"What's New in Bilingual Education? Five Short Talks"

Keynote address by Stephen Krashen (NJTESOL/NJBE Awards Dinner)

He's a very strong advocate of bilingual education throughout the United States. He answers every letter to the editor and he writes articles constantly. And he's been a very big friend to us at New Jersey TESOL/NJBE. I'd like to introduce Dr. Stephen Krashen, Professor Emeritus at the University of Southern California. (Dr. Krashen was introduced by Judith O'Loughlin, past-president of NJTESOL/NJBE, Inc.)

Judy, I have two and a half hours, is that right? O.K., good, let's do it. Till midnight, right? Nobody leaves. There is a little handout. Did you get one of those? Yeah? O.K. If you didn't, too bad, I don't have any.

Actually, I'm not going to give one talk. I'm going to give five talks, I've decided. I couldn't decide which of five topics, so I decided to do all five. Each one is going to last just a few minutes, and after each one I'll pause and wait for thunderous applause. So, this is a way of milking more applause from the audience, alright?

The first three are comments on something called the National Reading Panel. Have you heard of the National Reading Panel? Just curious. Raise your hands. O.K. You better find out about the National Reading Panel. The National Reading Panel is the foundation of George Bush's plan for schools in the—you've heard of George Bush—right.

Well, the bill that just passed, the big education act, the reading part of it—the other part of it was about testing, you know, no child left untested, you got that one?

Anyway, the other part was about teaching reading and about phonics, etc... That was based on the National Reading Panel report. And you can find out about the National Reading Panel by checking it out on your computer. So, if you, tonight, when you go home, if like me you have no life, you can type in "National Reading Panel" on your search engine and here's what you'll get. You'll get the full report, 600 pages. You can download it, O.K.? The government will send it to you at government—that's your money—taxpayers' expense. They'll send you all 600 pages. They'll send you two different summaries. And they'll send you a videotape of the whole report.

And you'll also see on the sheet, on the first page of the Web site, that the National Reading Panel—the government— has hired a public relations firm to publicize the report, as if they need it.

Let me tell you who was on the committee. The committee consisted of a chief executive officer of a corporation, a physician, a physicist and about six educational psychologists. To paraphrase my colleague, Susan Ohanian, these people wouldn't know a child if one bit them on the leg.

I love being professor emeritus. I can't lose my job. Alright? Actually there was one legitimate educator on the panel. Her name was Joanne Yatvin, and she's a former teacher—now principal—and she understood the situation. She wrote the minority report. Are there any questions?

And if you'd like the full gossip, her version of it, she did a very nice description that was written up in the Phi Delta Kappan, in the last issue. And the Phi Delta Kappan has selected articles on their Web site that you can download and, fortunately, hers is one of them. So, if you really want to find out what happened, how all the people went in with their minds made up, pretty much, you can read her report as well.

So, my first three points tonight are about the National Reading Panel. And the first point has to do with the National Reading Panel's stance on something called "phonemic awareness." Just curious, how many of you have heard of phonemic awareness? Please raise your hand.

I asked this group the same ten years ago and only three people raised their hands. And now everyone has heard of it. Those who raised your hand, how many of you are really confident you know what it is? The same people ten years ago who raised their hands.

O.K. Well, the government is now telling us that if the kids don't have phonemic awareness, they'll never learn how to read. Well, this is something we never heard of ten years ago. How did we all learn to read? It's some mysterious process.

Anyway, phonemic awareness is not phonics. It's kind of pre-phonics. It's your ability to break a word or syllable into its sounds, completely orally. And the idea is children must develop phonemic awareness or they'll never learn to read. And the way to do it is through phonemic awareness activities.

The National Reading Panel devoted 60 pages to research on phonemic awareness. Let me give an example of phonemic awareness. Phonemic awareness activities are divided into two basic types. One is blending, one is segmentation. In blending you take individual sounds—the child says, "kuh," "uh," "tuh" and combines these "kuh-aa-tuh" and makes the word "cat" out of it. The other is segmenting, where the child takes the word and divides it into various sounds.

I want to give you a segmenting activity, just to give you the feel for this. I'll give you a word and you tell me the first sound, what we call the first phoneme. O.K.? Are you ready? O.K. "Pit." What's the first sound? "Puh." Very good.

