Language learning

It is one of publicly available versions of explorations in Language Acquisitionof Stephen D.Krashen dated by 2003. Since it’s only available in PDF format, without the option to convert it to text, I dictated text using the Voice to Text app for the convenience of word-for-word translation in the browser for my grandson.
There are typos in the text and I completely missed some of the content, so it’s better to read the original, of course.
src: https://www.researchgate.net/publication/349255011_Explorations_in_Language_Acquisition_and_Use

Explorations in Language Acquisition and Use

The TAIPEI LECTURES Stephen D.Krashen

Contents

1 Principles of Language Acquisition

• Part 1 : Theory
• Part 2: Applications

1. Free voluntary reading: still are very good idea.
   • Research on free reading
   • Light reading as a bridge
   • Other advantages of Reading
   • Motivating students to read
   • Some innovation
   • Conclusion
   • Notes
2. Current issues and controversies: does Grammar teaching work? What about comprehensible output?
   • The role of grammar
   • comprehensible output?
   • notes

4.How reading and writing make you smarter or how smart people read and write -How we get smart -Some notes and incubation -The research on learning by problem solving -Reading and cognitive development -writing and cognitive development -oral language -Conclusion

Introduction

On November 17 and 18 2001, I presented a series of four lectures at national Taipei university in Taiwan. I attempted to review the fundamentals of second-language acquisition theory, present some of the original research supporting the theory as well as more recent studies, present counterarguments to criticisms and explore some new areas that appear to have promise for progress in both theory and application. This volume is based on these lectures chapters 1,2 and 4 correspond closely to the to three of the four lectures presented. Chapter 3 includes material presented in the four lectures but also includes material discussed in response to questions from the audience as well as a discussion of recent paper by Norris and Ortega dealing with the effect of grammar instruction.

Chapter 1 reviews the central hypothesis underlining what I consider to be current theory in the language acquisition. This hypothesis have not only survived well over the years but have also proven to be useful in other areas language education. So far research results remain consistent with hypothesis and there is no counterevidence. According to the rules of science this is all one can demand of a hypothesis. But the fact that where hypothesis have also helped explain phenomena in the other other areas in is equally impressive. The clearest example is the role of input hypothesis ( also known as their comprehension hypothesis). Has explored in chapter 2, the input hypothesis has been successfully applied in the area of reading a comprehensible input in the form of free voluntary reading has been shown to be highly effective for first and second language development I have argued in other publication that comprehensible input also helps explain the success of of whole language methodology in the beginning reading( Russian 1990 99) as well as the success of well designed bilingual education programs( gration 1996).

Chapter 3 is self-defense. The research community has devoted an extraordinary amount of energy in an attempt to show that grammar teaching works instead, they have shown only what many many language students have always realized: formal grammar instruction has a very limited impact on the second language competence. Even intensive prolonged instruction that is limited to just a few aspects of grammar results, in general, in only modest games on tests in which students students are encouraged to think about form. The researcher themselves in every case consider their results to strongly support the efficiency of grammar instructions. I argue in chapter 3 that results only show that the monitor hypothesis, reviewed in chapter 1, is correct.

Chapter 3 also contains a discussion of a current rival to the input hypothesis, talk comprehensible output hypothesis. Its originator,, did not consider it to be rival but supplement to comprehensible input. It much of current practice assumes the correctness of comprehensible output and considers it to be the major path the second language competence. The data, in my view, certainly does not support comprehensible output as the only way; in fact, there’s a little evidence that it plays any role at all. Parts of chapter 3 were originally published in foreign language annals and system.

Chapter 4 is an exploration into other areas. Good readers and writers, I argue, are those who have learned to read and write in a way that is consistent with way the brain learns and solves problems. Unfortunately, the most efficient way ah of using reading and writing are often different from the way we are taught in school. Good thing is I conclude those who have overcome the lessons they have learned in school. A previous version of this paper was originally presented at the Georgetown roundtable on languages and linguistics and was published in the proceedings.

It has been updated and, I hope, improved by the addition of recent work on the composing process, especially the interesting work of Robert Boyce. Voices inside have, in fact, been of great help to me in completing this manuscript. I highly recommend his 1994 book how writer journey to comfort and fluency.

I think my former student and now will you colleague professor sign Lee of national Taipei university, who organized the series of presentation in Taiwan. I also thank the chair of department of foreign languages and applied linguistics international Taipei university, Liu, for his hospitality.

Principle of Language Acquisition

Part 1: Theory

The following five hypothesis are the core of current theory on language acquisition. They are presented here as a summary without supporting evidence. For much more detail, see Krashen (1981, 1982, 1985, 1994a) Krashen and Terell (1983).

The acquisition-learning hypothesis

The acquisition learning hypothesis claims that we have two independent ways of developing language ability: acquisition and learning. Language acquisition is a subconscious process; while it is happening, we are not aware that it is happening. Also, once we have acquired something, we are not usually aware that we possess any new knowledge; the knowledge is stored in our brain subconsciously. The research strongly supports the view that both children and adults can subconsciously acquire language. Also, both oral and written language can be acquired. In nontechnical language, acquisition is sometimes referred to as “picking up” a language. When someone says “I was in France for a while and I picked up some French”, it means he or she acquired some French.

Language learning is what we did in school. It is conscious process; When we are learning we know we are learning. Also, learned knowledge is represented consciously in the brain. In not technical language, when we talk about rules and grammar we are usually talking about learning. Error correction helps learning. When we make a mistake and someone corrects us, we are supposed to change our conscious version of the rule. If a learner says, “I comes to school every day” and a teacher responds with, “Not it’s “I come to school”, the learner is supposed to realize that the -s doesn’t go on the first person singular. As we shall see, error correction and conscious learning are very limited.

The natural order hypothesis

The natural order hypothesis claims that require the part of language in a predictable order. Some grammatical items, for example, tend to be acquired early, while others come later. The order of acquisition for the first and second languages is similar but not identical it has been established, for example, that ink -ing marker in English, the progressive, is acquired fairly early in the first language acquisition, while the third person singular -s is acquired later. The third person singular may arrive 6 months to year after -ing. In adult second-language acquisition, the progressive is also acquired early but the third person singular may never come. It’s a common to hear people who speak English as a Second Language very well and yet have not acquired the third person singular. Not every acquirer proceeds in exactly the same order, but the variation among acquirers is not extreme. There clearly is an “average” order of acquisition.

There are three amazing facts about the natural order phenomenon, and I will list them in the order of amazingness:

1. The natural order is not based on any obvious features of simplicity and complexity. Some rules that look simple (e.g. the third person singular) are acquired late. Other that appear to linguists to be complex a acquired early. This presents a problem to curriculum designers who present rules to language students from “simple” to “complex”. A rule may seem to be simple to a linguist, but may be late acquired.

2. The natural order cannot be changed. Iit’s immune to deliberate teaching. We cannot alter the natural order by explanations, drills, and exercises. A teacher can drill the third person singular for weeks but it will not be acquired until the acquirer is ready for it. This explain a great deal of frustration language students have.

3. One might suppose that the solution into our problems is simply to teach along a natural order. We need only find out which items are naturally acquired early and teach those first, etc. The thought amazing fact is that this is not the solution: The natural order is not the teaching order. I will explain why later.

The monitor hypothesis

The monitor hypothesis attempts to explain how acquisition and learning are used. Language is normally produced using our acquired linguistic competence. Conscious learning has only one function: as a “Monitor” or editor.

Here is how it works: We are about to say something in another language.The form of our sentences pops into our mind, thanks to our subconsciously acquired competence. Then, just before we actually produce, the sentence just before we say it, we scan it intentionally, inspect it, and use our consciously learned system to correct errors.

We can also use our conscious Monitor to correct sentence after we have produced them; this is called self-correction (Of course we also self-correct, or edit, using our acquired system, or our feel for correctness. The monitor hypothesis reclaims that conscious learning responded this function; it does not contribute to our fluency.) While the Monitor can make a small contribution to accuracy the research indicates that acquisition makes the major contribution. Thus, acquisition is responsible for buff fluency and most of our accuracy.

It’s difficult to use the Monitor. In order to use the Monitor successfully, three conditions are necessary:

1. The acquirer must know the rule. This is very difficult condition to meet. Research linguists freely admit that they don’t know all the rules of any language. Those who write grammar texts know fewer rules than the linguist. Language teacher do not teach all the rules in the texts. Even the best students don’t learn all the rule or rules that are taught, even the best students don’t remember all the rules they have learned and even the best students cannot always use the rule they do remember. Many rules are too complex to apply while engaging in conversation.

2. The acquirer must be thinking about correctness, or focus on form. This is hard to do. It is hard to think about buffer form and meaning at the same time.

3. They acquirer must have time. For most people, normal conversation doesn’t provide enough time for the use of the Monitor. A few language experts can Monitor while conversion, but these are very advanced acquirers who only need to Monitor and occasional rule here and there, and who have a special interest in their structure of language.

Research shows that the Monitor use is only the obvious when all the three conditions fully met. For most people, this occurs only when we give them a grammar tests exclaim

The monitor is weak but it is not useless. Some conscious knowledge of language can be helpful. Acquisition doesn’t, typically, provide us with 100% of a language, there is often a small residue of grammar, punctuation and spelling rules that even native speakers do not acquire, even after extensive aural and written comprehensible input. In English, these can include why lie/lay distinction, the its/it’s a distinction, and spelling demons such as separate and commitment (how many -s). Because our standard for written communication is 100%, this aspect of language need to be learned, but they make up a small part of our language competence.

We pay a price for the modest amount of accuracy we get from monitoring. Some research shows that when we focus on form when speaking, we produce less information and we slow down (Hulstijn ans Huilstijn 1984). This can seriously disrupt conversation some people over monitor and are so concerned with grammar and accuracy that speech is slow and painful to produce as well as to listen to:

The major, who has been the great fencer, did not believe in bravery, and spent much time while we sat in the machines correcting my grammar. He had a compliment me on how I spoke Italian, and we talk together very easily. One day I had said that Italian seemed like such an easy language to me that I could not take a great interest in it; everything was so easy to say.”Ah yes” the major said, “Why then, do you not take up the use of grammar?” So we took up the use of grammar, and soon Italian was such a difficult language that I was afraid to talk to him until I had the grammar straight in my mind. (E.Hemingway, Men Without Women, 1997 [1927], 46-47])

The best advice is, I think, to use the conscious monitor when it does not interfere with communication, when we have time, as in the editing phase of writing.

The input (comprehension) hypothesis

The input hypothesis attempts to answer the most important question in the field of the language acquisition and language education: How does language acquisition occur? The evidence strongly support a simple hypothesis. We acquire language in only one way: when we understand messages; That is, when we obtain “a comprehensible input”. We acquire language, in other words, when we understand what we hear or what we read, when we understand the message.

In the recent years, I have used the term comprehension hypothesis to refer to the input hypothesis. Comprehension is a better description — mere input is not enough; it must be understood. This term also allows me to honor the lineage of the input/comprehension hypothesis; The idea is certainly not new with me. In the field of second language acquisition, James Asher, Harris Winitz, and Robins Berling proposed similar ideas year before I did, and in the field of literacy, Frank Smith and Kenneth Goodman had proposed that we learn to read by reading by understanding the message on the page.

Comprehensible input has been our last resort in language teaching: we have tried everything else— grammar rules, repetition drills, computers, and so on. They input hypothesis claims, however, that comprehending messages in the only way language is acquired. There is no individual variation in the fundamental process of language acquisition.

The input/hypothesis claims can be restated in terms of natural order hypothesis. Let’s assume a very simple version of a natural order hypothesis — that require the rules of language in the linear order: 1, 2, 3… The question of the how we acquire language can be restated as: how do we move from rule 3 to rule 4, from rule 987 to rule 988? More generally, if ‘i’ represents the last rule we have acquired, how do we move from i to i +1, where i + 1 is the next structure we are ready to acquire?

The input hypothesis claims that we move from i to i+1 by understanding input containing i + 1. We are able to do this with the help of our previously acquired linguistic competence, as well as our extra linguistic knowledge, which includes our knowledge of the world and our knowledge of the situation. In other words, we use context. (For a more detailed discussion of the role of context, including the issue of what happens when context is “too reach”, and no linguistic proccessin is necessary, see (Krashen 1999a). For beginner, pictures are a tremdous help in making input comprehensible, as are the body movement that the at the core of Asher’s Total physical response (TPR) method.

Now that we have some of idea of the input/ comprehension hypothesis, I can share to mystical, amazing fact about language acquisition. First, language acquisition is effortless. It involves no energy, no work. All an acquirer has to do is understand messages. Second, language acquisition is involuntary. Given comprehensible input and a lack of affective barrier (see below), language acquisition will take place. The acquirer has no choice. In a theoretical sense, language teaching is easy: all we have to do is give students a comprehensible messages that we will pay attention to, and they will pay attention if the message are interesting.

Corollaries of the input/ comprehension hypothesis

if the input hypothesis is correct, the following corollaries are correct:

Talking is not practicing

The input hypothesis maintains that speaking does not directly result in language acquisition: talking is not practicing. If you practice your French out loud every morning in front of mirror, your French is not improved. Rather, the ability to speak is the result of language acquisition, not a cause. Speaking can help language acquisition indirectly however. First, it can result in conversation, and conversation as an excellent source of comprehensible input, even though what counts in conversation, however, is what the other person says to you, not what you said to them. I suspect that speaking can help you in another way: it can make you feel more like a user of the second language, like a member of the a “club”. I return to this argument a bit later, in the section of the affective filter.

Given the enough comprehensible input, i + 1 is present.

The second corollary states that if we provide students with enough comprehensible input, the structures they are ready to require will be present in the input. We don’t have to make sure they are there, we don’t have to deliberately focus on certain points of the grammar. If this corollary is correct, it means the end of grammatically basic language teaching.

Before discussing this, it’s important to emphasize that grammatical accuracy is an important goal. What we are discussing is how to attain this goal. I am arguing that comprehensible input is a better way of developing grammatical accuracy than direct instruction in grammar.

We all remember grammatically based classes. Students focus on one rule at a time, the idea being to “master” one rule and then move one on to the next. It simply doesn’t work. I will discuss 4 problems with grammatically based syllabus, problems that I think are unsolvable for the grammatical syllabus syllabus, but that compensable input solves with easy.

1. What if a student misses classes one day? If the class is based on grammar, the student has missed “the ryle of the day”. If the class is based on comprehensible input, however, there is no problem. Every class will contain a rich supply of grammar and vocabulary, and there is will be plenty of chances for students to get comprehensible input containing i + 1. With a grammar based teaching, the student gets only one chance, unless review is constant and extensive. With comprehensible input, there are many many chances.

2. Even and though we all acquire language in the same way, there is individual variation in the rate of quisition. Some students in class will progress faster than others. Individual variation in rate is especially likely in second -language classes; some students get more input outside of class than others. If the “rule of the day” is the past tense, some students may have already acquired it, and some may be nowhere near ready.With comprehensible input, everybody is covered, even though i+1 may be different for different students. We need not know exactly where each students is in his or her development path; all we need do is to provide a great deal of comprehensible input.

