by Michael Swan

    Michael Swan is a writer specializing in English language teaching and reference materials. His publications include Grammar（in the Oxford introduction to Language Study series） and Practical English Usage(OUP). He is also co-author, with Catherine Walter, of the Cambridge English Course series, and of How English Works and The Good Grammar Book (OUP). His interests include pedagogic grammar, mother-tongue influence in second language acquisition, and the relationship between applied linguistic theory and classroom language-teaching practice. He has had extensive experience with adult learners, and has worked with teachers in many countries.

    The input to language learning: two out of three ain't good enough

Child language learning depends on:

1. extensive input: the flood of language in the environment.

2. concentrated input: repeated stories, nursery rhymes, the formulaic language of meals or bathtimes, etc. by unconsciously analyzing memorized material, children build up their knowledge of linguistic regularities.

Adults, but not children, also benefit from analysed input: grammar explanations, translations of vocabulary, ear-training, information about pragmatics, etc.

The language-teaching approaches of the last 50 or so years have rarely achieved a balanced focus on all three types of input. Concentrated input, in particular, is the Cinderella of current language-teaching theory, which has little room for things like dictation close text study, memorized dialogues, songs or learning by heart.

As many teachers instinctively feel, however, concentrated exposure to small samples of language is vital. The session will suggest some practical ways of integrating this element, so as to achieve a proper balance of input types.

    The ELT Journal / IATEFL debate 2006
    Sorry, but you've got to have a grammar syllabus

The fourth annual ELT Journal /IATEFL debate was held on Sunday 9th April at the IATEFL conference in Harrogate. The topic for debate was “Sorry, but you've got to have a grammar syllabus”. Keith Morrow, the editor of ELT Journal, chaired the debate and the two speakers were Michael Swan(for) and Scott Thornbury (against).

Over three hundred people came along, and after Michael and Scott had presented their views, it was time for the audience to have their say.

At the end, a vote was taken by inviting the audience to shout out‘aye' if they supported Michael Swan's view that you do need a grammar syllabus, or‘no’ if they supported Scott Thornbury's view that you don't. In the tradition of this debate, the chair was unable to decide who won–so please let us know what you think. Have a look at the summaries of the contributions from the two speakers, and give us your view on this interesting and controversial topic.

Michael Swan

The value of a grammar syllabus is self-evident. We can't teach everything, so we need to list high-priority items and make sure they are covered. A lower-level syllabus will naturally be progressive, dealing first with those elements which are essential for effective language use. At higher levels, syllabuses will be less ordered, focusing on points which are known to cause continuing difficulty. Grammar syllabuses should not dominate our teaching, but they are a vital part of our toolkit.

Critics of syllabus-informed instruction claim that teaching grammar points separately is ‘inappropriate' and ‘misguided’. Why? Pronunciation and vocabulary are most effectively taught if the most frequent, useful or problematic elements are listed and incorporated in teaching programmes. Why should this not be so for grammar? Music, skiing, surgery or piloting are commonly taught using progressive syllabuses. Why should the skills underlying language use be different?

Specific arguments against proactive syllabuses are either mistaken or difficult to interpret.

1. “A pre-planned syllabus is useless: you can't predict what grammatical problems your learners will have.”

2. “Learners acquire linguistic structures in a predetermined natural order; this progression cannot be affected by an imposed external syllabus.”
This is a highly controversial view for which empirical evidence is in short supply.

3. “The grammar of a language can only be acquired‘on-line' during communicative activity. Grammatical study decoupled from communication does not result in acquisition.”
There is no good evidence for this view, and massive evidence against it. Millions of people have learnt languages well by methods involving formal grammar study.

4. “Traditional methods involving grammar syllabuses have failed.”
No, they haven't — see above. Traditional methods haven't generally done very well, but nor have non-traditional methods. Languages are just hard to learn.

5. “Separating out discrete language points for attention subverts the holistic nature of language, and is in conflict with the natural conditions of language acquisition.”
You can't get the natural conditions of language acquisition(full-time exposure at an early age) in the language classroom. Our job as teachers is to compensate for the absence of these conditions, not to try insanely to recreate them. Especially not ‘holistically’ (teaching everything at once?).
Like other subtractive movements that try to stop us doing manifestly useful things, this proscription is counter-productive and makes little sense. We need grammar syllabuses.