O.K., ready? "Split." How did you learn to read? Well, some of you made the same mistake I did. You said, "Spuh, bluh." The answer is just the "s" sound. O.K.? Now, those of you who got it right: how did you do it? I'll give you a prediction. The way you did it is you spelled the word in your mind, right? You saw it in your mind's eye and pulled away the "s." Isn't that right?

What I just demonstrated to you is that phonemic awareness is the result of knowing how to read—not the cause.

This is nearly a hoax and a very dangerous one. But let me give you the research. I'm going to repeat what I just said, but seriously and give you all the research behind it. The reason the government thinks, the feds think that phonemic awareness is so cool—I was going to write an article, I was going to call it the unbelievable coolness of phonemic awareness—everybody thinks it's just the most amazing thing to study.

Anyway, the way they think they've demonstrated this is what's called training studies. In training studies, you take two groups of children: one group gets phonemic awareness activities—and these are four-year-olds, five-year-olds, etc..., and they do the kinds of exercises I just told you, they do segmenting, blending. And then there's a comparison group who does something else not related to phonemic awareness. They then give them pre-tests, post-tests, and guess what. The children with phonemic awareness training do better on tests of phonemic awareness.

I want to make two comments on this. First of all, the comparison group gets better, too. How did that happen? I think it's because of maturity. They grew up and they did a little reading. Alright? Also, the main point I want to tell you is I have isolated all the studies that the government could find and that I could find, and I went through quite a search in finding these, on where the test was not a test of phonemic awareness, was not a test of reading words in isolation, but was a real test of reading comprehension. What do we have that really shows this stuff helps kids learn to read? Well, I found only six studies in the entire professional literature.

I looked at the federal government's report. I'm in communication with some of the friendlier people on the committee, and they have been—Linnea Ehri is sort of the loyal opposition here—and she's been kind enough to send me her information, one of the studies. Also, on various listservs, I've asked people, you know, reading professionals who do research in this area, respond, do you know of any studies, blah, blah. All I could find were six studies.

Of the studies, three were done with English. The other three were done in other languages that are phonetically regular. The only study I could find where it was a clear difference, where phonemic awareness training really helped and it was significant and the effect size was large, etc..., was with one group of 15 children in Israel, learning to read in Hebrew as a first language.

So, the entire, huge phonemic awareness industry, this huge golf ball rests on this tiny, tiny tee of 15 children in a pre-kindergarten somewhere in Israel. That's it, folks. Now I'm not ready to dismiss phonemic awareness and say it's nonsense and all that, although I kind of think that's true. But I think a serious research hypothesis that has so many practical implications, that's changed everyone's life, should be made of much, much sterner stuff. The research reference is there for those of you who want to read it. The entire phonemic awareness movement, I think, is based on the slimmest evidence.

Well, now that phonemic awareness is taken care of, let's move on to phonics. The federal government claims, in 60 more pages in the National Reading Panel, that phonics is good for you, especially intensive, systematic phonics. And one of the many reasons they think phonics is good for you, in the 60 page report, is they compared children who have intensive phonics instruction with children who learn with something called "whole language."

And they claim that children in phonics do better. I've discovered that this has become common knowledge among virtually everyone in the United States. I found out that my barber believed this. He's no longer my barber. I got tired of arguing. He didn't understand bilingual ed either.

I now go to a hair stylist. It's so much more fun. My cousin Harvey got me into this. I was on T.V., on this program, and I didn't like the way I looked. So, I went to Harvey, who's my fashion consultant. I said, "Harvey, do something," so he says, "O.K., we're going to go to Sak's, we're going to get you some good clothes and I'm going to take you to my friend Jeff, who's a hair stylist."

Not only did Jeff give me a better haircut, but he understands whole language, he understands bilingual education. You sit around, you have a real conversation, and a nice cup of coffee. It's great. I understand now why women do all these things, it's great. I've discovered that Jeff and his colleagues at the hairstyling salon are all very good group therapists, as well. Isn't it true? Very nice conversation.

Anyway, this is what I was able to tell my hair stylist that I could not tell my barber. When you look at the research as I've done—I've looked at the research that the government says demonstrates that whole language fails, that intensive phonics instruction is better, and I did a few alterations.

