View context for this page View table of contents for this book View table of contents for LinguaLinksLibrary Go to LinguaLinks home page

1.3. Principle II: Engage in extensive extemporaneous speaking.


In my own experience, Krashen's input hypothesis has been enormously helpful. Yet it appears that few scholars agree with the hypothesis in its entirety. That is because Krashen doesn't just claim that comprehensible input is the most important factor in second language acquisition. He claims that it is the only factor!

1.3.1 Comprehensible input is not enough
1.3.2 You can't speak well unless you can speak poorly.

1.3.1 Comprehensible input is not enough

Merrill Swain (1985) examined the French ability of children who had been in a school immersion program for seven years. These children, who were from English speaking homes, had received all of their elementary education in French. Yet after seven years of receiving truly massive comprehensible input in French, they still did not control the French language like native French speaking children did. Why not? Probably there were a number of problems, but an obvious one was that the students didn't have much opportunity to speak the language. They mainly listened to the teacher. When they spoke to one another informally, they used English. When they spoke to the teacher they used French. But then, how much class time is devoted to any one student speaking to the teacher? And the only children they ever heard speaking French were their class-mates, non-native speakers like themselves, and that only happened when their classmates were addressing the teacher. You certainly couldn't say that the children used French in a very rich variety of life situations, or that they used French for a very wide variety of communication purposes. It appears that, since they didn't speak French very much, their speaking ability did not develop as well as we might have hoped. Even their ability to understand French appears to have suffered from the fact that they did not speak it very much. That's not to say they didn't learn French quite fluently. But Swain conjectures that they might have done better if they had been speaking French extensively, in addition to all their years of listening to comprehensible input.

You may know children of immigrant parents who can understand their parents' language quite well, but cannot speak it at all. Nancy Dorian (1981) noticed that although she had learned to speak Gaelic in the course of her research, young people with Gaelic speaking parents, although they could not speak the language at all, could often understand it better than she could. They had grown up with massive comprehensible input, and had developed a high degree of comprehension ability, but little or no speaking ability.

So it appears that massive comprehensible input can result in people having the ability to understand a language without necessarily being able to speak it well, or even to speak it at all. It appears that in order to learn to speak, you have to put a certain amount of effort into speaking. Somehow, I don't find that surprising.

You might ask whether it is possible to learn to speak without receiving much comprehensible input. Some linguists have told me that their speaking ability exceeded their comprehension ability, at least for a long time. Recently one told me that he could plan and execute a very complicated sentence in a certain African language, but that if he heard the same sentence in natural speech he would have difficulty understanding it. My experience in learning Blackfoot was similar. For a long time my speaking ability exceeded my comprehension ability. That is not a very good way to learn a language, for a number of reasons. For one thing, if you can say a lot more than you can understand, people will misjudge your general level of ability in the language, and speak to you in such a way that much of what they say will go over your head. This can make conversational experiences embarrassing and stressful, and discourage you from spending a lot of time conversing with people. So you want to keep the horse of comprehension ahead of the cart of speaking, while bearing in mind that comprehensible input in and of itself is not enough. You also need to speak.

In particular, it appears that you will need to put a reasonable amount of effort into conversational speech. Sure, you could concentrate on monologues, say by making long speeches to large audiences, but you might not know whether anyone understood you. By contrast, when you engage in one-on-one conversational interaction with people, it will often become obvious that you have failed to communicate, or have miscommunicated. What is more, the people with whom you converse will be able to help you to find ways to say what you are trying to say. It seems reasonable to think that this would contribute to your language learning.

Conversational interaction is an important source of comprehensible input as well. When you are involved in conversing with people, they will tend to adjust their speech to your level of ability. They may speak more slowly than normal, and use simpler vocabulary and simpler sentence structures, and repeat themselves a lot, and reword their sentences whenever you appear to be having trouble understanding them. Listen to yourself the next time you are talking to someone with limited English ability. You will probably find that you make these types of changes in your own speech in order to help them to understand you. Michael Long has demonstrated that the types of changes people make in their speech when they talk to foreigners really do make a significant difference in the ability of the foreigners to understand them (Long 1985).

