- extend learner experience, not discontinuity
- developing capability not competence conformity
Teaching is a matter not of denying but exploiting the learners’ experience and the process should continue after end of teaching – create momentum for further learning after teaching and beyond test. Certain objectives can’t be globally prescribed because how learners engage with a language depends on how they conceive of its foreignness. It is presupposed that all non L1 languages can be categorised as “foreign” depends on other factors like historical role (what symbolic or economic capital is associated with language). Russian is not foreign to Ukranian or Polish learners as it is to English speaker. Also how and why they are learned are likely to be different according to which languages are involved.
The popular textbook How languages are learned (Lightbown & Spada) makes the implicit claim that all languages are learned in the same way, and so taught in the same way. It is institutionally convenient and commercially profitable to make this assumption. But local contextualisation is more important according to Widdowson in spite of the vested interests of the ELT industry. Institutional resistance is to be expected, so this approach may be doomed to another kind of failure
New start #3
- local vs global perspective
- different perceptions of foreigness
Referring to the conference theme of valorizing practice in the form of grounded histories, Widdowson checks the definition of valorise (“to ascribe value or validity to an activity”) and asks whether the conference title refers to the activity of teaching or of learning? Since learning is mentioned first, perhaps it takes precedence? Little is known about the history of language learning, though this is the primary practice to be valorized. In the past, competence was measured by to what extent learners were good “teachees.” Learning has been undervalued and we should give priority to learning, and the development of lingual capability (not L2 competence). Widdowson judges this suggestion unlikely to be adopted but perhaps useful as a means of promoting critical thinking.