3. In order to teach grammar, a teacher has to know grammar, and this is a task that is getting harder every day. With each new discovery, with each new grammatical rule, each new rules of sociolinguistic competence, the curriculum gets more and more complex. And it will never end. But if instruction is based on comprehensible input, this problem disappears. If comprehensible input is plentiful, students will absorb their rules teacher and good authors use, whether teachers consciously know the rules or not, whether linguists have discovered them or not.

4. The final problem with a grammar teaching is the most serious: it’s boring. It is a very hard to say things that are interesting and comprehensible when you hidden agenda is the relative clause. But if instruction is based on comprehensible input, all we need to do is to present messages that are interesting and comprehensible, and grammar will take care of itself. As most teachers know. This task is difficult enough.

We can now return to the third amazing fact about the natural order hypothesis. With comprehensible input-based language teaching the syllabus is not based on the natural order. If the argument presented in this section are correct, the syllabus is not based on any grammatical order. Rather, students will acquire the language in a natural order as a result of getting comprehensible input.

The Affective Filter Hypothesis

The effective filter hypothesis claims that affective variables don’t impact language acquisition directly by prevent input from reaching what Chomsky has called the “language acquisition device”, the part of the brain responsible for language acquisition. If the acquirer is anxious, has low self-esteem, does not consider himself or herself to be a potential member of the group that speaks the language (see Smith 1988 for discussion of this last factor), he or she may understand the input, but it will not reach the language acquisition device. A block, the affective filter, will keep it out. The presence of the affective filter explains how 2 students can receive the same (comprehensible) input, yet one makes progress while the other does not. 1 student is open to the input while the other is not.

Application

Part 2: Applications

Do we need language classes?

Most people don’t think that language classes are necessary. Most people would say that the best way to acquire another language is to go to the country where it is spoken. But for beginners, this is a bad advice. If a beginner goes to the country, he or she will only encounter a great deal of incomprehensible input. Beginner are much better off in well- taught language classes. Good language classes will give the beginner the comprehensible input that the outside world will supply only very reluctantly. A beginner can get more comprehensible input in one session of a well-taught language class then from several days of being in the country.

The goal of the language classes is to bring the beginner to the point where he or she can go to the country and obtain comprehensible input. It’s important to point out the goal of the language classes is not to bring students to the highest levels of competence. The goal is to bring students to the intermediate level. When foreign language students reach this level they can go to the country and continue to improve on their own; they can have conversation and read at least some authentic texts. When a second-language students reach this stage, they can begin to get at least some comprehensible input from the environment and from the “mainstream” in school. They will not, however, be perfect.

For those who think this is too humble a goal, for those who expect perfection from language pedagogy, I should point out that the profession has not been particularly successful at this more modest goal. Moreover, this modest goal is consistent with a general philosophy of education that most of us subscribe to. After completing one’s basic education, one is not a master; true mastery comes only after years of experience. Education is, rather, a lunching pad: it prepares us to begin our profession, and we expect to grow and improve as we practice our profession. The idea applies to the beginning level language class.

The beginning level

At the beginning level, there are several methods that work. They are consistent with an underlining theory outlined here, and the research confirms that they work. Here is what we have in common:

1. The classroom hour is filled with aural comprehensible input. Teachers help make input comprehensible in several ways. First, they provide context in the form of the pictures and realia, and in the use of movement. In the powerful total physical response( TPR) method, language is taught using commands. The teacher gives the command, models the movement, and the student performs the action. Students are not asked to speak, only to try to understand and obey the command. The teachers modeling of their movement is the context that help make the command comprehensible.

2. In addition, teachers help make input comprehensible by modifying their speech. The adjustment they make, however, are not rigidly imposed. Rather, teachers naturally tend to talk a little slower and use somewhat less complex language as they try to make themselves understood.

3. The syllabus is organise it. A comprehensible input based method does not mean that we simply go in and talk to students. Comprehensible input based classes have less lesson plans and syllabi, but the syllabi are not based on points of grammar. Rather, they are based on activities ( games discussions on the topic of interest, projects) that students at that level and with their background will find interesting and comprehensible (they can be “enterprises”, as discussed in the chapter 4). Thus, an activity that might work for university-level French class in Boston may not work for the elementary school EFL class in Taiwan. Brown and Palmer (1988) suggest some very interesting activities for beginning language students that are useful for wide range of student: the magic tricks, simple scientific experiments, playing darts, playing card games, learning to do a headstand, etc. Brown and Thomas's suggestions illustrate the freedom comprehensible input teachers have. All that is required is that activity be interesting and comprehensible. There is no requirement that the activity provide practice with a particular grammatical structure. Is corollary to to the input hypothesis stated, given enough comprehensible input, i + 1 is automatically provided.

4. Demands for output our low and students are not forced to speak until they feel ready. Of course, students are not forbidden from speaking; in fact they are warmly encouraged to speak. As noted earlier, speaking per se does not cause language acquisition, but it can invite others to talk to you, and it can lower the affective filter by making the speaker feel more like a member of the group that speaks the language.

In comprehensible input-based methods, beginning students are able to participate in activities while saying nothing, or very little. Complete sentence are not required, and errors are not corrected. Theory predicts that grammatical accuracy is a result of comprehensible input, not output and correction, a prediction supported by the research showing disappointed result for error correction (Krashen 1981, 1994a; Truscott 1996).

1. Grammar is included, but only for older student (high school age and older), not for children. In the Natural Approach at the college level (Krashen and Terrell 1983), grammar is done as homework. Grammar is included for two reasons: first, to satisfy the curiosity of some student have about the structure of language in other words, is basic linguistics, a subject that is interesting and valuable. Second, consciously learned knowledge of grammar can be used to fill in some of the gaps left by incomplete acquisition ( see “the monitor hypothesis” above).

As noted earlier, acquisition will give us nearly all of a language, but not 100%. Writing that will be ready by other people must be 100% accurate. Comprehensible input-based methodology for older students therefore provides for the conscious learning of rules that many people, despite extensive listening and reading, may not acquire. Also is noted earlier, such rules should be used when they don’t interfere with communication, as in the editing stage of composing. It is not expected that rules learned in the grammar activities will be available for spontaneous use in conversation. There is no expectation. in other words, that learned grammar rules will become acquired.

Method comparisons: comprehensible input versus skill-building

Comprehensible input-based method have done very well in the published professional research literature when test I communicative, students in this classes show better result then those in traditional grammar-based classes.

When grammar tests are used, there is either no difference or comprehensible input students are slightly better. I present here a few samples of the research(see also Krashen 1981, 1982, 1994a).

Asher has published a number of studies in which TPR was compared to traditional foreign language methodology. Table 1-1 presents data from Asher. Subjects where public school children in the United states study Spanish as a foreign language. TPR students have received only 20 hour of instruction in TPR while comparison, traditionally taught children have received 100 hours (9th grade) or 200 hours (combine 7th and 8th grade classes). Traditional methodology included students repeating that instruction said, the use of translation to communicate meaning, and formal instruction in reading and writing, emphasising “Spanish grammar” (1044). The listening measures included 70 items and asked students to listen to sentence and view a picture at the same time and judge whether the sentence was true false or incomprehensible. The reading test was identical, except that subjects read the sentence.

The result was astounding. 6th grades with only twenty hour of TPR Spanish actually out for performance 9th grades with five times as much exposure to Spanish (d=0.58 for listening comrehension, d= 1.05 for reading comprehension). The combined 7th and 8th grades who did TPR also did better than the 9th grade comparison (d=0.81 for listening comprehension, d=1.18 for reading). Comparison with 200 hours of instruction outperformed the TPR students, but recall that they had 10 times as much exposure to Spanish!

This only one of my many studies show the superiority of TPR over traditional approaches. In Asher, Kusudo and de la Torre (1974) TPR students after 90 hours exceeded the 15 persentile of the standard Spanish test designed for students who had had 150 hours of instruction. In Asher (1972), adult TPR students of German who had had only 32 hours of instruction outperformed to control groups who had traditional instruction, one that had 40 hours and another that had 80 hours. Several additional replication of these results have been published (Asher 1966, 1969, 1972; Swaffer and Woodruff 1978, Wolfe and Jones 1982).

Recent Studies

Hammond (1988) compared the attainment of eight randomly selected classes of university level Spanish who experienced a comprehensible input based method, The natural approach (Krashem and Terrell 1983) with 52 classes that experienced “modified grammar translation”. On a grammar test given at the end of semester. Natural approach students were slightly better (according to my calculation).

Nicola (1990) compared “grammar audio lingual” methodology to a method that focused more on comprehensible input and meaning and less on form, for students of Arabic at the Defense language institute. The treatment lasted about 30 weeks and students met six hours per day, but about half the material was common to both groups. We have no detail about the nature of the of the test, no was it possible to compute effect sizes, but it’s clear from an inspection of the data that students who received more comprehensible input consistently did better on tests of listening and reading. For the first two groups studied, traditional students were better on an oral test but CI students in the third group studied better.

Winitz (1996) compared the progress of college Spanish students in the United states of 1 semester of an implicit grammar approach in which the focus was on the presentation of comprehensible input (36). Activities were organised so that basic grammatical forms were covered, but the focus was on meaning. Students in this section were free to speak when they wished, but error were not corrected. The explicit students follow a traditional approach, using a text that gave explicit description of a grammatical rules before or after many examples of usage. Table 1-3 presents the results of a test of grammaticality judgments. The test consisted of 54 sentence in Spanish and subjects were asked to indicate if the sentences were grammatical or ungrammatical. The implicit grammar group did better.

Nikolov in Krashen (1997) were a comparison of grammatical accuracy and fluency in two EFL classes in Pecs, Hungary, followed over seven years. The experimental class had a story-based syllabus, and a focus on the content with no formal grammar instruction and no focus on form until grade 8, the last year of the study. The comparison group followed a structural syllabus, with explicit rules, drills and exercises. An analysis was done of accuracy and fluency in the interview situation in the which students were asked to talk about themselves, describe a person they knew, a book they had read, or a film they had seen.

An analysis of nine grammatical morphemes in obligatory occasions showed that the experimental group was more fluent ( 3.366 obligatory occasions produced, compared to 2.742), and was slightly more accurate ( 87% correct, compared to 82%). The experimental group was more accurate on five of the nine items, and there was no difference on 2. This results confirm that comprehensible input can produce both accuracy and fluency. In addition, after the study was complete Nikolov maintained contact with the students: 13 of the 15 in the story-based class passed a form-focused proficiency examination in English at the university level, and the four became English majors.

Isik’s (2000) study shows that a combination of 75% comprehensible input and 25% grammar is more effective than 80% grammar and 20% communicative activities. Isik compared 2 groups of 20 students, low intermediates in EFL starting in Turkey in high school. The comprehensible input group devoted 7 hours a week to formal grammar study. The rest of the time (22 hours) was spent on TPR and communication based activities, with minimal correction. Students in this section also read 2 graded readers per week. The grammar group devoted 24 hours per week (out of 29) to form based activities, moving from mechanical to meaningful practice, with a focus on correct production “ … meaning was secondary and immediate correction was provided” (251). The duration of the study was 36 weeks, a total of about 1000 hours. Results presented in table 1-4 show that the comprehensible input group was far superior in all tests.

My goal is this section was not to present a complete survive of method comparisons, but simply to present some sample studies, and to give readers an idea of how robust the advantage for comprehensible input based construction is.

The intermediate level: sheltered subject- matter teaching

As effective as comprehensible input based methodology is, it’s not enough. Methods such as Total Physical Response and Natural approach provide conversational language. Second language students need more: they need advanced, or “academic” proficency (Cummins 1981), the language of history text, story problems in math, and the language of business, science, and politics. It’s also the language of the classics. A very effective way of develop academic language is through wide reading, a topic that is discussed in the next chapter. In this chapter, I present an additional way of doing this: sheltered subject-matter teaching.

Inspired by the success of Canadian immersion programs ( see Lambert in Tucker 1972 ), sheltered subject matter teaching derived from one important concept: subject matter teaching, when it is comprehensible, is language teaching, because it provides comprehensible input. Sheltered subject-matter teaching has two important characteristics:

1. It is not for beginner and not for native speakers of the language. It sheltered classes, only intermediate second language acquires participate. The input will not be comprehensible for beginners. Beginners are better off in TPR, Natural approach and related methods. When we allow native speakers of language into the class, there is a real danger that the input will not longer be comprehensible for the second language acquires. When all students are more or less in the same linguistic boat, it is easier for the teacher to make sure their input is comprehensible.

2. In shelter classes, students and teacher are focused on subject matter, not language. This emphasis on meaning and not form, results in more comprehensible input, and those more language acquisition. Sheltered the subject-matter classes are, thus, not” ESL math” or “ESL history: but a “math” and “history”.

Research on sheltered subject-matter teaching

Research on sheltered subject-matter teaching has shown that students in these classes require considerable amounts of second language, doing at least as well as students in a regular intermediate language classes, and they also learn an impressive amount of subject matter. Thus, shelter teaching is very time efficient; Students get both language and subject matter at the same time. Also, sheltered subject-matter teaching provides exposure to academic language. I present here 2 examples; for others see Krashen(1991).

The first study done with adult students on sheltered subject- matter teaching showed that university students at the University of Ottawa could learn both psychology and make progress in the second language at the same time (Edwards, Wesche, Krashen, Clement and Kruidenier 1985; Hauptman, Wesche, and Ready 1988). Participants, who were voluntaries, has already started one semester of college psychology in the first language (English or French), and had at least a low intermediate language knowledge of the second language (French or English). The shelter course was second-semester psychology ( in Haupman et al., one experimental group did sheltered psychology for two semester), and was supplemented by half our weekly session with a language teacher, who did no direct grammar teaching, but focused on comprehension of content and “developing strategies for effective reading and class interventions” (Haupman et.al. 445). In general, subjects made progress in second language acquisition equivalent to students in a regular second-language classes, and acquires subject matter just as well as students who took the same course in their first language.

Lafayette and Bosgalia ( 1985 ) reported that four 4th semester university level students of range who studied French civilization and culture did just well as well as traditional 4th semester class on several measures of French proficiency(listening and reading), and might better gain on the speaking test. Comparison were slightly better on a grammar test, but more than 20% of the items on this test deals with this subjunctive, and late acquired aspect of grammar that was emphasis in the traditional class.

Continuation studies

Indirect evidence of hypothesis that comprehensible input-based methods are effective finding showing that more students in these classes continue on and do additional study more than students in traditional classes. Swaffer and Woodruff (1978) reported that enrollment in second semester German classes increased after students experience a comprehension based first semester course. The attrition rate between the 1st and 2nd semesters and the traditional instruction was 45% and 47% in the two years of study. After comprehensive basic construction attrition dropped to 28% and 22% in two consecutive years.

Cononelos (cited in Sternfeld 1992) compared students who had completed five quarter of traditional skill-based foreign language instruction (German) with students at the same university who had completed five quarters of an immersion multi literacy program which was sheltered subject matter teaching focusing on culture and civilization. Of 109 traditional students, only 4 went on to take more advanced course in the foreign language. In contrast, 9 out of 22 for my shelter students went on to higher levels; according to my calculation, this difference is highly significant. While immerson/multiliteracy students made-up only 17% of fifth quarter students survived, they accounted for fully 69% of the student enrolled in upper division courses. Similarly in Lafayette and Buscaglia 1985, discussed above, most students from sheltered 4-semester French class said they intended to enroll for advanced French ( 50%, compared to 36 percent of the comparison students), and and 94% said the course was more interesting than other French courses they had taken at the same university.