Consider the nature of a conversation between a sympathetic native speaker and you as a language learner. You have a meaning you wish to express. You make a stab at it, but the person with whom you are conversing is either unsure of what you meant, or wonders whether you really meant what you appear to have meant. So she helps you to clarify yourself. Likewise, when she says something to you, it may go over your head, and so you get her to clarify her meaning. This back and forth process of a language learner and a sympathetic native speaker working together to achieve success in conversational communication is referred to as the negotiation of meaning.

During Stage II, the most convenient context for conversational practice may be in structured language learning sessions, where someone is consciously helping you to learn the language. If you have one or more persons who are employed to help you on a daily basis, those people will be accustomed to speaking with you. They will have a good feel for your current level of ability, and thus will be in a good position to make sure that what they say to you is comprehensible. There is a low stress level involved in conversing with a familiar person in a familiar setting, when compared with conversing with all the people you happen to encounter out in the world at large.

In addition to formal language sessions, you will increasingly be able to engage in conversation with friends. For them, conversing with you is hard work at this point, so it requires some commitment on the part of your conversational partners. But again, people who know you well will be able to communicate with you far more effectively and easily than people who do not yet know you. With time, you can systematically expand the number of regular conversational partners with whom you visit (see Thomson, 1993c). So once you are past the very early stages of language learning, an obvious way to increase your comprehensible input is to engage in a lot of social visiting. You may not be a person who normally does a lot of social visiting. If so, it will help if you can view social visiting as part of your daily work routine.

Not all of your early speaking efforts need to be in the form of two-way conversation in the strictest sense. As a matter of fact, while you are first trying to loosen up your tongue and develop some fluency in the language, you will benefit a great deal from activities in which you do most of the talking. These activities are probably best carried out in formal language sessions, where you are employing someone who understands that she is there for the purpose of helping you learn the language. I'll have more to say about these activities below. Once you gain a degree of fluency through such structured activities you will be increasingly comfortable with unstructured social visiting as a means of getting conversational practice on a grand scale. You can use your formal language sessions as a means of preparing for your general social visiting. For example, when you learn to discuss some topic in your language sessions, you can then make a point of discussing that same topic during informal social visits. You can even tell your friends, “This is what I have been learning to talk about with so-and-so”, and then go on to talk about the topic with your friends.

To sum up, Principle II is another way of saying that you learn to talk by talking. You might say that you learn how to talk by being exposed to massive comprehensible input, but ultimately you only learn to talk if you talk.

Given what we have said about Principle I, and Principle II, we might consider the following formula to come close to the truth:

Assuming you have a strategy for getting comprehensible input, and for getting conversational practice, the path to powerful language learning could hardly be more simple.

1.3.2 You can't speak well unless you can speak poorly.

Now you may be thinking that I'm ignoring your main concern. You feel that no matter how you struggle, you are unable to get the grammar right. If you have been learning the language through a formal language course, mastering the grammar may seem to be the central challenge. Perhaps you even got low marks because of all your errors of grammar. Well, I have good news for you. Errors are great! From here on in, you get high marks for errors, at least in my book, and hopefully, in your own book, too. If you're not making errors, you're not breaking new ground. The pathway to accurate speech is through error-filled speech. I therefore suggest that you move your concern for grammatical accuracy away from center stage. Concentrate on getting comprehensible input and conversation practice, and watch your grammatical accuracy improve without your even focusing on it. I will later suggest ways that you can focus on grammar, as well, but that will be more with a view to mopping up persistent problem areas.

When I was in High School, a language learning method came into vogue which was based on the belief that from the very outset students should speak the language perfectly. My high school French teacher responded to a student's complaint with the comment “I didn't write the textbook, but if I had I'd be a millionairess.” Such was the enthusiasm of many teachers for the new method. That enthusiasm was followed by disappointment, when it turned out that few students developed much ability to use the language extemporaneously for real communication.

Have you ever observed a real person learning English as his or her second language? If you have observed such a person over an extended period, you will have noticed that s/he began by speaking English very poorly, and gradually improved until, hopefully, s/he came to speak English well. It always works like that in real life. Granted some people do better than others both during the early weeks, and in terms of their overall rate of progress, and ultimate attainment, but nobody starts out speaking perfectly. Developing good speaking ability is always a gradual process. I can't understand why my high school French teacher and others like her hadn't noticed that.

