Stephen KrashenStephen Krashen is one of my heroes. He is a linguist, researcher, education activist, and professor emeritus at the University of Southern California. I have wanted to meet him since I began studying linguistics in university, and finally had my chance at Ming Chuan University’s 2009 “Annual Conference on Applied Linguistics” in Taipei, Taiwan. He then agreed to conduct the following interview via email.

Note that this interview was originally only available to newsletter subscribers, but since I am now offering Language Mastery Insiders more than a dozen bonuses, I decided it was time for everyone to have the chance to enjoy Krashen’s unique brand of intellect and humor. Enjoy!


JF: Could you try to summarize the results of the research you have done over the last 30 years in a few sentences?

SK: Of course. We acquire language when we understand what we hear and read, when we understand what people are saying to us, not how they say it. To borrow a phrase from the Jewish philosopher Hillel, “the rest is commentary.”

JF: Can you provide some of the commentary?

SK: With pleasure. We do not acquire language by learning about it, by consciously learning rules and practicing them. Consciously learned rules have very limited functions: We use them to edit what we say and write, but this is hard to do, and sometimes they can help make input comprehensible, but this is rare.

We do not acquire language by producing it; only by understanding it. The ability to produce is the result of language acquisition, not the cause.

Language acquisition proceeds best when the input is not just comprehensible, but really interesting, even compelling; so interesting that you forget you are listening to or reading another language.

Language acquisition proceeds best when the acquirer is “open” to the input, not “on the defensive”; not anxious about performance.

Language acquisition proceeds along a predictable order that can’t be changed by instruction. Some grammatical rules, for example, are typically acquired early and others much later.

JF: If all this is true, what happens to language teaching? Doesn’t this mean the end of language classes?

SK: Not at all. In fact, the comprehension hypothesis makes life much more interesting for both teachers and students. Classes are great places to get comprehensible input. Even if you live in the country where the language is spoken, it is hard to get comprehensible input from the “outside world”, especially if you are an adult. The language you hear is too complex. The beginner can get more comprehensible input in one hour from a good language classes than from days and days in the country.

Here is an example from my own experience. After having spent about six weeks in Taiwan, on and off over six years, all I could say was “I like ice cream” and maybe four more words, and I understood nothing. Then in the summer of 2007 I took a nine- hour short course in Mandarin, taught by Linda Li, using TPRS, a very good method for providing comprehensible input for beginners. Linda made the input comprehensible in a variety of ways, including pictures, actions, and the use of the first language.
I got much more comprehensible input in the first 30 minutes in that class than I had in Taiwan during the six weeks I was there.

The comprehension hypothesis helps clarify what the goal of language classes is: Acquire enough of the language so that at least some authentic language input, input from the outside world, is comprehensible. Then the acquirer can improve without a class.

JF: I noticed that you said that language acquisition “proceeds along a predictable order” with some grammatical items acquired early and others late. This finding must be a big help in teaching – now we know when to teach which grammatical rules, right?

SK: That’s what I thought at first, but I have changed my position: I don’t think we should teach along any order. There are strong arguments against using any kind of grammatical syllabus.

First, we don’t know the natural order. We know enough to be confident that the natural order exists, but researchers have not worked out the order for every aspect of grammar.

Second, if our hidden agenda in a reading passage or discussion is the relative clause, or some other aspect of grammar, it is very hard to make the input truly interesting.

Third, we have to constantly review the target structures: Every language student knows that one set of exercises and a few paragraphs are not enough.

Finally, we don’t need to use a grammatical syllabus. In fact, it is more efficient not to have a grammatical syllabus. I have hypothesized that if we provide students with enough comprehensible input, the next structures they are ready to acquire are automatically provided and are reviewed regularly and naturally.

JF: I assume that translation is out of the question…

SK: Too much translation can interfere with delivery of comprehensible input. This is because there is a tendency to pay attention only to the translation and not the second language input.
But there are ways of using the first language to make input more comprehensible, including doing background reading or having discussions on topics that are especially complex and hard to understand in the second language. This is part of the basis for bilingual education: Providing background knowledge in the first language that makes second language input more comprehensible.

In class, the first language can also be used for quick explanation or for providing the meaning of a problematic, but crucial word. This may or may not help much with acquiring the meaning of the actual word, but will serve to make the entire discussion more comprehensible and thereby aid in acquisition of other words and grammatical rules. Linda Li did this very effectively in the Mandarin class I attended.

JF: This sounds nice for developing conversational language. But we also need to talk about what Jim Cummins has called “academic language.” That’s the real goal for many students of English today. Now that English has become an international language, many people need high levels of English literacy and knowledge of specialized vocabulary.

SK: Again, the comprehension hypothesis is a big help. It predicts, and predicts correctly, that there are several ways of developing academic language proficiency. The one I think is the most powerful is wide, self-selected reading, also known as free voluntary reading.

There is an overwhelming body of research that shows that free reading is the main source of our reading ability, our writing style, our “educated” vocabulary, much of our spelling ability and our ability to handle complex grammatical constructions, all important aspects of academic language proficiency.

A second way is through sheltered subject matter teaching, that is, making subject matter comprehensible for second language students in special classes, a form of “content-based” teaching.

Studies show that students in these classes typically make good progress in second language development and learn subject matter at the same time.

JF: One more question; a very important one. You have claimed that there is research supporting these hypotheses. But it is very hard to find the actual studies, especially these days when money is a problem for nearly everyone. How can we access the actual studies?

SK: I think the prices of technical books and journals are outrageous, and do a disservice to educators and concerned citizens. My approach is to make as much as possible available on the internet, for free.

I have my own website,, and readers of this interview are free to download, share, and cite anything on this website. I am adding articles as quickly as I can. There is already one book on the website and there will be more.

The website also has a mailing list, if people are interested in seeing short items I come across, and my own letters to the editor. I write several letters to the editor to newspapers all over the world every week. Again, readers are free to share anything from the website with others, including with their students.

We also started a free open-access internet journal a few years ago, which includes many of the research papers my colleagues and I have done, the International Journal of Language Teaching (IJFLT). Just go to and you have easy access. The journal emphasizes short, readable papers, a real contrast to the usual thing you see in some professional journals in education these days. And for those interested in the political as well as the research controversies in language education in the US today, I recommend two more websites which have been very important for me:

  •, which I regard as the center of gravity for the “resistance movement” in American education.
  •, the website of the Institute for Language and Education, a new organization dealing with policies related to children acquiring English in the US.

JF: Thank you, Professor Krashen.