Scott Thornbury

Ten reasons why you don't have to have a grammar syllabus:

1. Plenty of people (and not just children) have learnt a second language to a high degree of proficiency without ever having followed a grammar syllabus, eg, through simply “picking up” in naturalistic settings, or by following a content-based or immersion program—suggesting that at least for some people a grammar syllabus is not obligatory.

2. A grammar syllabus is an idealized representation, by linguists, of native speaker competence, and has only an accidental relation, if any, with the second language learner's own syllabus, that is, with the internal developmental path that language acquisition seems to take, irrespective of instruction or the learner's first language.

3. A grammar syllabus, especially in its traditional, canonical form, is a distorted and impoverished representation of competence, and bears little or no relation to the distribution and frequency of grammar items in naturally occurring language, nor does it distinguish between spoken and written grammar.

4. By prioritizing grammar, a grammar syllabus ignores the key-role of lexical and phrasal items in the development of the learner's internal representation of the language. Michael Hoey(2005), for example, notes that “Grammar is … the sum of the collocations, colligations and semantic associations of words like is, was, the, a and of, syllables like ing, er, and ly…” (p159)

5. … and by prioritizing grammar, a grammar syllabus ignores the predominance of lexical and phrasal items in language performance: corpus studies show, for example, that over 50% of written English and nearly 60% of spoken English consists of‘formulaic sequences’ of various types (cited in Schimitt and Carter, 2004).

6. By implying that language acquisition is rule-driven, a grammar syllabus ignores the fact that language is acquired through the piecemeal but massive exposure to instances of language in use:“Language learning is exemplar based… the knowledge underlying fluent use of language is not grammar in the sense of abstract rules of structure but a huge collection of memories of previously experienced utterances” (Nick Ellis,2002, p.166).

7. In practice, the grammar syllabus—even when combined with other syllabus ‘strands’(lexical, functional, etc) – tends to predominate at the expense of those other strands, since grammar is the most thoroughly described and organized aspect of language, and – crucially – the most easily tested.

8. By organizing language into teachable units (or grammar macnuggets) the grammar syllabus encourages a delivery-style, transactional-type model of teaching, at the expense of allowing language development to emerge organically out of the natural process of communication.

9. A grammar syllabus, by focusing on the forms of the target language, ignores the more productive potential of organizing learning around the meanings that the learner needs to engage with, as is the case with semantic syllabuses such as notional or task-based ones.

10. A grammar syllabus, in turn, requires that the materials that are designed to deliver it (in the form, for example, of course book texts) are chosen not because of their usefulness or relevance to the learner, but because they embed particular pre-selected and pre-graded grammar items, contributing to the generally anodyne, pasteurized nature of most ELT materials.

 

导读

    2006年4月9日《英语教学杂志》（ELT Journal）与国际外语教师协会（IATEFL）在英国Harrogate举行了第四届年度辩论会。辩论的主题是：“是否需要语法大纲”辩论的正方是著名语言学和语言教学专家Michael Swan，反方是Scott Thornbury。

    Michael Swan在辩论中肯定了语法大纲的必要性。他认为重要的语法项目应该优先考虑。低层次的大纲应该循序渐进，首先处理那些对有效使用语言不可或缺的项目。较高层次的大纲也应重视那些给不断给学生的学习和使用带来困难的语法点。他认为不能让语法大纲来主宰教学，但语法大纲是所有教学手段中的至关重要组成部分（a vital part）。

    在谈到对语法教学的批评时，他指出：在语音、词汇教学中把那些最常见、最有用、最容易出问题的地方列出来，综合处理，教学效果非常有效。他还说：教音乐、滑雪、做手术、驾驶飞机都需要循序渐进的教学大纲，为什么教语言技能就应该两样？

    随后他列举了反方的具体说法，指出这些看法要么是错误的，要么令人费解（either mistaken or difficult to interpret）。

    1． “学生遇到的语法问题无法预测”。

    Michael Swan认为：可以预测，因为这些问题有典型性，而且是长期存在、经常接触到。

    2．“语言习得有其自然顺序（natural order）不受从外部强加于学习者的影响”。

    Michael Swan：这种观点很有争议，实证不足。

    3． “语法只能从日常交际中习得，不能通过学习学得。”