When you are first learning a new language, your personal version of the language is very different from the version used by the native speakers. Let's suppose you are learning Chukchee, and your native language is English. The new “language” that you speak, say, after a couple of months, is sure not English. But is it Chukchee? It doesn't appear to be Chukchee in the strictest sense. However, it is obviously derived from Chukchee, and not from English. Six months later you will be speaking another “language”, which is much more like Chukchee in the strictest sense than the “language” you speak after two months. After a couple of years, the language you speak may be enough like that of native speakers that you can justifiably call it Chukchee. However the “language” you spoke after two months, and the one you spoke after six months, were quite different from Chukchee in the strictest sense. What were those “languages”? Chukchee speakers could understand you when you spoke to them, and you could understand a lot of what they said when they spoke Chukchee. What you spoke was a real language (despite all of my scare quotes). More precisely, it was a series of languages, each one more like real Chukchee than the last. You invented these languages as you went along, on the basis of the Chukchee you heard. I have to say you invented these languages, because they were unique to you. You didn't hear anyone else talking like that, so you can't really say that you learned them. No. You invented them, using as your source of building blocks all of the comprehensible input you were exposed to.

You may prefer to think that you didn't invent anything. Rather, you may say, you only learned something. You learned Chukchee, only you learned it poorly at first. But if we may return to the example of someone learning English we'll see that there is quite a bit of inventing going on. Wode (1981) examined the forms of negative sentences used by people learning English. Learners first learned to use the word “no” in response to questions or statements. Then they started adding it to sentences, so that if they wished to say that someone had not finished doing something, they might say “No finish.” Later they would use the word “no” in slightly fuller sentences, as in “That's no good”, meaning what you would mean by “That's not good.” Later they would learn to add a form of the helping verb “do”, and say something like “You didn't can throw it” (all of these examples are cited in Cook, 1991, p. 19 ). I think it is fair to say that sentences like “No finish,” and “You didn't can throw it,” come from an invented language. They are not simply copied from normal English. Rather the speakers know bits of English, and use those bits to invent their own language. These invented languages that are derived from the language being learned, and which gradually become more and more similar to the language being learned, have been called interlanguage (see especially Selinker, 1992).

The existence of interlanguages is one of the main reasons we know that brains know how to learn languages. The interlanguages of people learning a given language, let's say, learning English, go through similar stages, regardless of their mother tongue. For example most people go through this same sequence of patterns in learning to form negative sentences in English. Why do different people's interlanguages go through the same stages while learning English negation? The answer is that when it comes to learning a language, your brain has a mind of its own. It will invent the interlanguages, and refine them, until it has succeeded in reinventing the language as it is spoken by natives, or at least some reasonable facsimile.

I say all of this to reassure you that if you keep exposing yourself to comprehensible input, and keep persisting in conversational practice, your speech will keep getting better. Some perfectionistic people don't like this. They would prefer to speak perfectly, or not at all. Well, if you are such a person, swallow your pride. Speak badly. The way to come to be able to speak well is to speak badly for an extended period of time.

So then, speaking the language imperfectly is essential. There is a whole body of research on how people manage to cope while they are still not very good at using their new language. They use a variety of strategies in order to communicate, strategies which have been labeled, appropriately enough, communication strategies.

There have been a number of efforts made at classifying the strategies people use when communicating in a second language (these are surveyed in Bialystok, 1990 ). One well-known system of classification makes a distinction between reduction strategies and achievement strategies (Faerch and Kasper, 1983b, 1984), summarized in Ellis 1986). When you use a reduction strategy, you may simply avoid trying to say something that you would like to say, because you can't think of any way to get your point across. Or you may find a way to say something which is related to what you wanted to say, but not really the same. For example, you may wish to say that you are worried about something, but realizing you don't know how to say that, you may resort to simply saying, “I don't like it.”

In using an achievement strategy, you will find a way to express what you wish to express, even though you don't know the normal way to express it. For example, you may not know the word for a crank on a machine, and so you say “this thing” while making a circular motion with your hand. Some people are probably better than others when it comes to using communications strategies. I mention them here to reinforce the point that it is normal to speak “poorly” first, and gradually improve. That is the name of the game. If you put high demands on yourself for premature excellence, it will discourage you from speaking as much as you need to, and thus hinder your progress. So get out there and start making mistakes. And give yourself extra credit for extra mistakes.

Context for this page:

Go to SIL home page This page is an extract from the LinguaLinks Library, Version 3.5, published on CD-ROM by SIL International, 1999. [Ordering information.]

Page content last modified: 11 September 1997

© 1999 SIL International