In this chapter I reviewed some of the basic of language acquisition theory and some general applications. The next chapter discuss a powerful means of helping student move to more advanced level of proficiency, one that has been nearly completely neglected: free voluntary reading.

Free voluntary reading: still a very good idea

Free voluntary reading may be the most powerful tool we have in language education period. In fact, it appears to be too good to be true. It is an effective way of increasing literacy and language development, with strong impact on reading comprehension, vocabulary, grammar and writing. It’s also very pleasant. In fact, it is more than pleasant: it’s extremely enjoyable. Free reading may also be an important part of the solution to two related problems: making the transition from elementary level to authentic language use and from “conversational” language ability to “academic” language ability.

Free voluntary reading works, I propose, because it is a form of comprehensible input delivered in a low-anxiety situation (Krashen 1994a; Chapter 1 of this volume). In this chapter, I briefly review the evidence for the free reading, some practical issues, and, even though it’s hardly necessary, evidence has shown that free reading is enjoyable.

Research on free reading

Correlation studies

Studies in both second and foreign language acquisition confirm that those who read more do better on a wide variety of tests. I include here some recent studies in foreign and second language acquisition (see Krashen 1993b for early studies). In Stocks, Krashen and Kartchner (1998), students of Spanish as a foreign language in the United states were tested on their knowledge of the subjunctive on the test that attempted to probe acquired competence (in the result presented below, only subject who were not aware that the subjective was the focus of the test were included). Formal study was not predictor of subjective competence, not was length of residence in a Spanish speaking country. Stocks, Krashen and Kartchner also asked subjects about the quality of instruction they had had specifically in the subjunctive. This variable also failed to predict performance on the subjunctive test. The amount of free reading in Spanish however was clear predictor. Lee, Krashen and Gribbons (1996) reported that for international students in the United states, the amount of free reading reported (number of years subjects read newspapers, new magazines, popular magazines, fictions, and nonfiction) was a significant predictor of ability to translate and judge of the grammatically of complex grammatical construction in English (restrictive relative clauses). The amount of formal study and length of residence in the United states were not significant predictors. Results for the grammatically judgment task are presented in table 2-2.

Constantino, Lee,Cho, Krashen (1997) reported that the amount of free reading international students living in the United states said they did before taking the test of English or as a foreign language TOEFL was an excellent predictor of the score on this examination (table 2-3). In this study, formal study and length of residence were also significant (and independent) predictors.

Case histories

Cho and Krashen (1994 ) demonstrated substantial and obvious growth in the vocabulary in English as a second language in adults English acquirers who were encouraged to read novels in the Sweet Valley High series. Subjects had had some instruction in English as a second or foreign language (heavily grammar based), and began this Sweet Valley Kids (second- grade level) series, moving eventually to Sweet Twins (fourth grade level) and to Sweet Valley High (fifth and sixth grade level).

They showed clear gains in vocabulary, and vastly increased confidence in speaking English.

Segal (1997) describes the case of L, a 17 years old 11 th grade student in Israel. L. speaks English at home with her parents, who are from South Africa, but has serious problems in English writing, especially in spelling, vocabulary, and writing style. Siegel , L.’s teaches in grade 10th, tried a variety of approaches:

Error correction proved a total failure tried correcting your own mistakes ,tried proccess writing, and tried just copying words correctly in your notebook. Nothing worked L.’s composition poorly expressed and her vocabulary was weak. We conferenced together over format and discussed ideas before writing. We have made a little progress. I gave a L a list of five useful word to spell each week for six weeks andd tested her in an untreatening way during process. L. performed well in the tests in the beginning, but by the end of six week she are reverted to misspelling the words she had previously spelt correctly.

In addition, L. mother got her private tutor, but there was little improvement.

Seagull also taught L. in grade 11. At the beginning of year, she assigned an essay . When I came to L.’s composition I stopped still. Before me with me was an almost perfect essay. There were no spelling mistakes. The paragraph were clearly marked. Her ideas were well put and she made good sense. Her vocabulary had improved. I was amazed but at the same time uneasy. Segall discovered the reason for L.’s improvement. She had become a reader over the summer. L. told her I never read much before but this summer I went to the library and I started reading I just could not stop. L.’s performance in grade 11 in English was consistently excellent and her reading habit has continued. Cohen (1997) attended an English-language medium school in her native Turkey, beginning at age 12. The first two years were devoted to intensive English study, and Cohen reports that after only two months, she started to read in English as many book books in English as I could get hold of. I had a rich, ready-made library of English book at home. I became a member of the local British Council’s library and occasionally purchased English books in bookstore. But the first year of middle school I had become an avid reader in of English. “

Her reading, however, lead to an “unpleasant incident” in the middle school. I had a new English teatcher who assigned us two compositions for homework. She returned them to me ungraded, furious. She wanted to know who had helped me write them. They were my personal work. I had not even used the dictionary. She would not believe me. She pointed at a few underlining sentences and some vocabulary and asked me how I knew them; They were well beyond the level of the class. I had not even participated much in class. I was devastated. There and then and many years later I could not explain how I knew them. I just did.”

In- School free reading

In- school free reading studies includes evaluations of several kinds of program: in sustained silent reading: students read whatever they please (within reason) for a short time each day and there’s no accountability required. In extensive reading programs, a small amount of accountability is included; for example a short description of what was read. In self-selected reading programs, the entire class period is devoted to reading, and occasional teacher-students conferences are scheduled.

I have reviewed the available research on in-school free reading in several places (Krashen 1993,2001). In my most recent summary, I found that student who participated in this program did as well or better than comparison students in traditional language arts or second-language programs on tests of reading comprehension in 51 out of 54 comparison. The results were even more impressive when one consider only studies lasting 1 academic years or longer. In eight out of 10 cases, participants in in school reading programs outperformed comparisons and in two cases there there were no difference.

The national reading panel (NRP), supported by the US Government, also reviewed studies of in-school reading, and reached the starting conclusion that there is no clear evidence supporting this practice. They were, however, able to find only 14 comparisons, all lasting less than one academic year, between students in the school free reading programs and comparison children, devoting only 6 pages of their massive report to this topic (as compared to approximately 120 pages devoted to research of on phonetics phonemic awareness and phonics). Interestingly, in-school reading did not fare badly even in the limited analysis done by NRP, with in-school reading doing better in four cases and never doing worse. Note that even the finding of “no difference” suggests that free reading is just as good as traditional instruction, an important theoretical and practical point. Because free reading is so much more pleasant than regular instruction (see below) and because it provides readers with valuable information, a finding of no difference provides strong evidence in favor of free reading in classrooms.

I have also argued (Krashen 2001) that the NRP not only missed many, many studies, they also misinterpreted some of them once they included. I present here a discussion of recent studies that have particular relevance to the EFL situation. In Elley and Mangubhai (1983), 4th and 5th grade students of English as a foreign language were divided in the three groups for their 30 minutes daily English class. One group had traditional audio lingual method instruction, second did only free reading, while I thought did the “shared reading.” is a method of sharing a good book with a class, several times, in such a way that the students are read to by the teacher as in a bedtime story. They then talk about the book they read it together, they act out the story, they draw part of it and write their own caption, they rewrite the story with different character or events. After two years, the free reading group and the shared reading group were far superior to the traditional group in tests of reading comprehension, writing and grammar. Similar results were obtained by Elley (1991) in a large-scale study of second-language acquirers, ages 6 through 9 in Singapore.

Elley’s recent data comes from South Africa and Sri Lanka. In all cases, children who were encouraged to read for pleasure outperformed traditionally taught students on standardized tests of reading comprehension and other measure of literacy. Table 2-4 presents the data from South Africa. In this study EFL students who lived in the print-poor environments were given access to sets of 60 high interest books, which were placed in classroom, with another 60 mate available in sets of 6 identical titles. The book were used for read- alouds by the teacher for shared reading and for silent reading. Table 2-4 presents data from different provinces; In every case the readers outperformed those in comparison classes and the gap widened with each year of reading.

Mason (Mason and Krashen 1997) developed a version of extensive reading for university EFL students in Japan in which students do self select reading of a pedagogical readers as well as a easy authentic reading. Accountability were present but minimal; Students only had to write short appreciation of what they have read. In three separate studies, Mason found that extensive readers made greater gains than comparison students who participated in traditional form based EFL classes. Table 2-5 presents the details of a three study in the form of effect sizes comparing with extensive reader to the traditional taught students.

Lao and Krashen (200) compared progress in reading over 1 semester between university level EFL students in Hong Kong who participated in popular literature classes that emphasized reading for content and enjoyment, including some self-selected reading, and the students in a traditional academic skills class. Application of statistical tests, including those that accounted for pretest differences, confirmed that the superiority of the popular literature group was statistically significant. As shown in table the popular literature students made better gains in vocabulary and reading rate and, at the end of the semester, clearly felt that what they had learned in the course would help them in their other university courses.

Shin (2001) examined the impact of six-week self- selected reading experience among 260 and 70 graders who had to attend summer school because of low reading proficiency. Students attended class 4 hour per day; during this time approximately 2 hours were devoted in to sustain silent reading, including 25 minutes in their school library. The district invested $25 per student on popular paperback and magazines, with most book pushesed from Goosebumps series. In addition, about 45 minutes per day were devoted to reading and discussing novels such as “Holes” and “The Island of a blue dolphins”. Comparison children followed a standard language art circulum during the summer. Attrition was high for both groups but similar (class size dropped from 20 to 14.3 among readers, and from 20 to 13.2 among comparisons) as was the percentage of limited English proficient children (31% in the reading group, 27% in the comparison group). The reader gained approximately 5 month on the Altos test of reading comprehension in vocabulary over six-week period, while comparisons declined. On the Nelson Denny reading comprehension test, the summer readers grew a spectacular 1.3 years ( from grade 4.0 to grade 5. 4). On the vocabulary section, however , the groups showed equivalent gains.

The Author recognition test: A methodical breakthrough

Stanovich, in series of study, has verified the value of simple procedure for starting an impact of reading. In the other recognition test, subjects simply indicate whether they recognize the names of author on a list. For speakers of English as a first language, scores on the other recognition test have been shown to correlate substantially with measure of vocabulary, reading comprehension and spelling. This results have been confirmed using other first languages as well: significant correlations have been reported between performance on author recognition test and writing performance on Chinese and Korean, and between author recognition test performance and vocabulary development in Spanish (Rodrigo, McQuillan, Krashen 1996).

Those who report reading more also do better on the other recognition test. This true for English speaker, Korean speaker, China speakers, and Spanish speakers. One study also reported a positive correlation between performance on the author recognition test and the amount of reading subjects were observed doing. West, Stanovich and Mitchell observed airport passenger waiting for flights and classified them as either reader those who were observed to be reading for the least 10 min continuous minutes) and nonreaders. Readers did significantly better on an author recognition test as well as on a vocabulary recognition test.

Only one study thus far has examined the performance of foreign language students on the author recognition test. Kim and Russian reported that for high school students of English as a foreign language, performance on an English author recognition test was a good predictor of performance on an English vocabulary test. In addition, those who reported more free reading in English also tended to do better on the author recognition test.

In addition to providing confirmation of the relation between recreational reading and language development, the author recognition test and similar measures (magazine recognition test, title recognition test) promise to simplify work in this area.

Light reading as a bridge

Of course, a great deal of free reading will be a light reading research by Hayes and Arthur (1998) supports the idea that lighter reading can prepare readers for heavy reading. According to the finding, it is highly unlikely that much educated vocabulary comes from conversation or television. Hayes and Aherens found that the frequency of less-common words in ordinary conversation, whether adults-to-child or adult-to-adult was much lower than in even the lightest reading. About 95% of the words used in conversation and television are from the most frequent 5000. Printed text includes far more uncommon words, leading Hayes and Ahrens other to the conclusion that the development of lexical acknowledge beyond the basic words requires literacy and extensive reading across a broad range of subjects. Table 2 presents some of the data, including two of the three measures they used for word frequency. Note the light reading ( comics, novels, other adults book, and magazines), although some that closer to conversation, occupies a position between conversation and abstract of scientific papers.

All the adventures of reading

In Krashen (1994b), I proposed the pleasure hypothesis: pedagogical activities that promote language acquisition are enjoyable, and those that do not are not enjoyable (and may even be painful). Of course, just because an activity is enjoyable doesn’t mean it’s a good for language acquisition; some activities may be very annoyable but may not help at all. Enjoyment is no guarantee of effectiveness. It is, however, interesting that there is a strong evidence that free voluntary reading is very anjoyable.

The evidence includes work by Cskszentmihalyi (1991) who introduced the concept of law. Flow is the state people reach when they are deeply but effortlessly involved in an activity. In flow, the concerns of everyday life and even the sense of self disappear - our sense of time is altered and nothing but the activity itself seems too matter. Crosscultural studies indicate that flow is easily recognised by members of widely different culture and groups. For example, members of Japanese motorcycle gangs experience flow when riding and rock climber experience flow when climbing (Massimini, Cskszentmihalyi and Della Fave 1992).

Of special interest is the finding that reading is currently perhaps the most often mentioned flow activity in the world. This finding is consistent with reports of individual pleasure readers. A resident in Walse and Northern Italy said that when he reads “I immediately immerse myself in the reading and the problem I usually worry about disappear”. One of the Nell’s subjects reported that “reading removed me from the irritation of living for a few hours a day I read ‘trash’: I escaped the care of those around me, as well as escaping my own cares and dissatisfactions.” W. Somerset Maugham quoted in Nell (1988), has similar comments: “Conversation after time bores me, games tire me and my own thoughts which we are told are the unfailing resource of sensible man, have tendency to run dry. Then I fly to my book as a opium-smoker to his pipe (232).

Nell provided interesting evidence showing why bedtime reading is so pleasant. Pleasure readers were asked to read a book of their own choice while their heart rate, muscle activity, skin potential, and respiration rate were measured; level of arousal while reading was compared to arousal during other activities, such as relaxing with eyes shut, listening to white heart noise, doing mental arithmetic, and doing visualization activity. Nell found that during reading arousal was increased as compared to relaxation with eyes shut, but a clear decline in arousal was recorded in the period of just after reading, which for some measures reached a level below the baselines (eyes shut) condition. In other words, bedtime reading is arousing, but then it relax you. Consistent with this finding are Nell’s results showing that bedtime reading is popular. Of twenty-six pleasure readers he interviewed, thirteen read in the bed every night and eleven “almost every night” on “most nights”.

Free reading has additional benefits. Lee and Krashen (1997) proposed that those who read more have less writing apprehension because of their superior command of the written language. They reported a modest but positive correlation between the amount of reading done and scores on writing apprehension questionnaire for Taiwanese high school students. The modest size of a correlation (r=.21) maybe because other factors affect writing apprehension, such as a mastery of their composing process. It is a consistent, however, with reports that those with less writing apprehension enjoy reading more. Free reading is also an excellent source of knowledge: those who are read more know more.

Motivating students to read

If reading is so annoyable, do we have to worry about motivating students to read? I think we do, but the task is much simpler than we thought. There is a good evidence that rewards an incensitive play no role in increasing the amount of reading done nor does it impact gains in the reading comprehension (McQuillan 1997). The simpler solution is to provide students with access to plenty of interesting and comprehensible reading material and also provide some time for them to read. There is evidence that this works.