First of all, there's quite a bit of confusion as to what is "whole language." These people aren't really sure. I decided to try to define whole language. I think the most prevalent definition of whole language, before I tried to repair this, was that "whole language means nice." Anything that you're doing that's warm and nice must be whole language. So, you can do spelling drills and give music in the background, and therefore it's ... no, it's not.

Whole language is a specific thing. I will claim this, and is a Ken Goodman, Frank Smith approved definition. Whole language must contain a great deal of real reading. If it doesn't have real reading, it's not whole language.

I discovered in the research literature that the government used that whole language, some classes were doing hardly any reading at all. In fact, sometimes the skills classes were doing more real reading. So, I made that correction in the research. I also looked at tests of reading comprehension, not just tests of phonics. My conclusion, which is in a journal called Talking Points, when you define whole language

correctly and you look at tests of reading comprehension, the result changes dramatically. The very studies the government used to, in their opinion, demonstrate that phonics was superior demonstrates whole language is better.

Children in whole language classes read better. They like reading more. They do more reading on their own. They do better on tests of telling stories. And when you look at tests of reading nonsense words, tests of skills, what they call words that are just made up that have no meaning so you can see whether they reflect knowledge of phonics, there's no difference between the groups. So, there's nothing lost. Again, the research changes dramatically when you take a close look.

So, we've now handled two of three major results of the National Reading Panel. Let me review. The federal government claims that phonemic awareness is important and helpful to learn to read and I claim there's no evidence showing this. Phonemic awareness appears, in my opinion, to be the result of reading, not the cause. Number two, is phonics better than whole language? No, it's not. When you look at the research correctly, whole language shows itself to be superior.

The third one, though, I think really takes the cake. The federal government—and this is such an amazing result that I had to actually put it on the handout so you'd see I didn't make this up. The National Reading Panel says, "We are unable to determine from the research whether reading silently to oneself helps you learn to read." Does reading for pleasure help children learn to read? The federal government says it doesn't know.

Now, the cosmetic consultant at Sak's Fifth Avenue knows that's false. Alright? This is the most incredible result I have ever seen. Now, the feds devoted 60 pages to phonemic awareness, they devoted 60 pages to phonics. They devoted six pages to sustained silent reading studies. Free reading in school.

You remember sustained silent reading. It's called SSR. Used to be called USSR, remember that? Anyway, the federal government looked at sustained silent reading studies—studies in which children read for pleasure, five, ten, fifteen minutes a day. The comparison group did skill-building activities and they said they couldn't find a difference. Well, they found 14 comparisons. That's all they were able to come up with. Actually the government's conclusion wasn't all that bad. They found in four studies the students in sustained silent reading read better, in ten there was no difference, and in no case were they worse. That's not so bad. That's not so bad at all. But they still concluded there is absolutely no difference.

Well, they found 14 comparisons, and this is the entire federal government. I found 54, working just a little harder, I think. In 51 cases out of 54, the children in sustained silent reading classes read at least as well or better. And when you look at long-term studies—my table is at the bottom of the page, there—the children in sustained silent reading studies never were worse, eight out of ten times read better. The federal government included no long-term studies, only short-term studies, and the studies that they included they made serious mistakes.

I'll give you the gossip on this one. I love gossip. Who was it who said, "If you don't have anything good to say about someone, sit next to me." My point of view. The Education Week summarized the results of the National Reading Panel and made this outrageous claim in the Education Week article that they showed that free reading doesn't help, recreational reading doesn't help. And I responded with a detailed letter to the editor to the editor—this is about a year and a half ago in Ed Week. And Ed Week, by the way, is edweek dot org or edweek dot com. And they have a free archive where you can look at any article in Ed Week without paying any fees. So, if you just type in my name you'll see the original letter.

And I wrote a letter to the editor which they published in its entirety, and I said that the National Reading Panel basically blew it, that they didn't look very hard in finding the evidence. They only looked at English as a first language. They ignored all English as a second language studies. They made serious mistakes in reporting. It was a sloppy, sloppy job.

Well, the feds responded. I don't want to mention any names. Timothy Shanahan, who was responsible for that section, who is now "the reading czar" for Chicago Public Schools, by the way. Mayor Daley has hired him to do this. I don't want to mention his salary. 120,000 a year. It's a lot more than anyone I know makes in education.

