Double speak

What's the best way to teach a second language? Alan Thomson reports on a debate that has occupied academics for over 40 years

How people learn a second language is the subject of a 40-year-old debate that has done little to help Britons tackle their linguistic ignorance. Academics are deeply divided about how to teach students to speak a foreign language. Some argue that people pick up a language best by being immersed in it. If you want to learn Spanish, live in Spain, listen to nothing but Spanish and you will soon be speaking fluently. Others maintain that teaching the nuts and bolts, and the rules of any language, is vital.

In practice university courses usually mix the two approaches, sending students for a year abroad while also giving them three years' grounding in grammatical niceties. But the past two to three decades have seen the emphasis on the old-fashioned teaching of rules take a back seat.

The debate started in the late 1950s when Noam Chomsky, the influential philosopher of linguistics, argued that humans were innately equipped to create and understand language. He claimed that we are born with a rule book, or "universal grammar" - a sort of template that underpinned all languages. Certain formal similarities, said Chomsky, were common to all languages. His theories gained many adherents, perhaps as a reaction to the educational orthodoxy that insisted that learning a language meant grammar, tenses and learning by rote.

The Chomskian approach encouraged linguists to favour "experiential learning" - imparting a second language by surrounding the learner with speakers of that language rather than by stifling natural communicative abilities with grammatical rules.

This is how young children learn their native language. Children born to bilingual parents have no trouble learning two languages simultaneously despite the fact that they are not taught grammar and syntax explicitly by their parents. They just imitate the words they hear around them and somehow assimilate this into a working knowledge of language. Chomsky argued that children carry an innate ability to form new sentences, sentences they have never heard, into adulthood.

But recently there has been a backlash. Language lecturers complain that while their first-year students can chatter happily in a foreign language, they care little about sentence construction or proper use of tense and grammar. Disgruntled academics are calling for a different approach.

Keith Johnson, senior lecturer in linguistics and English language teaching at Lancaster University, says: "Having had a generation who were taught languages without much grammar, we are seeing the results in students who cannot put a sentence together. The pendulum is swinging back."

Derek McCullough, senior lecturer in German at Surrey University, agrees:

"I am a grammarian, and I do not think that a lot of language teaching is bearing fruit. How a child acquires language at two or three is not how they learn it at 18. In most school situations language immersion is irrelevant. How can a pupil be immersed in French in a secondary school in Slough?"

Even strong Chomskians see the need for revision, but they remain opposed to old-fashioned ideas of drumming grammar into people's heads. Antonella Sorace, reader in applied linguistics at Edinburgh University, says: "I believe that there is such a thing as universal grammar. It is virtually impossible to prove empirically but there is evidence pointing to its existence. All children end up with a knowledge of language more complex than the words they have heard spoken.

"Fifteen years ago the tendency was to throw the grammar books out the window. Now there is a revival of older methods. Perhaps the best way would be to teach the grammar in the language so you get double the benefit."

Finding the middle ground is the key. But the debate continues on where to put the emphasis. Chomskians such as Dr Sorace say that part of the learning conundrum would be answered if more were known about how universal grammar interacts with adult comprehension. The problem, says Dr Sorace, is that two different types of knowledge are at work when we learn a second language - our conscious understanding of grammatical rules and our innate understanding of language construction. "I would not say we know very much about how these interact but there is evidence that knowledge acquired through studying grammar is not very useful when using the language."

"Children learning their first language rely only on universal grammar. But I do not believe adults learn a language in the same way. They can rely on universal grammar only to a certain extent but much is due to their cognitive powers."

Richard Towell, head of modern languages at Salford University, says there is "no authoritative view" about the best way to teach languages. "I think there are elements of both types of learning, although Chomsky has been over-influential in schools of linguistics over the last 30 years. The difficulty is that we take language for granted and when we look at it, it begins to fall apart."

Genes may be the key to learning a second language, according to German lecturer Derek McCullough, who bases his argument on observations of how successfully people in foreign countries learn English. Scandanavian children, for example, are particularly articulate, while children in Latin countries are much less fluent in foreign languages, says McCullough. The British are poor linguists.

"The French are as bad as us. Whereas even relatively indolent German school children can construct a meaningful sentence in English. I am sure that someone may prove that hereditary factors play a part," he says.

How you learn to play a musical instrument may shed light on the mechanics of language learning, says linguistics lecturer Keith Johnson.

"According to the Chomskian school of thought - that learners need to be surrounded by speakers of a language if they are to learn it properly - a child learning to play the violin would not have to practise scales. But we know that practice is very important in learning to play an instrument, although it is clear one would never learn to play the violin just by practising scales," he says.

"Learning to swim offers another parallel. Following the Chomskian argument, the way to teach people to swim is to push them in at the deep end. Some will start swimming, but is it the most efficient way to learn? If you send someone to Germany for six years they will learn the language but is it the most cost-effective way of teaching?"