The impact of reading itself.

Those who participate in the in school free reading programs are motivated to read more (Pilgreen and Krashen 1993). Greany and Clarke(1973), in fact, rported that children who participated in a sustained silent reading program reported reading more than comparison students six year after the program ended. Tse(1996) describes the case of Joyce, an adult ESL student in the United States who did not view reading as a leisure activity and had never read a book in English before coming to the United states. After participating in an extensive reading class, her attitude towards reading change it dramatically, and she continued to read after the end of the course, and she recommended that her husband take the same class, rather than a traditional class.

The impact of one trip to the library

Ramos taught in elementary school that had a inadequate school library. He and his fellow teacher organized a field trip for the second grade student to nearby Public Library, at a time when the library was closed to the public and the librarian was available to help and interact with the teachers and children. Ramos documented a clear and dramatic growth in interest in reading among the children after this visit. Cho and Krashen (2002) documented a clear increase in interest in reading and in promoting pleasure reading among teachers after one exposure to interesting and comprehensible children’s literature.

The impact of One positive reading experience

Jim Trelease (2001) has suggested that one positive experience with reading can do the job, one “home run” experience. Two recent studies have confirmed that a surprising percentage of elementary school children report that they did indeed have one very positive experience with reading that got them interest in reading. (Von Sprecken, Kim, and Krashen 2000; Kim and Krashen 2000). In both cases, children reported a wide variety of home run books which strongly suggests that reader should have exposure to a reach variety of reading material.

There are several ways of helping ensure that at home run experience take place. The best ways to make sure interesting reading is available, reading that students really want to do. Simply recommending books is an obvious step. Others include read aloud (Trelease 2001), modeling reading (e.g. reading while children are reading during sustained silent reading time; see Whelldall and Entwhistle 1988 for evidence), and interesting book discussions (the core of language arts) as well as providing time to read. The time issue is an important one: there is evidence that interest in reading remains strong as students get older, but the pressure of school, and sometimes work, result in their having less time to read (Krashen and Sprecken 2002).

Providing time to read

Simply providing time to read results in reading. Von Sprecken and Krashen (1998) observed sustained silent reading sessions (SSR) in middle school in the middle of the school year and reported that 90% of the students weere reading. More reading tended to take place in those classroom in which more books were available in the classroom library, in which teachers also read while students read, in which students were not required to bring their own books, and in which teachers made deliberate efforts to promote certain books. In one of the 11 classes observed, there were few books, no modeling of reading, no promotion of books, and students had to bring their own books. Nevertheless, 80% of the students in the class were observed to be reading during SSR.

Cohen (1998) unobtrusively observed 128th eight-grade students during SSR time over 2 week period, and found that 94% were reading during SSR. She noted the enthusiasm for sustained silent reading was not high at the beginning of the school year, but increased after one or two months.

Herda and Rarnos (2001) reported that 63% of students in SSR sessions in grades 1 through 12 were actively reading; in grades 1 through 5, the percentages were much higher, ranging from 76% to 100%. In the upper grades, students were given the options of starting or pleasure reading, and a substantial percentage took advantage of the study option. Nevertheless, a surprising percentage were reading for pleasure, ranging from 29% in grade 12 to 65% in grade 9. Overall, 21% of the sample were studing during SSR time and only 17% were neither reading nor studing.

Some Innovations

Handcrafted Books

a problem with free reading in the 2nd and foreign language situation is that it is hard to find texts that are both interesting and comprehensible; The beginning student will find authentic text too difficult. There are two solutions to this problem. One is simply to find the best pedagogic readers and make them available for free voluntary reading. A second is a recent innovation called“ handcrafted books”( Dupuy and Mcquillan 1997 ). Handcrafted books are written by intermediate students, corrected by the teacher, and are to be read by beginners. Writers are instructed not to look up word while writing; if intermediate students don’t know a words, the chance are good that beginner won’t know it either. Handcrafted books thus have a good chance of being interesting and comprehensible; They are written by peers who are slightly more advanced than their readers.

Sheltered Popular Literature

A very useful adjunct to sustain a silent reading in a class of popular literature. Even foreign language students who are well read in their first language might not be aware of the options for pleasure reading in the second language. Sheltered popular literature expose student to the different kind of light but authentic reading available, moving from comics and magazines to novels. Such a course is taught as a literature; that is, with discussion of the values expressed in the reading as well as the insights they provide on their culture (for suggestions, see Dupuy, Tse, and Cook 1996 ). Our hopes is that such a course will help students discover one or more kinds of light reading they would like to do on their own. For evidence that such a course can actually work, see Lao and ` Krashenn ( 2000), discussed earlier.

If students become enthusiastic readers of any type of reading, they will progress enormously; better readers are typically“series” readers (Lamme 1974) - readers of Nancy Drew, The Black Stallion, John R. Tunis, Sweet Valley High, Goosebumps and Fear Street, and so forth). Narrow reading build language and literacy competence rapidly, thanks to the familiar context and resulting higher level of comprehensibility. In addition, acquisition of any written style should facilitate comprehension of any others; While there are differences among of different type of pros, there is also substantial overlap (Biber 1988 ); Someone who can read light fiction easily has acquired much of that what is needed to read academic prose.

Conclusion

There is a overhealming evidence for recreational reading as a means of increasing second language competence. In fact, it is now perhaps the most thoroughly investigated and best-supported technique we have the second-language pedagogy. Only one aspect of recreational reading remains uninvestigated: why is not it used more frequently in second language program?

Notes

1. Sze (1999) evaluated an extensive reading program in Hong Kong. 496 students from 5 schools, aged 13 through 15, were engaged in an extensive reading project (the Hong Kong extensive reading scheme) that had the following features,

2. Each class of 40 was given 100 books; So they suggest that this is a large number, but it’s not. It is only 2.5 books per student.

3. Students had put free choice in book selection but had to answer comprehension questions; “ question and answer cards” were provided with each book ( 64 ).

4. Students had regular conferences with teachers, and teachers gave support through“ awards” and “ praise” ( 65 ).

5. One to two periods per week were devoted to extensive reading.

Note that this is a version of“extensive reading”( free reading with accountability), and has a few features that may not be optimal: book access was limited, incentive were used (for evidence showing the lack of positive effect of incentive with, see Macquillan 1997 ), and reading was massed (all at once), rather than distributed ( some reading each day; see Pilgrin 2004 suggestive evidence that distributed SSR is a better option).

The readers responded to a questionnaire after two years, the comparison after one year. Readers reported reading more, reporting that they typically read about two hours per week, compared to about 1/2 hour per week for comparisons. Those in the extensive reading group also reported reading an average of 26 books over the last year, while comparison only reported a reading five. There was, however, considerable variation within the reader group, with some reading over 100 books in the last year, others very few. Those in the extensive reading group reported a modest increase in the interest in the rating, with 7% reporting that their interest in the reading increased “a great deal” and 62 person reporting that it had increased open court moderately”. Only 4% reported a decline in interest in reading. Readers also reported increased confidence in reading English. For example, 74% agreed with the statement “I can read English books independently without much help from the teacher”, as compared to 64% of the comparisons.

Readers also felt that they had improved; most felt that reading had improved their vocabulary ( 77%), with less perceived improvement with other aspect of language competence ( 62% per trading improved their reading comprehension, 50% that is had improved their writing, 46% their grammar, and 19% their speaking).

This extensive reading program was clearly successful - there was a clear increase in reading, a modest increase in the interest in reading (with clearly five negative reactions), and perceived improvement. With more access, less accountability, and distributed reading times, it might have done even better.

Subjects in Yang (2001) were students are in for evening adult EFL classes in Hong Kong. All had passed an exam in a level equal to 450 on the TOEFL in grade 11. They attended class for free (consecutive) hours per week for a total of 15 weeks. Students ( A and B below) read two Agatha Christie novels in addition to the reading materials done by all students in all four classes. Students read about 40 page per week. About an hour was spent in class per week discussing the book (“… plots, characters, and social issues students found in the book and how those issues could be related to present day life” [455]).

The pre- and post tests were identical, and multiple choice tests of “grammar, sentence structure and usage” ( 454-5). Yang performed an omnibus analysis of a variation of parenthesis (ANOVA), which revealed “ strong evidence that at least one class is different from the rest” (457) but did not perform post - hoc comparisons. He noted, however, that classes A and B made about twice the gains while the other two classes made. Combining race course for classes A and B (readers), and C and D(non readers), I calculated an effect size of 6.3, which is enormous and easily statistically significant (posttest means for reader = 74.6, standard deviation( SD)= 1.26; post tests means for non readers= 66. 9, standard deviation = 1.18; pre test scores were nearly identical for all 4 groups).

Results of a questionnaire administered showed that most readers understood their books, and felt that reading was beneficial. Only 20% had read a novel in English before. Is Yang points out, there are comfounds. Those who did the reading spent more time on English, and also had writing assignments related to the novels. His conclusion is reasonable:“… The extra time on reading in English is time well spent” (460).

Of course, one could argue that extra time spent doing grammar is also well spent, but studies of in class sustained silent reading and related programs in which students spent the same time in skill building and reading show reading to be more effective, is not there in the text. Also is noted in the text, low in Krashen (2000) reported that university EFL students in Hong Kong who participated in popular literature- based classmate greater gains in vocabulary and reading rate than students in traditional classes. Students in the literature class reported more reading outside of the of school, but those in the comparison class spend more time watching TV and movies in English, used English more in conversation, and significantly more time in academic study of English. This result confirm the time spent in reading is indeed very well spent.

1. Horst, Cobb, and Meara (1998) provide evidence supporting that hypothesis that vocabulary is required via reading, but conclude that reading is“not a very effective way” for those at lawer level of competence to increase their vocabulary. Subjects in their study were “low-intermediate” students of English as a foreign language in Oman. Students read a simplified version of the Mayor of CasterBridge, consisting of 21.232 words. The procedure was“rather unorthodox”: students followed along in the text while the story was read aloud in the class by the teacher in six class sessions. This was done to ensure subjects covered in the entire text and to prevent students from looking up the words while reading. Horst, Cobb and Meara assure us the students were “ absorbed by their story”( 211 ).

Horst, Cobb, and Meara constructed a multiple-choice vocabulary test of 45 words considered to be potentially unknown to the subjects. On a pretest given a week before the reading, subjects averaged 21. 64 correct (sd = 6.45). Thus, 23 roots remain for potential acquisition. On the post test following the reading, subjects averaged 26.26 correct(sd = 6.43), again of four. 62, or 22%(effect size based on the pre and post test = 0.72). This rate is somewhat higher than that seen in previous studies using adult second language acquirers ( e.g., Pitts, White, and Krashen 1989; Day,Omura, and Hiramatsu 1991; Dupuy and Krashen 1993), which Horst, Cobb, and Meara attribute to the fact that a longer text was used. They describe the increase as a “ small but substantial” (214). Despite this conclusion, Horst, Cobb and Meara argued that for acquires this level, reading is not enough. Reading a 20.000 word book resulted in 5 word increase: even if they read one such book a week this would translate into a gain of only 250 words per year, insufficient progress to reach the 5000 word level considered by some to be the minimum to read authentic texts. Since the students have “limited time”,“vocabulary growth needs to proceed more rapidly” ( 221).

There are several problems with this conclusion:

1. It is not clear that direct teaching resulted in true acquisition of vocabulary; Direct reaching results in learning, not acquisition, a fragile kind of knowledge that is unavailable unless stringent conditions are met, and that fades fairly quickly with time (see chapters 1 and 4 ).

2. The treatment may have underestimated the impact of reading. As noted above, the students did not read at their own space, but followed along in the text as it was read aloud, a method that prevents the reading and pausing that naturally occur with reading. In addition, subjects may have acquired words from the text not included in the test. ( and, of course, readers get other linguistic benefits from reading, such as better grammatical development and acquisition of “planned discuss”, as well as knowledge and pleasure; see text.)

3. There is no evidence that those who have reached the 5000 words level did it via direct instruction and study of vocabulary. Native speakers with a good vocabulary, in fact, attribute their attainment to reading, not study. Smith and Supanich (1984) tested 456 company presidents and reported that they had significantly larger vocabulary scores than a comparison groups of adults did. When asked if they had made an effort to increase their vocabulary since leaving school, 54.5% said they had. When asked what they did to increase their vocabulary, about half of the 54. 5% mentioned reading. Only 14% of those who try to increase their vocabulary ( 3% of the total group) mentioned the use of vocabulary books. Smith and Supanich’s presents were more advanced than the subjects in Horst, Cobb and Meara; it would be a great interest, however, to determine how second-language and foreign-language acquirers who are successful in reaching their 5000 word goal digit. It is hard to imagine that they studied 5000 flash cards.

Current issues and controversies: does grammar teaching works? What about comprehensible output?

In this chapter, I examined two current issues in second-language research and teaching, beginning with what is certainly the most persistent question in the field: the role of grammar. I then examine a recent claims, that we require language by a comprehensible output

The role of grammar

In Krashen(1992, 1993a) I argued that studies that attempted to demonstrate the efficacy of direct instruction in grammar showed only the grammar teaching has a peripheral effect. I review here a number of studies that have appeared since that time, and come to the same conclusion, in disagreement with the researchers themselves, who is in nearly every case conclude that they have shown that grammar instruction are focusing on the form works.

I define direct grammar instruction here is consisting of two components: (1) focus on form and (2) presentation of the rule. It is possible to do (1) without (2). that is focus student on form without presenting rule. This is done in several studies described below. It is not, however, possible to do (2) without (1): when we present a rule, we are also focusing students on form.

Condition (2) can take one of two forms. In one version, the students are given the rule: this has been termed deductive grammar learning. In another version, the students are asked to try to figure out the rule on their own (inductive rule learning). The latter version is termed rule search in some of the studies discussed here. It is not acquisition because the goal is the discovery of any explicit rule.

I first review a set of experiments in which the impact of direct instruction in the grammar is measured directly. I then review studies in which it is claimed that the impact of grammar can be felt not only on grammar test but on the “free-responses” test as well.

Experimental studies claiming to show an effect for grammar

The studies reviewed in this section have several characteristics in common:

1. Subjects were experienced adult language students, which means they were used to direct teaching of grammar, expected it and had survived it.

2. Comparison groups had either no treatment at all, or received what can only be described as improverished comprehensible input. Comparison subjects, we are told, sometimes focused on meaning, but this focus was always in the extremely contrived situation in which context and interest was minimal. In addition , it’s clear from some of the studies that some of the comparison students were, in fact, focused on form. These studies, thus, investigated only the impact of more direct instruction (more focus on form and more explicit presentation of rules) versus less, not learning versus acquisition— contrary to what some of the investigators claim.

3. Tests were given soon after the treatment was finished; Thus, only the short-term effect of conscious learning was investigated.

4. In the in the studies discussed in this section, the free conditions for monitor use (see chapter 1 ) warm it on all measures used. (I reserve for the second section studies in which the measures are claimed to be “free-response” on monitor-free.) I repeat here the conditions for Monitor use:

5. The acquirer must know the rule. In the study reviewed here, subjects had thoroughly studied the rules that were being tested.

6. The acquirer must be thinking about correctness, or focused on form. This was clearly the case for all experimental groups discussed here, and was often the case for the comparison group as well.