    Michael Swan：这种说法无说服力（no good evidence），但反面证据却很多（massive evidence against it）。通过正规语法学习学好语言的多达millions。

    4．“传统语法大纲的教法已经失败”。

    Michael Swan：不，没有失败。传统法的效果不很好，非传统法效果同样不佳，因为语言太难学。

    5．“分散、孤立地教语言破坏了语言的整体性，与语言习得的自然条件冲突。”

    Michael Swan：课堂教学中无法获得语言习得的自然环境。教师的职责是去补偿语言环境的不足，而不是愚蠢地再造环境。

    Michael Swan最后指出：阻止、剥夺人们去做那些显而易见的有益的事情的权利具有破坏性（counter-productive），而且没有道理（makes little sense）。他的结论是：We need grammar syllabuses。

    Scott Thornbury列举了十条理由，反对语法大纲。

    1．语言是靠自然习得的，语法大纲至少对一些人是不必要的。

    2．语法大纲理想化了，与语言的实际使用和语言习得的内在路径不一致，语法教学作用不大。

    3．语法大纲，尤其是传统语法大纲，对语言能力的表述是扭曲、无力的，与实际使用中语法项目的分布和频率缺乏关联。

    4．语法大纲突出语法的作用，忽视了在自然语言中出现频率很高的词语固定搭配（lexical and phrasal items）在语言习得和使用中的关键作用。

    5．语法大纲突出语法的作用，忽视了词语规定搭配在语言使用中的支配地位。语料库研究证明：50%的书面语和60%的口语是由各种类型的固定结构（formulaic sequences）构成的。

    6．语法大纲暗示语言习得受规律制约，语法大纲忽视了语言是通过零星而大量的接触实际使用中的语言范例而获得的。语法只能在使用中学习、掌握，语言学习是“以范例为基础的”（exemplar-based），是在使用中根据对大量语言实践的记忆感悟出来的，而不是靠抽象的语法规则和结构学会的。

    7．在实际使用中，即使语法大纲和词汇、功能大纲（lexical, functional, etc）结合，也往往独占一统，因为语法最容易表述、安排，也最好测试。

    8．语法大纲把语言分割成容易讲授的条条块块，鼓励注入性教学，影响语言在自然的交际过程中健康发展。

    9．语法大纲重视语言的形式，忽视围绕语言意义组织语言学习（如意义大纲和任务型教学理念）。

    10．语法大纲为了教学方便，在选材上不是根据语言对学习者的用途和相关性，而是因为所选的材料中包含事先选定并分成等级的语法项目。

    仔细研究正方和反方的观点，对于在外语教学中是否需要有语法大纲？我们会有自己的结论。

    Michael Swan在本文的开头就指出：语言学习仅有两种输入（intensive input和concentrated input）是不够的，还应该有analysed input。他指出：Adults, but not children, also benefit from analysed input: grammar explanations, translations of vocabulary, ear-training, information about pragmatics, etc.

    虽然他认为这种analysed input是用于成人，而不是儿童的。然而不可否认的事实是：在中国语言文化环境中成长起来的我国的中学生在心理和智力上已经不是幼儿，而是正趋向成人，或者接近成人。目前我国英语教学中的一种不良倾向，正是把学生“小儿化”，不是积极开发他们的智力，而是把他们当成幼儿，让他们在模拟的英语环境中通过感悟、体验来摸索。这样做的效果究竟是“多、快、好、省”，还是“少、慢、差、费”？实践已经、并将继续做出证明。

    Michael Swan所说的analysed input包括语法讲解（grammar explanations）、词语翻译（ translations of vocabulary）、听力训练（ear-training）、语用知识（information about pragmatics）等等。这在英语二语教学中是必要的，在英语外语教学中更是必不可少。

    诚然，我们目前使用的中学英语教材大多是按照“话题、功能、结构、词汇”这一体系，而不是按照语法大纲编写的。对此教师无法改变，但又不是无能为力。   

    虽然教材不是按照语法大纲编写的，但教师完全可以有自己心中的语法大纲。也就是说：教师应该对教材中涉及到的语法项目有个总体安排。有些项目，正如Michael Swan所说的，如不讲解，会影响学生正确理解和使用语言，该明示就明示，先讲规则，再举例子；有些项目，（如冠词的用法、动词的时态、主从复合句等）可以先引导学生注意观察，然后归纳总结。