For instance in her chapter, “The Educated Philistine: the revolt against grammar”, Phillips accurately quotes me in Grammar! (CILT 1994) saying that “the prime goal of teaching French is the communication of meaning”, but then perversely interprets my argument as meaning that teachers should go to some lengths to avoid teaching their pupils any ”‘unnecessary grammar’ - unnecessary being defined, it would appear, as anything which might require systematic learning”.

Nowhere in this chapter do I use the word “unnecessary”. Not only do I believe that systematic learning and practice of language structures is required, but I illustrate my belief both in the chapter from which the quote is taken and earlier in the book.

She similarly makes up another belief which I have never held. Apropos of my strong proposal to teach phrases such as “je voudrais” or “je suis allee” initially as vocabulary, she adds that I then suggest that “what teachers should not bother to do until much later, is teach pupils the conditional or perfect verb forms along with the correct word endings”, with the result that “such pupils are quite unable to work out for themselves what ‘we would like’ or ‘they have gone’ might be in French.”

In fact I make no such suggestion in the book. I do suggest that teachers create for learners the opportunity to work out for themselves how to say anything they need to say. But my approach to grammar not only enables them to understand “why endings change”, which Melanie Phillips says my suggestion precludes, but also enables learners to explain why these changes take place.

Grammar may be deemed as “a description of a language and its constituent elements”. As such, it is something learners need. For example, they need the description if, having practised and used the past tense of verbs conjugated in French with etre, it helps them to see a systematic representation of the verb category they have successfully used to talk about past events. It is clearly impossible in any one text to make reference to descriptions of all the grammatical features contained within it; learners and teachers therefore make judgments about what has high communicative value in a particular context and select what shall be described and what shall be left until later.

Grammatical description can also serve if learners need to check linguistic accuracy, which is important if inaccuracies are likely to prevent communication, lead to ambiguity or cause offence. However, grammar description, when encountered out of context and before any practice and use by the learner, seems unlikely to help communication. By the time the learner has worked from description to use, fluency has been sacrificed and communication normally breaks down.

The word “grammar” may also be used to mean the terminology in which that description is formulated. Again this can serve a useful purpose. It can provide a commonly understood vocabulary which enables teacher and learner to talk about language, and clarify the language patterns and systems they have put into practice.

The process develops through cycles of practice, use, explanation, re-use andor extended use depending on the level of understanding the learners have and need. In class learners may need to identify verbs and nouns in the language they are learning but few need to know at first that le mien, le tien, le sien are possessive disjunctive pronouns. The use of grammatical terminology grows with grammatical understanding; it does not necessarily, of itself and out of context, lead to understanding and subsequent competent language use.

Grammar commonly also refers to grammar rules, the rules to which the language operates, and to the patterns created by those rules. Knowing a rule of grammar can also help learners, provided they are familiar with the terminology and have used the language to which it refers. Try to express a thought by working from the rule to the communication, and fluency crumbles.

This is not to say that grammar rules cannot be used to generate language: they can. At issue is what kind of language use we envisage. If we are aiming at fluency in a given situation then the emphasis is best placed on communicating the meaning, rather than being too concerned during the process with the form of the message. If we are aiming at presenting and practising language patterns we can concentrate on learners operating within prescribed grammatical rules.

Much discussion about learners’ ability or inability to apply grammar rules confuses these two aims. It seems that the two can co-exist happily and indeed need to complement each other. Communicating meaning automatically involves the implementation of some grammatical rule, albeit with error, necessary for learning to take place. In school, practising the application of grammatical rules in contexts which continue to convey meaning and are not entirely unreal exercises, can enhance understanding and, as far as we know and if put into use, lead to greater communicative competence.

Systematic practice of patterns in both speech and writing, and freedom to experiment in order to convey a message in either medium should be offered to all learners; but since learners differ and since learning goes through different stages we cannot automatically and without thought offer the same mixture to all.

Melanie Phillips, All Must Have Prizes. Little, Brown, 1996 Grammar! edited by King L, Boaks P. Centre for Information on Language Teaching and Research, 1994.

Barry Jones is a principal lecturer at Homerton College, Cambridge