7. The acquirer must have time period in all measures used in the studies reviewed here, subjects had time to apply the rules they had studied. While it is acclaimed that the open court “time” condition was not satisfied in several of the studies, in only one case was time pressure quantified, and I argue that there results show that there was sufficient time for rules to be applied.

The consistent result is that those who had more rule based instruction and formed focus did better, but in nearly all cases the effect is quite modest in some cases it is completely absent. The status does only show that more instruction means a bit more cautiously learned competence, a conclusion that is consistent with the claims of the Monitor hypothesis (Krashen 1982).

Master (1994)

Subjects: subjects Aurora university ESL students at UCLA in California State University, Fresno. All were considered to be the intermediate level at most all in ( 72% in the UCLA sample, 69% in the Fresno sample) head study the target rule, the English article, before the treatment occurred.

Treatment: experimental subjects received a six hour of systematic direct instruction on the article over a nine week period.

Comparison group: subjects in comparison groups had instruction on writing during the experimental period. Errors in the articles were corrected on essays, but there was no formal teaching of the article or in these classes during this time.

Measure: The same measure was used as a prayer and post tests, a fill- in-the- blanks test in which students supplied the correct form of the article. Master provides these examples, some involve just one sentence: “Carol is__ student at our university”; some involving pairs of sentence: : Once there were many trees here. Now, trees are gone”; and o ther involving a paragraph: “ favorite food of __ jaguar is __ wild pig __ wild pigs move in __ bands of fifteen to twenty. They have __ great courage and ___ strength in __ groups.”

This measure clearly focused on students of form. Master suggests, however, that the test measured acquisition, because “subjects were given the test without prior announcement and they were only given enough time to answer without deliberating upon their responses. It was hoped that the test was thus reflective” ( 232). The nature of the time constraint was not discussed in any more detail.

Results: table provides the results of the original study and replication (Fresno). Master reports that experimental group gains were statistically significant, but comparison group gains came close, reaching there one tail .10 level in the UCLA study in falling just short of this level in the Fresno study.

Most important, the gains were very modest: after six hours of intensive study, the two experimental groups gains only 6.5% and 9%, respectively. Master also calculated the effect size for the difference in gain scores between the experimental and control group and the UCLA study, based on the means and standard deviations, and reported as an effect size (d) of .664.

Rather than demonstrating that instruction works, Mmaster has confirmed the limits of conscious learning: using subject who are supportive of an experienced with grammar learning ( international start students at the ….e parenthesis comma and who underwent intensive study(actually review in the most cases) of the target rules, and using and discretepoint grammar test focused on exclusively on their target role administered very soon after the treatment, gains were very modest.

The average UCLA student went from about of C grade to low B, where Fresno students went from about a C- or D+ to a C.

Master also reported no significant correlation between amount of formal study and performance on the pretest, confirming the inefficiency of instruction. He also reported no significant correlation between length of residence in the United States and article pretest performance. It would be interesting to see if amount of pleasure reading an article use related, is other studies have reported a clear relationship between acquisition of complex syntactic forms and reading.

The impact of instruction on the free response tests

in the previous section, I reviewed a number of studies of that claim to show that instruction is effective. In my view, the results of this study simply confirmed that consciously learned knowledge can be displayed on test of the consciously learned knowledge that is, other tests in which the condition for the use of conscious monitor are met. I also argued that the impact of instruction is not impressive; After what is often a great deal of study and direct instruction, gains, even on grammar tests, are modest. These results only confirm that direct teaching results in a fragile kind of knowledge that is applicable only when we are a severe condition are met, the conditions for Monitor use. In fact, this kind of evidence has been available to the profession informally for the centuries.

The crucial test of the efficiency of instruction is whether it has an effect for monitor free tests given a substantial amount of time after the treatment. No study has shown that consciously learned rules have an impact on monitor free tests over there long term.

In this section, I review studies but, according to appendix a in Norris and Ortega (2000), evaluated the impact of direct instruction and used “free constructed response”, measures that “required participants to produce language with relatively few constraints and with a meaningful communication as the goal for L2 production. Of the 11 studies cited by Norris and Ortega that, exclaimed, utilize it such tests, I was able to locate ten of them. No one provide any evidence for the efficiency of direct instruction. One study, in the fact, had nothing to do with impact of instruction (Yang and Givon 1997) and in another it was unclear whether there was a focus on the form in the treatment so given to any of their groups studied. Inner thought, Salaberry, subjects produced very few instances of the tag instructor on the free response test, which made the interpretations of the result difficult.

In Salaberry (1997) and in all other studies subjects have had intensive instruction on a specific target form. This was clearly not the case in the studies in which it is a claim to that acquisition is primarily at work; Subjects in these studies were not tested after a pedagogical treatment focusing on specific roles (the morpheme studies). Also, post tests were not delayed long enough for their conscious knowledge to fade; The longest delay in any of the measures in study using free response was three months. All postests will administrator immediately after their treatment in Doughty (1991),Jourdenials,Ota. Some tests were delayed in other studies, but never for more than three months. The delay in White,Spada LightBrown, Ranta(1991) was 5 weeks, ….

Another crucial point is that comparison groups in these studies did not have exposure to the target rules in any form; They were not in an acquisition rich environment. Thus, they started really compare the presence or an absence of formal instruction, no acquisition versus learning. Thus, regardless of the kind of test used, subjects in these studies were focused on form, in most cases were presented with the actual rule, and were tested soon after instruction occurred. There is, however, good evidence that the conditions for the use of the conscious monitor were met during the free response tests.

In some cases, students took other tests that clearly had a focus on form, which confirmed the subjects that the name of game was accurate performance on certain structures. In several cases, subjects were actually prompted during the administration of the test to use target form when they did not initially do so and at least two studies the test was similar in format to some class activities obviously designed to teach the target forms. In no case those subjects subjected to the time pressure that exists in a regular conversation. On the oral tests, the subjects were asked to describe a picture or sequence of pictures or engage in the information gap activities in which they asked questions of a native speaker, and in two studies, tests were written; the oral tests were an interview, but no indication is given whether time pressure existed or not. It is thus no surprise that in most cases there was a positive effect for instruction wow that is a comma instructed groups did better than uninstructed groups even on the free response tests. Norris and Ortega did not provide details of the studies or a description of how they arrived at their results. They also do not inform us what the effect sizes were for individual studies. I therefore review the nature of their treatment, the measure, and the results for each study available to me in which so-called free response measure was used. Studies are presented in the order in which way appeared in print. ….

Summury of Studies

in this section I examine studies using the free response tests. Despite their label, students who take such tests, I conclude, are indeed focused on form, thanks to the treatment, other measures, and the nature of the test themselves. It is thus no surprise that one sees some impact of instruction on these tests: the conditions on the use of the monitor are met.

The monitor hypothesis predicts that one will see more impact of instruction on tests when more conditions for the use of the monitor are met, or when conditions are met more fully. This predicts that the free response tests will show less impact of instruction than in performance on what are clearly discrete-point grammar tests in which there is more focus on form.

Norris abd Ortega reported a small mean effect size for studies using what they considered to be free response measures then for study using measures involving more focus on form, which is consistent with the predictions of the monitor hypothesis:

Norris abd Ortega’s overall results are entirely consistent with the monitor hepatitis column measures using selected response and constraint response have the most focus on form. Grammaticality judgment focus tests takers on form, but they can be done on the basis of learning or acquisition; Rule knowledge is not essential. Free response typically invokes the monitor least, but still invokes it to some extent, as argued here.

We have a clear test of this hypothesis, however, when the same subjects take a free response test and that either I selected or constrained response test, the two that are the most form focused period of the 8 studies in which free response measures were used, I was able to find 4 in which the same subjects took from focus to test as well.

….

It is unreasonable to support that test given that to students in a classroom situation preceded by intensive study of a particular grammar rule will involve some form focus, especially when the test allows students some processing time. Be then be the case that the free response tests will produce not a zero effect but a smaller effect than a traditional grammar tests. This is precisely what the Norris and Ortega found. In addition, the impact of grammar study, even on the grammar tests, is not spectacular. The Ortega review should not be interpreted as a dream for direct instruction, but if as fully consistent with the monitor hepatitis. Norris in Ortega note that the vast majority (about 90 percent) all studies of the effectiveness of instruction use measures that involve“script and focused linguistic tasks” and that the research, consequently, has been based primarily on the application of “explicit declarative knowledge under controlled condition, without much requirement for fluent, spontaneous use of a contextualized language”. This was, of course, my conclusion in the first section of this chapter. Investigation of studies using free response measures, however, shows that no matter how we label these measures, they, too, invite the use of consciously learned the grammatical knowledge.

Comprehensible output?

The comprehensible output(CO) hypothesis states that require language when we attempt to transmit a message but fail and have to try again. Eventually, we arrive at the correct form of our utterance, our conversation partner finally understands, and we acquire the new form we have produced.

The originator of the comprehensible output hypothesis, Merrill Swain (1985), does not claim that CO is responsible for all or even most of our language competence. Rather, the claim is that “ sometimes, under some conditions, output facilitates second language learning in ways that are different from, or enchance, those of input” (Swain and Lapkin 1995, 371). A look at the data, however, shows that even this weak claim is hard to support.

The scarcity argument

A problem all output hypothesis have is that output is surprisingly rare (KRASHEN 1994a). In the case of CO, the problem is especially severe. A recent confirmation of the scarcity of output is in Ellis, Tanaka, and Yamazaki (1994), which examined vocabulary acquisition under 3 conditions, task in which EFL students heard (1) “premodified” input (input recorded from a task performed with a native speaker and non-native speaker who could request clarification); (2) interactionally modified input (the nonnative students could interact with the native speaker); Or (3) unmodified input (input recorded from a native speaker doing the task with another native speaker). Of interest to us here is a finding that of the 42 learners in the IM (interactionally modified) group, only seven engaged in meaning negotiation. The others simply listened.” (211).

Even when acquirers do talk, they do not often make the kind of adjustments the Co hypothesis claims are useful in acquiring new forms.

• Pica (1988) concluded that instances of comprehensible output were “relatively infrequent”(45). In her study of 10 one-hour interactions between low-level ESL acquirers and native speakers (teachers), only 87 potentials instance of comprehensible output were found, that is, introductions in which the native speaker requested “confirmation, clarification, or repetition on the NNS utterance” (93). These 87 interactions contained only 44 cases in which the nonnative speaker modified his or her output (about 4 per hour), and of this 44, only 13 modifications involved grammatical form, about one per hour.

• Pica, Holliday, Lewis, and Morgenthaler (1989), intermediate ESL acquirers interacted with native speakers. Because the situation in Pica (1988) did not promote negotiation (an interview), some of the conversations in this study were in situations designed to require a negotiation and comprehensible output. Of 1952 native speaker utterances, 327 “signals indicated a clarification or confirmation of what the NNS had said” (74). In reaction to these 327 utterances, the nonnative speakers produced 116 responses containing “modified output.” in other words, they produced comprehensible output in response to about 6% of their native speakers utterances (116/1952).

• Interactions were also contravened to promote negotiation in Van den Braden ( 1997). 11 and 12 year old speakers and acquires of Dutch interacted with peers or with a teacher in an activity in which speakers had to describe a drawing to a partner. In pure peer-peer dyads, 51 instances of a negotiation of meaning were recorded, and of these, their speakers modify their output 20 times period. In peer- teacher dyads, there were 49 instances of negotiation of meaning and 20 instances of alteration of output. In both cases, this amounted to about one every 5 minutes. We do not know if the alteration improved the grammatical accuracy of the output; We are told only that “these modifications often involved, or included, formal modifications…” ( 616). Even if every case resulted in improvement, however, this data in confirms that even in contrived situations, comprehensible output is infrequent.

Lister and ranta( 1997 ) recorded 18. 3 hours of French immersion language arts and subject matter lessons involved four and 5th graders. The lesson contained a total of 3, 268 student turns, of which about 1/3 ( 1,104 or 34%) had at least one error. While teachers provided some kind of feedback to 62% of these errors, only 73 were in the form of a clarification request, “a feedback type that can refer to problem in either comprehensibility or accuracy, or both” (47). Of this 73, 20 were followed by student repair, or correction. This amounts to about one per hour, a result very close to that reported by Pika (1988) for conversations.

The situation in writing is similar. Cumming (1990) examined the think-aloud protocols of second-language writers, hypothesizing that instances in which writer appeared to be attending to both form and meaning at the time are potential instances for language acquisition, according to comprehensible output hypothesis. Only 30% of their verbal reports made by the writers in his sample were of this kind (490). In addition, the nature of the episodes makes it unlikely that they play a major role in language acquisition: in most of the episodes writers were searching for the right word or searching for first language equivalents. The latter is a familiar strategy of falling back on the first language when competence is lacking in the second language (Newmark 1966).

In Swain and Lapkin (1995), grade 8 early-French-immersion subjects were asked to write a short essay in French ( one to two paragraphs) and then to edit it, and “to think aloud” as they made a decisions. For the draft and editing stage combine d, there were open “190 occasions in which students consciously recognize a linguistic problem as a result of producing, or trying to produce, the target language” (384). This amounts to an average of 10.6 occasions to students. If students wrote a short essay every day (which they do not), this would mean about 10 chances to improve through writing per day - not very much. As was the case with Cumming (1990) study, many of the decision were lexical (looking for the right word), not grammatical ( 50% in the first draft). In addition, Swain and Lapkin note that there was no evidence that any of the episodes they described led to improvement.

Acquisition without output

There in numerous studies that confirm that we can develop extremely high level of language and literacy competence without any language production at all(Krashen 1994a). Laboratory studies show that subjects typically acquire small but significant amounts of new vocabulary knowledge from a single exposure to an unfamiliar word in a comprehensible text (Nagy, Herman, and Anderson 1985 ), enough to account for expected vocabulary growth, and similar results have been reported for the second language development (Pitts, White, and Krashen 1989; Day, Omura, and Hiramatsu 1991; Dupuy and Krashen 1993). It has been argued that a similar effect exists for spelling (1989). In addition, there are case histories of those who have developed very high level of competence from input alone (Richard Boydell suffered from cerebral palsy and acquired language through listening and reading alone [see Krashen 1985]; Malcolm X and Richard Wright credit the literacy development to wide reading, discussed in Krashen, 1993b).

Ellis (1995) is an additional analysis of Ellis, Tanaka, and Yamazaki (1994), discussed above, and provides another instance of acquisition without output. The “premodified” group, a group that did not speaking at all, made modest but clear gains in vocabulary, gaining, in fact, more words per minute than the group that interacted with the native speaker.

Does CO lead to language acquisition?

Nobuyoshi ans Ellis (1993) claim to have provided data showing that comprehensible output results in actual improvement. In their study, 6 adults EFL students in Japan of “fairly low-level proficiency” but “capable of using at least some past tense verb forms correctly” (206) were asked to participate in a jigsaw task with their teachers in the which they described actions in picture that, they were told, occured the previous weekend or previous day. During the first session of the study, the three experimental subjects received request for clarification if there verbs was not in the past tense or if the past tense was incorrectly formed. During the second session, one week later, they received only a general request for the clarification (when the teacher did not understand). The three comparison subjects received only general requests for clarification each time.

Nobuyoshi and Ellis report that comparison subjects that did not improve their past tense accuracy. Two experimental subjects (E1 and E2) were able to improve their performance in response to requests for clarification at the first session, but the third experimental subject (E3) did not. Nobuyoshi and Ellis claim that E1 and E2 sustained the gains to session 2 with E1 increasing accuracy from an original level of 31% to 89% and E2 increasing from an original 45% to 62%. Nobuyoshi and Ellis conclude that their study “provides some support for them claim that“pushing” the learners to improve the accuracy of their production result not only in immediate improved performance but also in gain in accuracy over time” (208).

As Nobuyoshi and Ellis point out, however, their conclusions are based on a very small sample size. In addition, they are based on very low number of obligatory occasions. E1, who showed that clearest gains, went from 4 correct out of 13 at session 1 to 8 correct out of 9 at session 2. E1’s wend from 9 correct out of 20 a session 1 to 16 correct out of 26 at session 2. E1’s gains are statistically significant (Fisher exact Test, two-tail, p = .0115) but E2’s gains are not (chi square = . 69). Thus, for one subject there was no evidence of the value of comprehensible output (E3), and for another, gains were not statistically significant. Data supporting a central hypothesis should be made of sterner stuff.

Note also that all three subject had started the past tense rule, and had been clearly focused on it in session 1. It is reasonable to expect that when subjects are focused on form, then put back in the same environment, they will be focused on form again, especially if the conventional partner is their teacher. The significant effect on E1, in other words, may have been a performance effect - E1 was simply more inclined to try to use a consciously learned rule for the past tense and was a more successful Monitor user than E2 and E3.

Van den Branden’s subjects (Van den Braden 1997, discussed earlier) who participated in sessions that encouraged negotiation of meaning increased their output relative to control group that did not engage in interaction but was not superior in grammatical accuracy. Each subject, however, had only 7 to 9 minutes of interaction.

Taron and Liu (1995) suggest that CO might have played a role in the second language development of a Liu’s subject, “Bob”. Bob was recorded interacting with peers, with teachers, and within an “adult- friend” (Liu). Tarone and Liu note that language use was much more complex in the latter interactions, and, in general,“new structures that appear first in interactions between Bob and their researcher, spread to the interactions with his peers, and appeared last in his interaction with his teacher” (119). They know that it is likely that Leo provided Bob with more complex input, but also suggest that Bob’s attempts to produce CO in interacting with Leo played a role. While interacting with Liu, Bob used English in a much wider range of speech acts than in the other situations, and this may have pushed Bob to “go beyond the limits of his interlanguage competence in production” (121). Taron and Liu show that the CO hypothesis, as well as the input hypothesis, is consistent with the what is known about the Bob’s development. As they note (123), data is lacking on the frequency of CO, which prevents us from determining whether CO resulted in language development and whether Bob produced a significant quantities of CO.

The discomfort of CO

The CO hypothesis predicts that we acquire language when there is a communicative breakdown and we are “pushed to use alternative means to get across… the message… precisely, coherently, and appropriately” (Swain 1985, 248-49). In addition to the research that shows that CO is an unlikely candidate, there is a additional evidence that “pushing” students to speak is unpleasant for them. When asked what aspects of foreign language classes are the most exciting provoking, students put “talking” at the top of the list“ (Young 1990). Loughrin Sacco (1992) reported that for students in the beginning French classes,“for nearly every student… Speaking was the highest anxiety-causing activity” (314).

Ten “anxious” foreign language students interviewed by Price ( 1991) stated that the greatest source of anxiety “was having to speak the target language in front of the peers”(313). Of great interest here is the finding that another source of stress “was the frustration of not being able to communicate effectively” (105).

These results suggest that it is a “pushed output”, having to utilize structures they have not acquired, under demanding conditions, that students find uncomfortable. Methods based on the comprehensible output put students in this situation constantly.

CO and the interaction hypothesis

The Co hypothesis is linked to what is sometimes called the interaction hypothesis, the hypothesis that we acquire language from interacting with others. As stated in this way, the interaction hypothesis is vague— is interaction necessary or just helpful? Is it the only way to acquire language or one way to acquire languages? Also, what occurs during interaction that causes language acquisition?

I have argued that a part of interaction that does not contribute to language acquisition if the output produced by the language acquirer. In addition, there is evidence that a strong version of the interaction hypothesis, one that asserts that interaction is necessary before language acquisition, is not correct. Such a hypothesis denies that acquisition can occur from reading and listening. In addition to their massive data showing that reading can promote language development (e.g., Chapter 2, this volume), the results of Ellis, Tanaka, and Yamazaki (1994), discussed above, confirm that acquisition is possible without actually participating in the interaction. A weaker version of the interaction hypothesis is that interaction can be a good source of a comprehensible input (Krashen 1982).

The Need Hypothesis

the Co hypothesis is closely related to the need hypothesis. I have never seen the need hypothesis discussed explicitly in print, but it is widely assumed to be true. The need hypothesis says that we acquire language only when we “need” to communicate, when we need to make ourselves understood. If this hypothesis is correct, language acquirers must be forced to speak the second language. The need hypothesis thus implies that submersion is a good thing, in that it forces students to try to communicate.

The need hypothesis is not correct. An excellent counterargument was presented by Garrison Keillor on the Prairie Home Companion, in a segment called “The Minnesota language school”. The Minnesota language school operates on the assumption that we acquire language when we need to use it. Their method is to take someone who speaks no German at all, fly them up in the helicopter, and threaten to push them out if they don’t start speaking German immediately. If they need hypothesis were correct, this would work.

According to the input hypothesis need can be helpful when it puts the acquirers in a position to get comprehensible input. All the need in the world, however, not result in language acquisition if there’s no comprehensible input. In addition, interesting and comprehensible input will result in language acquisition whether a need is present or not.

Summary and conclusion on comprehensible output

The comprehensible output hypothesis has numerous difficulties:

• Output and especially comprehensible output is to scare to make a real contribution to linguistic competence.

• High levels of linguistic competence are possible without output.

• There is no direct evidence that comprehensible output leads to language acquisition.

In addition, there is some evidence that suggests that students do not enjoy being pushed to speak.

The original impetus for the comprehensible output hypothesis was the observation that students in French immersion, despite years of input, were not so as good as observer felt they should be in grammatical aspects of their second language (Swain 1985 ). Input, it was suggested, was therefore not enough. It can be argued however that we have not yet given comprehensible input a real chance. We have it to see how students will do if their classes are filled with comprehensible input, if they have access to a great deal of very interesting reading and listening materials (films, tapes), and if the acquisition situation is generally free of anxiety. (There is evidence that children in French immersion do very little pleasure reading in their second language; Romney, Romney and Menzie s(1995) reported that the French immersion students they studied spend “… an average of 3 1/2 minutes a day reading French books and one minute reading French comics, magazines and newspapers…” (485). By comparison, they averaged 26 minutes per day reading English books and seven minutes reading English language comics, magazines, and newspapers. When asked to name their favorite French author, only 3% of the students could name an author; In contrast, 81% were able to name their favorite English author.)

Given the consistent evidence for comprehensible input(Krashen 1994a) and the failure of other means of developing language competence, providing more comprehensible input seems to be a reasonable strategy than increasing output.

Notes

1 The comparison-group teacher promoted a focus on form, frequently correcting student`s use question forms, but students in the experimental classes produced more questions and had more feedback. Spada in Lightbown point out that the comparison teacher might have emphasized form more in the mouth preceding the treatment, which in their view explains why this group also did well on the delayed posttests. Comparison students, however, also heard far more questions ( 1993, 214, Table 3 ). Clearly, this is one of the study does not help us decide among competing hypotheses.

1. The discussion here is of durability of the effect. Interestingly, Norris and Ortega reported a slightly smaller effect size for studies with a longer treatment, measuring treatment length in terms of hours devoted to the instruction. As they note, factors such as a complexity of the target rule and intensity of instruction were not considered in the analysis.

2. As noted early only 7 of the 42 subjects in the interaction group actually spoke: Ellis, Tanaka, in Yamasaki (1994) found, however, that these 7 “did not enjoy a clear advantage in either comprehension or vocabulary over those who just listening” (212).

3. Kitajima (2001) claimed that output-based activities are superior to input based activities for learning vocabulary. In her study, 5 students of the 5th semester Japanese at the university level were exposed to 19 verbs in two input based sessions of 25 minutes duration each, and to 20 verbs in two sessions combined input and output (termed output here). Words used in the two conditions were different but were considered to be of equal difficulty. In the input session, students heard the target words in a meaningful context, and were then given “a simple explanation” of the word (475), which was followed by a question in which the target words were used. Students were not required to use the target word in the answers. In their output sessions, that presentations and explanation were the same is in the input session, but they were followed by activities in which students had to produce the target words in a small group discussions with other students and in class presentations. There was no difference in retention of a words in a test given immediately after their presentation, but words presented in output activities were retained better on tests given one month and two months later. (No statistical tests were performed because of the small sample size.)

Does this result provide a counterevidence to the input hypothesis and evidence in their favor of some kind of output hypothesis? No. The input hypothesis deals with a subconscious language acquisition, not conscious language learning. Very little acquisition was involved in this study. It was, rather, a comparison of the impact of different amounts of conscious practice of previously consciously learned material on retention. Consider the following points:

1. Subjects were given the definitions of the words, and were thus not required to acquire the meaning of the words from context. The study was, therefore, no study of quisition or new learning, but the study of retention of consciously learned vocabulary.

2. The input condition was a very weak acquisition environment: while some comprehensible input was present, students were clearly focused on the form. As noted above, they were told what the target words meant and you that they had to try to remember them. Words were initially presented in comprehensible way, with context, but were not part of the interesting text or story. Rather, they were presented in isolated and contrived contexts. Here is an example of the kind of question used in the input condition and in the first part of the output condition.

For example, to introduce the target word “be successful,” the instructor showed the video scene of the computer room at the university and asked students questions, for example about who became successful in the use of computers, in what field would they be hoping to succeed, and what skills and knowledge would be required for success in those fields.” (473)

Each target word received a different scenario. It’s likely that students would not be particularly fascinated by such contexts; They would not be“lost in this story” but would be fully aware that they are practicing a new vocabulary words.

1. There was clear focus on the form in both conditions, but students were more intensively focused on the form and the output condition than in input condition. In both conditions, subjects were aware of all times that they were learning vocabulary. In addition, subjects in the output condition were told that they would be required in actually produce the target items at the end of session. During the output sessions, target words were available on the blackboard and students were required to use those words.

2. The conditions for Monitor use ( 1982; Chapter 1, this volume) were present for all tests given. The tests are clearly focused their students on form, students had been told the meaning of the words, and the test allowed sufficient time for students to access their conscious knowledge; In no case subjects forced to produce the words under the time constraints of ordinary conversation.

3. Subjects were all the university students in their 5th semester of study of Japanese. It is thus safe to say that they were all highly successful “learners”, familiar with conscious learning of language, and a accustomed to it. In fact, Kitajima notes that the less diligent students in the class were not utilized in the study, those who did not attended regularly or who dropped the course while their study was taking place (475).

Kitjima ‘s study is useful and interesting, but has little to do with language acquisition and therefore little to do with the input hypothesis. It is a study of the effect of different amounts of form-focused practice on the previously learned (not acquired) vocabulary. It shows that a highly trained learner remember consciously learned knowledge better when they are more intensively focused on form than when they are less intensely focused on form. It does not show that learning is better than acquisition or that output is better than input.

How to become smarter

How reading and writing make you smarter or, how smart people read and write

How we get smart

To explain how reading and writing make you smarter, I first have to discuss how we get smart. To do this, I present a model derived leisurely from Graham Wallas “The art of thought”, published in 19 26, and frank Smith comprehension and learning (1975). According to this model, we go through the following stage in thinking and creating new ideas:

1. Preparation: in order to come up with new ideas, we have to prepare, or clarify, our current ideas and the problem we are working on. Wallas (1926) states, our mind says not likely to give us a clear answer to any particular problem unless we set it a clear question” (44). Elbow (1973) maybe referring to the same stage when he discusses“wrestling with ideas” ( 129) and “perception of a major mass close Cortana ( 131).

2. Incubation: In this stage, the mind goes about solving the problem. Elbow ( 1973, 1981) refers to this as a“cooking”. Incubation accuracy subconsciously and automatically. When given a clearly state problem, we involuntarily attempt to solve it.

3. Illumination : elimination is the emergence of a new idea, the result of incubation period it is often perceived by the thinker as a sudden insight (“Eureka”).

4. Verification: ideas that emerge from the incubation stage are fragile and easily forgotten. To enter long term memory, they need to be confirmed or verified. 1 conjecture is that this happens when the thinker notes that he or she has arrived at the same conclusion from a different source or when he or she discovers that someone else has the same idea.

Wallas (1926) points out that the five stages can overlap: a psychologist watching an experiment, or a business man go and throw this morning’so letters, may, at the same time, be“incubating” on a problem which he proposed to himself a few days ago, be accumulating a knowledge in “preparation” for second problem, and be “verifying” his conclusion on a third problem. Even in exploring the same problem, the mind may be unconsciously employed in preparing or verifying another aspect.(42)

a very exciting hepatitis is that the process our planet here is a gateway to long term memory and the development of new cognitive structures. In other words, we learned by solving problems, and not by deliberate study.

Some Notes on Incubation

The incubation stage is of particular interest, because it appears to require that we do nothing when we take a break from creative work, and the mind is relaxed, not focused on the problem. Wallas (1926) note that “in the case of more difficult forms of creative thought… It is desirable that not only that there is should be an interval free from conscious thought on the particular problem concerned, but also that that intervals should be so spent that nothing should interfere with the free working unconscious process of the mind. In those cases, the stage of incubation should include a large amount of actual relaxation” ( 95).

Wallas reports that he first heard about the idea of incubation from the physicist Helmholtz. In speech delivered in 1891, Helmholtz described how new thoughts came to him: after previous investigation, “in all direction… happy ideas come unexpectedly without effort, like an inspiration… They have never come to me when my mind was fatigued, or when I was at my working table… They came particularly readily during the slow ascent of a wooded hill on a sunny day” (91).

Einstein clearly knew about incubation: according to Clark( 1971), Einstein would “allow the subconscious to solve particularly tricky problem. Whenever he felt that he had come to the end of the road or in a difficult situation in his work, her eldest son said, “he would take refuge in music, and that would resolve all his difficulties.” (106). Clark notes that for Einstein, “with relaxation, there would be often come the solution” (106).

Cskszentmihalyi and Sawyer (1995) interviewed nine “creative” individuals, all of whom had made a creative contribution in their field, were 60 or older, and were still actively involved in creative work. All mentioned that inside security during idle time, and Sevilla mentioned that the way accurate while we were doing something else, during “repetitive, physical activity” such as gardening, shaving, taking a walk, or taking a bath (348).

Mind On, Mind off

This is not to say, of course, that hard work is unnecessary. Quite the opposite is true. Many studies have confirmed that high achievers put in a tremendous amount of work, far more than a less accomplished colleagues( Simonton 1988). The hard work of thinking and problem solving comes in the preparation and verification stage, not in the incubation stage. Wallas suggests that educated person knows how to prepare and then allow incubation to happen: the educated person “can put his ‘mind on’ to a chosen subject, and ‘turn his mind off’.(92) Also, the “elimination” that is the result of the incubation needs to be followed by more conscious work. Ideas that arise as a result of incubation need to be evaluated (Smith 1994); our new insight may not be right.

Scheduling incubation time and not being ashamed of it

Some of Cskszentmihalyi and Sawyer (1995) subject actually scheduled“a period of solitary idle time that follows a period of hard work… Many of them told us that without this solitary, quiet time, they would never have their most important idea”( 347). One respondent actually began with his interview with this statement:“I am fooling around not doing anything, which probably means this is a creative period … I think that people who keep themselves busy all the time are generally not creative, so I am not ashamed of being idle”( 352).

the research on learning by problem solving

There is a formal and informal evidence supporting the hypothesis that we learn through solving problems. I begin with the informal evidence because, in my opinion, it is much more convincing.

Thr Informal Evidence

The Fox Hills Mall

it has been said that if Americans are not at home or at work, they thought most likely place you will find them is in a shopping mall. I live near the fox hills mall in the cooler city, California, for many years, and my experiences in this small land support to the hypothesis that we learn by solving problems and not by“study”. After some reflection, I have come to the conclusion that I probably know about 1000 facts about the fox hill mall ( and I’m sure you know 1000 facts about your shopping mall). I want list them all, just enough to make the point:

1. I know where the Fox Hills mall is ( corner of Slausen and Sepulveda, underneath a very short freeway).

2. I know where to park at the fox hills mall. There are at least twenty options for the parking, and each option has its own consequences.

3. I know where the telephone are(the ones in the center of the mall are usually broken or in use; I recommend the ones in my company), and I know where the bathrooms are.

4. I know a great deal about some of the stores in the mall, and practically nothing about the others. Of course, I know about the stores I have shopped in. I know, for example, if LensCrafters will really give you a new pair of glasses in one hour.(Actually, yes.)

how did I learn all this? Where did I get all this detail knowledge? I never studied. The manager of the mall does not give shoppers a add manual describing the mall and require them to get at least 80% correct on a test before they are allowed to shop. I got my knowledge of the mall the same way you learned about your mall - by finding a telephone, by buying things… By solving problems.

This is clearly the way all experts gain their detailed knowledge of their fields. Linus Pauling, I’m sure, did not master the entire field of chemistry by studying flash cards. As Frank Smith has pointed out (Smith 1988), the “laws of learning” are irrelevant when we are involved in the real problem solving. The man proposes marriage to the woman. He doesn’t ask her, 5 minutes later, what your answer was, claiming he forgot. When the information solves a problem, when it’s relevant, one repetition is often enough.

The formal evidence: breaking the intentional learning barrier

experimental evidence for the hypothesis that we learn by problem solving comes from studies of “incidental learning”. Hayden Jenkins( 1969 ) presented subjects with written words that were flashed for a brief moment, not long enough for the subjects to examine the words in detail. One group of subjects were asked to estimate the number of letters in the world(the current group). A second group was asked to determine tour weather the letter E was in the word (the current group). A second group was asked to determine whether the letter E was in the word open the search group). 1/3 group was asked to rate the words after that open quarter pleasantness” (people would probably rate 3 as more pleasant than tire). Hyde and Jenkins then surprised their subjects by asking them to recall as many of the words as they could. The pleasantness group remember the most words. The pleasances group also did just as well as four group that deliberately tried to remember the words. Thus,“ incidental” learning was shown to be just as effective as “intentional” learning, if the problem that the incidental learner are trying to solve is interesting enough.

Wilson and Bransford (reported in Bransford 1979) did a similar study, but added another condition, the “desert island” condition: They asked subjects to rate how important the objects denoted by the presented words (nouns) would be on a desert island. The desert islands subjects remembered the words better than a group that deliberately studied.

Wilson and Bransford’s results are very important; they show that incidental learning can be more effective than intentional learning. In other words, they break their intentional learning barrier. In my opinion, it is very easy to break the intention of learning barrier. Many things we do in everyday life, many problems we solve ( such as shopping in the fox hills mall) are more interesting than the desert island condition in Wilson and Bansford’s study.

We now turn to the main point column how reading and writing make you smarter period to reveal the punch line early, I’m going to claim that to at least some extent,“smart people” are people who have learned to read and write in ways that are consistent with their thinking process presented earlier. They are user reading and writing, in other words, to solve problem period and in order to do this, they have had to overcome the lessons they learned in school.

Reading and cognitive development

There is a little doubt that reading influences cognitive development common but it’s surprisingly difficult to find direct evidence. Ravage and theme( 1987 ), in their study what do our 70 years old now find that those 70s years also who knew more, read more: those who lived in print richer environment did better overall on tests of history and literature, and there was a clear relationship between the amount of reported leisure reading and performance on their literature test. Stanovich and Cunningham (1992) confirmed that college students who read more did better on the same test of history and literature knowledge that Ravitch and Finn used, and this relationship held even when nonverbal ability factors were controlled.

Those who read more also do better on various measures of cultural knowledge. West and stanowicz “1991” created a cultural literacy test, a checklist of 30 names of the artists, entertainers, explorers, philosophers, and scientists. Those who had more print exposure did better on this test period even when other factors, such as cat scores (West and stanovich 1991 ), age education, exposure to television(W, sternal which, and Mitchell 1993), and non verbal abilities( stanovich, W, and Harrison 1995) well controlled. Stanovich in Stan urgent Cunningham (1993) found that those with more print exposure did better on a test of“practical knowledge” and the test of science and social studies.

Good thinkers read more

status of“good thinkers” also give us some reason to believe that reading makes you smarter. Good thinkers and, however they are defined, rate a great deal and have a read a great deal. Simonton (1988) concluded that open code omnivorous and adolescence correlates positively with ultimate adult success” (11). Schaefer and Anastasi (1968) reported that high high school students considered to be a creative read more than average students, with more creative students reporting that they read over 50 books per year. Emery and Csikszentmihalyi (1982) compared 50 men of a blue collar background who became college professors with 15 men of very similar background who grew up to become blue collar workers. Their future professors lived in a much more print rich environment and did far more and they were young.

And thus appears to be the case that good thinkers, as a group, read more than the general population does, after a certain point, however, the relationship between the amount of reading done and thinking is less clear. Goertzel (1978) started 300“evident personalities of our age” (subjects of biography in the Menlo Park library published after 1963) and reported that almost half of the group“ omnivorous readers”( 11 ). Simonton( 1984) did reanalysis of this data, however and found only a 0.12 correlation between “ achieved eminence” an amount of reading done.Van Zelst and Kerr (1951) reported a modest 0.26 correlation between number of professional journals read regularly and productivity resulted in a bimodal curve - some less productive scientists read a gread deal. Apparerntly, good thinkers do read a lot, but it is possible to overread. Wallas (1926) was aware of this, noting that .. may interfere with incubation.

Selective reading

What may maybe crucial it’s not simply reading a lot, but reading selectively - reading what you need to read to solve the problem you are currently working on. Brazerman( 985 ) provides support for this idea. Brazerman examined the reading habits of the top physicist , and reported that they read a great deal, visiting the library frequently to keep up with the current research period they distinguished, however, between open quarter” and “peripheral” reading, reading carefully only whatever was relevant to their interests at the time.

It may be the case that reading is useful to us only when it’s relevant to a problem we are working on, when it helps gather information for preparation or verification. When we read selectively to help us solve the problem, we remember what we read. When we read material that is irrelevant, we don’t remember it. This is certainly my experience. I have, it seems, nearly total recall for some articles and books I read years ago. Quite often,however, I ran across an article or book on my shelf that has my underlining it it, my notes in the margin, and I have no conscious memory whatever of having created, even if the book or journal is fairly recent. Whenever this happens, it is something I read because I felt I should read it, not something that related to a problem I was working on at the time.

School

School tells us are the opposite. School does not encourage selective reading for program solving, but tells us that all reading is a core reading, and that we should try to remember what we read. School does this by assigning a certain amount of reading for each class, and by testing us on our reading. This works against the cognitive development.

Consider what happens when you have a 25 page assignment to read in one evening. You read the first paragraph on the first page and, stimulated by what you read, it’s an idea: incubation has taken place. This is good. This is the purpose of the school - new learning. Ideally, you should stop reading and write the idea down (see the section on the role of writing below). But you have 24 and half pages left to read! If you stop, you won’t finish the assignment on time.

The problem, in other words, is that incubation and illumination occur beyond our conscious control and can happen anytime. When we have reading assignment, new ideas, instead of being welcomed, are an annoyance. Good thinkers need to overcome the lesson they learned in school.

Writing and cognitive development

writing makes its contribution in the preparation stage. When we write, we attempt to represent our cognitive structures, our current thoughts, on the page. The act of doing this is a powerful stimulus towards creating new cognitive structures, new ideas. In terms of Wallace model, writing prepares us for incubation.

Evidence suggests that certain writing activities such as not taking, summary writing, and answering comprehension questions help learning. According to ladders( 1988 ) “the preponderance of evidence strongly favors note taking” (616); students who take notes during her lectures are typically retained more than those who do not don’t. Similarly, several studies show that students who write a summaries of what they read or hear remember more than those who don’t( Doctorrow, Withrock, and Marks 1978; Bretzing and Kulhavy 1979; Pepper and Meyer (1986); studies also show that answering comprehensive question is more effective in promoting learning than require multiplechoice responses (anderson and Biddle, cited in Langer and Applebee 1987). In these studies, however, the benefit of writing is not tapped, becouse real problem solving is typically not involved.

In a series of studies, Landry and Applebee (1987 ) came closer to showing the impact of writing and thinking. The third study is, in my view, the most revealing. Ninth and 11th grades were asked to read two social studies passages, one considered easy and one considered difficult. One group simply read the passages (“read and study”), another answered comprehensive questions, another wrote a summary of each passage, and another wrote an essay that require them to “reformulate and extend” the material from the passage (104). Subjects were given a variety of tests, including “topic knowledge” test developed by Langer. The topical knowledge test was given the day after the reading and again five days later; Only the result of the delayed tests are discussed here.

Results for the easy passage appears to be contrary to a hypothesis that writing leads to more learning. Those who simply read the text but did not write (“read and study”) did just as well as those who wrote essays. Summaries, and answer to comprehension questions period but the results for the harder passage were different. On the hard passage, essay writers did the best, and the read and study group actually did worse than those who didn’t read the passage at all (control group).

This result suggests that writing especially essay writing, works best when problem solving is involved. As a Langer and Applybee conclude, “ if contact is familiar a relationship are well understood, writing may have no measure effect at all” (131 ). Even the essay written in response to the hard passage does not reveal the full power of writing, however. Subjects were given only 20 minutes to write their essay, and the topic was assigned. We would get a better picture of what writing can do if we examine real writing, done by real writers, solving real problems that are important to them. The framework presented here predicts that this kind of writing results in the exceptional learning, both of new concepts and facts.

Some evidence that appears to support this prediction comes from the studies of scientific and artistic achievement. It is a well established that good thinkers produce a great deal: “voluminous productivity is the rule and not the exception among individuals who have made a noteworthy contribution”(Barron, cited in Simonton 1988,60). Simonton (1988) provides some striking examples: “Darwin could claim 119 publications at the close of his career, Einstein 248 and psyhology Galton 227, Binet 277, James, 307, Freud, 330 and Maslow,165…” (60). Simonton also reports that correlations between total productivity and citation counts range from 0.47 to 0.76 and provides additional data showing that quality and quntity are related.

Is this also evidence that writing makes you smarter? I think it’s, but there is some problems with this interpretation. And obvious 1 is that good thinkers are typically recognized as a good thinkers early in their careers, before producing much for public view. It may be the case, however, that good thinkers wrote a great deal privately before their work was known.

Another problem is the common perception that good thinkers do their best work when they are young. Simonton (1984), however, reports that quantity of work declines only slightly with age, and quality remains constant. It may be that quality actually increase with age. Simonton suggests that early contribution may simply get more attention:

… later creative offerings may not be perceived by the scientific community to be nearly as innovative as the initial milestones, this perception may be partly an illusion of contrast. It may be precisely cause of the early efforts have revolutionized the field so thoroughly that the later works, being interpreted in a new context, may seem to lack any revolutionary quality. (99).

Simonton notes that Einstein’s general theory of relatively is “a contribution no less revolutionary than special theory” produced a 10th year earlier. But the special theory “has changed the way scientists viewed the universe, making the general theory “looks less momentous than it was” (100).

Regular writing and new ideas

Strong confirmation that writing helps thinking is the work of Robert Boice . In an extremely important study, Boice (1983) concluded that regularly scheduled writing sessions encouraged more writings and the emergence of more creative ideas than did a “spontaneous” writing (writing when the writer “felt like it”). Boys ask the college professors to write under several different conditions: not to write for several weeks (control group), to write only when they felt like it, or to write regularly at scheduled sessions each day. Subjects were asked to keep track of the number of page written and the number of creative or novel ideas that came up. Regular daily writing resulted in about double in the number of page written and doubled the number of new ideas, as compared to writing when the one felt like it. The control group reported the fewest number of new ideas.

This result is consistent with reports from professional writers who maintained writing schedules, regardless of mood(Wallace and Pier 1977 ). It is also consistent with the results of other studies by boys( 1987, 1989, 1994) showing that regular, planned writing sessions result in more writing than waiting for long blocks all the time (“binge writing”). African theme invoices work is about inspiration does not typically precede writing an experienced writers; It is the result of writing.

The Composing process

research in the language arts has revealed that better writer have mastered what is now known as a “the composing process”. These studies typically compared more advanced and successful writers with beginning and remedial writers, and have shown, among other findings, that

good writers engage in more planning before they start to write, but are willing to change their plans; Their plans, the other words, are flexible( 1971).

Good riders are more willing to revise, and focus far more on meeting when revising. Faigley and Witte (1981) reported that advanced college writers made more content revisions and delayed mechanical and words choice changes until “they had satisfied factorially dealt with their subjects” (409).

Sommers (1980) has confirmed that experienced writers understand that the early draft are tentative, and that as they go from draft to draft they come up with new ideas. Average and remedial writers don’t know this but think that all of the ideas are in their outline or first draft, and regard revision as simply making a neater version of the first draft.

The composing process works because it is in harmony with the way the brain naturally learns. Writers who use their composing process use writing for preparation by planning, and they respect the incubation process; When new ideas arise through writing they are willing (and eager) to change their plan and revise. Writers who have no mastered the composing process have a much harder time coming up with new ideas. And unwillingness to revise means an unwillingness to accept new ideas while writing. Excessive focus on mechanic and other aspects of editing while writing pulls attention away from meaning and may be sources of writer’s block (Rose 1984).

Incubation and the composing process

own secret to coming up with good ideas through writing is, I suspect, understanding the importance of incubation period for many writers, good writing cannot be rushed. Forcing writers to sit without break and write nonstop denies that the possibility of incubation: as Smith (1994) notes, “composition is not enchanced by grim determination” (131).

Blocked and fearful writers may be under the false impression that writing should always flow, and that hesitations are a sign of incompetence. Hesitations, however, are not usually true writing blocks. They may simply be signs that a problem has come up, and taking a break may help the subconscious solve the problem period this happens to me probably 100 times a day: the problem with the word choice, a discovery that I have contradicted myself, a vague malaise that the argument are not the right order, and so forth. At least half time, a very short break, even 2 minutes or less, is enough to solve the problem period and to solve the problem often means new learning, and deeper understanding.(Note: I took five breaks in writing this paragraph, during which time I filled some papers, took some vitamins, and checked my e-mail).

Many professionals have recognized this: Irving Stone (cited in Winokur 1990) noted that“when I have trouble writing, I step outside my studio into the garden and pull weeds until my mind clears…” ( 325). Richard Condon (cited in Murray 1990) says“ I’ve never been blocked, but there are times when the words won’t come. When I feel dry-up, I deal myself a few games of solitary…”(72).

Writing an illumination

Thus far, we have been discussing the role of writing in preparation. Wallas (1926) points out that writing is also valuable after incubation, when illumination has occurred. When a new idea first occure to a thinker, before it is verified, it’s fragile, it need to be recorded: “in modern life, the range of observations and memory which may start a new thought-train is soft that is almost incredibly easy to forget some thought and never again pick up the train that led to it. The story be true of the man who had so brilliant an idea that he went into that his garden to thank God for it, found on rising to his knees that he had forgotten it, and never recalled it” (86). Smart people carry writing material with them.

School

School teachers are the opposite. School teachers ask that you write to display what we already know, not to discover a new ideas. In class assess and assess exams that needs to be done, start to finish, in one class period, actually penalize students for coming up with new ideas while writing, and certainly allow no time for incubation.

Recall your history class in high school. Your sit down exam question is to give three reason for the start of the World War Two period you think of three reasons and begin to write. Midway through your second reason, stimulated by your writing, incubation and illumination take place and you think of three better, more valid reasons for the start of World War Two period you look at the clock, however, and see that you only have the 10 minutes left. You have to suppress the new ideas and finish writing out of original three reasons, or you will fail the exam completely.

This kind of thing happens in school not once, but thousands of times, and students learn that writing functions mainly to show what they already know. Once again, good thinkers need to overcome the lesson they learned in school.

Oral Language

My focus here is on reading and writing, but I don’t mean to discount the value of oral language for problem solving. There is a good reason to believe that discussion can solve problems or solving very well discussion for preparation as Elbow (1973) points out:

If you are stuck writing or trying to figure something out, there is nothing better than finding one person, or more, to talk to. If they don’t agree or have trouble understanding, so much better so long as their minds are not closed. This explains what happens to me and many others countless times; I write a paper, it’s not very good; I discuss it with someone; After 15 minutes of back and forth I say something in response to a question or argument of his and he says, “But why didn’t you say that? That’s good. That’s clear”. (49)

Conclusion

I have argued elsewhere(chapter 2 ) that reading is the primary source of our competence in writing style and grammar, as well as a vocabulary and spelling (see also croissant 1993 b ). Figure 4-1 attempts to combine these hypothesis with the hypothesis presented in this chapter.

What remains to be discussed is what goes to the left of Reading and Writing in figure 4-1: What do we do in class? Smith (1988) suggests an answer: “Enterprises”. Enterprises are problem - real, not realistic, problems that students genuinely want to solve, programs that naturally entail reading, writing, and discussion. Finding the right enterprises is, in my view, a major goal of the teaching profession. Enterprises may include a chemistry class project in which students analyzes the water in their community and publishes result in the local newspaper, writing a history of community that will become the official history and be on public record, running a small business(and keeping the profits), and writing book reviews that remain in the school library permanently for future student use, rather than writing book reports.

Figure 4-2 expand figure 4-1 , editing enterprise and free reading. An important characteristic of Figure 4 dash is that the arrow goes from left to right, and not from right to left. As a Smith ( 1988 ) has point out, we have a confused cause and effect in the education period we don’t learn parts of language and “facts” so that we can eventually read and work on real problems. We read for interest and pleasure and we engage in problem solving, real enterprises; language acquisition, literacy development, and intellectual development occurs as a result.

Obstacles

If we are going to use enterprises in school, we face tremendous obstacles. The major obstacles is that many people, including professionals in education as well as the public, have a view of learning that is quite different from the one presented here. Danskin and Burnett’s article, “Study Techniques of those superior students”, illustrated the point of view of some professionals. Danskin and Burnett (1952) analyzed the study habits of thirty-seven university students. Every students in their sample was an excellent student, ranking in the top 12 of their class.

Danskin and Burnett were quite disappointed with these students’ study habits. Contrary to what is advised in study skill courses and books, 81% of the students waited until the last minute to study before tests, and only 8% attempted to predict that questions would be on the tests. Why am I need study skill books are recommended that students study in the heart chair, 41% of the sample said they started in an easy chair or in the bed, in 14% said they didn’t care where they studied. While many study-skill books recommend that students carefully schedule their time, 48% of a sample said that they studied when they had the time and only 11% said they had a schedule. In the face of this contradictory data, danskin and Barnett, undaunted, concluded that these students, even though they were successful, could use a good course in study skills!

According to the framework presented here, Danskin and Barnett’s results confirm that “study skills” are not crucial. Far more important then how students schedule their day, when they study, and what kind of chair they sit in, is that they focus on mentally- weather or not they are involved in real problem solving. This was confirmed by bloom ( 1963 ), who studied for my graduated students at the University of Chicago who finished the PhD and went on that successful research careers. Bloom reported that one characteristic of the successful students was on involvement this problem solving during their Graduate School careers, “preoccupation this problem rather than the subject matter of courses… The relatively complete acceptance of the role of research worker and scholar (rather than the roles of student)…”( 257- 58 ).

The public seems to equate reading and writing with study, not problem solving. I see evidence of this all the time. I do a fair amount of reading and writing in public, on airplanes, while in offices, and so forth. Occasionally a friendly person will see me working, or just see my books and notebook, and ask, “are you studying for the test?”

Notes

1. Incubation sometimes require a very long break: Richard Feynman noted that “you have to do 6 months of very hard work first and get all the components bumping around in your head, and then you have to be idle for a couple of weeks, and then - ping - it suddenly falls in place…” (Csikszentmihalyi and Sawyer 1995,350). Incubation can also occur with break of shorter duration. Piaget told Gruber (1995) but after he worked for a few hours open court he would go for a walk, not think about very much, and when he went back to his desk his ideas would be Cleary. Call Scott( 526 ). And it can also happen in very short breaks, a few minutes or even moments. In my experience extremely short breaks are all that is necessary to solve many problems and losing many blocks. In agreement with Wallas, I have a phone that breaks rock best when they are devoted to something fairly mindless: washing just a few dishes, feeding just a few papers, or doing something light exercise.

2. Experimental evidence also suggests that problem solving is more important for learning than additional efforts or hard work ( Walsh and Jenkins, 1973), then an additional time to own task ( Craik and Tulving 1975), then additional exposures, all the repetitions of stimulus ( BOBROW and BOWER 1969), and is more important than increasing her rewards ( CRAIK and TULVING 1975).

3. STANOVICH and colleagues used interesting measures of print exposure, the other recognition test and the magazine recognition test, which consists simply of checklists of authors and magazines titles. Many studies have confirmed that this simple measures are an excellent means of estimate print exposure. For a detailed discussion, see chapter 2 of this volume.

4. CTANOVICH and CUMMINGH AM( 1993) is an especially interesting study. Subject row 268 N American undergraduates who filled out questionnaires related to print exposure(Arthur, magazines and printer recognition tests) and TV exposure and who completed several tests of general knowledge. General knowledge tests included a test of knowledge of science and social studies, and acronym test(items included NATO, EEC, UHF), and checklists of names of the well known individuals in a variety of fields.

Performance on the printer composite correlated highly with the performance on the composite tests of general knowledge ( r = .85), but exposure to TV did not ( r = -. 05). Moreover, the relationship between print exposure and general knowledge help even when controlled for measures of “general cognitive ability” (nonverbal cognitive ability[ Raven MITRICES], reading comprehension, grade point average on and TV exposure.)

STANOVICH and CUNNINGHAM also compared to the performance of those who scored high on test of reading comprehension but read little, and those who scored lower on a test of reading comprehension but read more. The latter group did better on the test of general knowledge, indicated that amount of reading count more than a reading ability in terms of gaining information.

An analysis of performance on certain items of the practical knowledge test was particularly revealing. Among the questions asked was “at the time of Normandy invasion( 1944), what country was germany’s primary ally?”. Those who were the readers scored higher than those who were not, and TV watching made no difference. In fact, among “law print/ high TV” group, 28% thought that the Soviet Union was germany’s primary ally at that time. Below I also present the results for the question, “name a country in which lighting is currently the primary language”. The investigators scored this liberally, accepting “Rome,“the Vatican”, and “none”. Once again, readers did best and TV had no impact. The overall scores of this college students are distressing, however.

1. GLUECK and JAUCH( 1975 ) provide evidence that suggests that thinkers read primarily for verification and not for preparation OK. They found that a productive scientist did get some ideas from professional journals, but rely more on their own ideas and previous as input for their thinking.

2. Results of think-aloud protocol analysis done this 8 subjects revealed that those who answered comprehension questions simply “searched the passage for the correct response, copied it… and never rethought that response or returned to it to change an answer…” (221). Summary writers searched for more relationship than did those who answered a comprehension questions, but tended to maintain the temporal order of the text in the summary. Essay writers, however, used the text “to corroborate rather than find the ideas they wanted to write about” (121). Thus using writing for their verification stage of Wallas’ model.

3. BOICE also included a fourth condition: writers agreed to produce 3 pages per session, and a strong motivator was instituted: if subjects did not meet the goal on any day, they would have to donate $ 15 to “despised organization”. This group wrote the most, and produced the most ideas, but was the least efficient, according to my analysis ( KRASHEN, 2002): they produced their fewest ideas per page, and only barely made the quarter, averaging just over the three pages per day. Writing and all other conditions resulted in about three-fourths of an idea per page, but those are in the “forceed writing” produced about half that. It is possible that this group found writing to be aversive; They wrote just enough to avoid a punishment. It would be interesting to see how much writers in such condition continue to write after the negative consequence are removed, and to see whether the quality of their ideas was similar to those produced by the other groups.

Conclusion

In the first 3 chapters of this volume, I reviewed studies dealing with the effect of a comprehensible input based methodology (Chapter 1 ), the effect of free voluntary reading (Chapter 2 ), and the effect of grammar study (Chapter 3 ). Table 5 one summarize the result of studies.

The data we have so far are reasonably consistent with the generalization. When we compare comprehensible input based methods with traditional, form based methods, and test communicative, students in comprehensible input classes clearly do better. This is a supported by Asher’s studies (1972,1977) and by Isik (2000), reviewed in Chapter 1. When the test is a grammar test, there is no difference, or comprehensible input students who do somewhat better. This is supported by Hammond ( 1988 ) and WINITZ (1996). Students in NIKOLOV and KRASHEN (1997) who had a comprehensible input, however, did only slightly better than the traditional students, but ISIK’s students did far better on all types of tests.

The generalization proposed here parallel results in the beginning literacy development in English as a first language. I argued in a KRASHEN( 1999 ) that students in whole language classes, when compared to those in classes in phases skill building, did better on tests of reading comprehension, and just as well on tests of uh“skills” (reading nonsense words).

Intermediate-level students in free-reading programs appear to outperform comparisons on all kinds of tests. Elley’s studies allow us to compare the impact of a free reading on different tests. In ELLEY and MANGUBHAI (1983), the impact of sustained silence reading tended to be larger on the tests of reading comprehension than on tests of “English structures” (see ELLEY 1991, 387, table 2 ), but in ELLEY( 1991, 395, Table 3 ) the effects were similar.

Students who undergo intensive direct grammar instruction on a limited set of rules outperform those with little or no grammar instruction on test of grammar, but the effect is modest. The monitor hypothesis predicts no impact on grammar instruction on communicative tests, tests in which the conditions of for monitor use are not met. I have concluded that we still lack data for this kind of tests, as a student taking the open quarter communicator for close quarter tests used and published studies were, in fact, focused on the form and could apply the conscious knowledge of grammar when taking the tests. The impact of the grammar instruction on these tests, however, appears to be less than on tests that are more obviously open what grammar” tests and that encourage more focus on form.

I have devoted a great deal of space in this volume to criticize studies interpreted as supporting grammar study and relatively small percentage of space to studies supporting the comprehensive hypothesis. One reason for that is that I have devoted so much space in previous publication to the latter. But another reason is because the professional has spent so much energy trying to show that grammar study works. I have presented only some of the efforts here; There are many more, but they suffer from the same flaws as those discussed here. The professional has, in my view backed the wrong horse; paraphrase Frank Smith, it has spent a great deal of energy studying things that the brain does not do very well.

The energy devoted to futile attempts to show that grammar study is effective has diverted energy from a more worthwhile endeavor, trying to determine the true potential of compensable input. In my view, their studies are supporting comprehensible input, as impressive as they are, underestimate its power; There have been a few attempts to really see what comprehensible input can do. What would happen if we fully utilize comprehensible input, if we use methods such as TPR, natural approach, and sheltered subject matter teaching, and supplement them with super rich print environment that includes accessible and well stocked libraries filled with a compelling books, magazines, newspapers, comic books, audio tapes, and video tapes? If we do this, we my prediction is that we will see unprecedented success in our classes. We will see students who have a genuinely reached the level where they can read authentic test and can comfortably communicate with speakers of the language, students who are in a position to improve on their own.

Chapter 4 argues that we have also backed by the wrong horse and other areas of education: we have backed the hepatitis about cognitive development is the result of deliberated study, instead of the hypothesis that it is the result of problem solving. I discussed some of the implication of this for reading and writing arguing that effective uses of reading and writing aid the problem solving process. These users, however, are very different from the often conflict with that we learn in school.

We have thus confused cause and effect for both language acquisition and cognitive development according to the comprehension hypothesis(academic input the hypothesis), our ability to spell, our vocabulary, our grammatical competence, and so on, are the result of comprehensible input. According to the problem solving hypothesis, our knowledge of concepts and facts are a result of problem solving.

School makes the opposite assumption: we first learn vocabulary, learn to spell, learn grammar, and so both, and practice in contrast situation to “automatize” them. Only after this “basics” are mastered are we allowed to actually use language for real communication. Similarly, school has assumed that only after effects and concepts are mastered are allowed to engage in real problem solving, in real enterprises. As Frank Smith has noted, traditional schemes of education “represent the world turned upside down” ( Smith 1986, 75).

Clearly, tests are a major force in promoting activities about going the wrong direction. When a language test focus students on form, the effect is for teachers to teach form directly. If, for example, where final Test focuses on vocabulary, instruction will emphasize vocabulary lists. We know, however, that it is much more efficient to pick up vocabulary through reading ( NAGY, HERMAN, and Anderson 1985). In other words, those who do things in the right direction do well on tests, even on those tests that focus on form, a conclusion supported by the research reviewsd in chapters 1 and 2. (Exceptions are language tests that focus on late acquired items so that even a fairly advanced language students will typically not acquire without extensive reading and/or residence in the country).

The public has assumed that some things must first be “learned”, that a lot of hard worker is necessary before students can go on to more interesting activities. But when students engage in the real problem solving and are exposed to interesting and comprehensible input, they acquire language and learn concepts and facts much more easily. There is no need to wait. There is no need for delayed gratification.

ALFIE KOHN please recognize this: “for all our talk about motivation, I think we often fail to recognize a truth that it it is a staring us in the face: if educators are able to create the conditions under which children can become engaged with academic tasks, the acquisition of intellectual skills will probably follow” (Kohn 1993, 146).