Anyway, Tim Shanahan responded and, in my opinion, his response to me made the matter even worse because he made even more mistakes in the response. Which I pointed out in my follow-up letter, which Ed Week did not publish. However, the Phi Delta Kappan published my long, long article on all of this.

Let me close this section by telling you a little about the Phi Delta Kappan and what has gone on there for the last year. The last year, the Phi Delta Kappan has been absolutely heroic. They have published article after article attacking the National Reading Panel, beginning with a spectacular article by someone named Elaine Garan—G-A-R-A-N—last year in March, followed by my article, an excellent article by Gerald Coles reviewing the phonics research again, finding more flaws, then Joanne Yatvin's paper and, two issues back, the editor of the Phi Delta Kappan, Pauline Gough, in an editorial, said, "Because of what we've published here the last year the National Reading Panel should be dead in the water."

It's not. In fact, this is not just academic quibbles. This is not just an academic squabble. Because of the National Reading Panel, huge amounts of money are now invested in phonemic awareness, phonics, in programs that conform to the results of the National Reading Panel, mostly by McGraw-Hill, "Open Court," etc..., which you've heard of. And money has been diverted from libraries into phonics programs, to make it consistent. The federal government, according to Bush's law, that just was passed a few months ago, you cannot get federal money for anything in education unless you show that your work is consistent with the results of the National Reading Panel. This is now the law, ladies and gentlemen, and what I've tried to show is that it is seriously flawed. We are in big, big trouble in language arts.

Well, that's my cheerful first three points. Oh, I forgot the extra applause. I'd like three times the amount of applause. Thank you very much.

Number four and five—we can go to the flip side now—are also related, because they both deal with what we might call "urban legends." You know what an urban legend is. There are alligators in the sewers in New York City. That's the most famous urban legend. There are many, many more. An urban legend is defined as "a captivating story that may or may not be true, that spreads rapidly, that people like to believe." Some of my favorites. Humphrey Bogart was the original Gerber baby. False, but it's a great story, isn't it? My absolute favorite, which you heard in elementary school, so did I. If everyone in China were to jump up at the exact same time, right? remember that one?, it would cause a tidal wave that would engulf the entire United States. I wonder if they have one in China that says the opposite. If everyone in North America were to jump up at the same time...

Well, I have two more urban legends to add to this, and they both have to do with test scores. One has to do with test scores going up; the other with test scores going down.

Urban legend number one. When whole language was introduced in California, in 1987, test scores plummeted. The next one? When Prop 227 was passed in California, to dismantle bilingual ed, test scores skyrocketed. You’ve heard both of these. I’m sure my former barber, Billy, spoke of them. These are part of the national fabric. They’ve been carried in newspaper article after newspaper article. They’re extremely serious. And they’re extremely wrong.

Let me do the first one in some detail. I’ll give you another version of it, because this is kind of what the public believes. In 1987, a group of fuzzy-minded hippies took over the California language-arts framework. I was in that group. In fact, I became one of the ring-leaders. And I was recently told that I, Steve Krashen, because of my evil role in this committee, am responsible for the testing hysteria that has gripped America. Alright? Because of this urban legend. This group introduced whole language. Test scores immediately plummeted until by 1992 our children in California were last in the United States on reading comprehension tests. And it’s only clear-thinking, fair-minded, all-American phonics that’s slowly bringing things back around. Now, I’m not asking you if you agree with the urban legend, but do you agree with my description? Is this pretty much what people believe?

Well, let me tell you what really happened. Yes, I was on the 1987 committee. That part is true. Did we bring in "whole language"? We never mentioned the word "whole language," not once. None of us knew what it was. In fact, you can look at the transcripts. You can look at the actual report, and you will see that we never mentioned whole language. What we did say is that language arts should be literature-based. I thought we were defining terms. I didn’t know what else language arts can be? That’s what it is.

Did anything change in California? There has been no study in California that a single teacher changed their behavior as a result of our committee meeting. We have no idea whether they did or not. Did test scores drop? Well, one part is true. In 1992, California children did rank at the bottom of the United States on a national examination—it was called the NAEP examination, N-A-E-P, national assessment of educational progress—and our fourth graders in California did in fact come in last in the country. That part is true.

But was there a drop? I’m pausing for effect. You better sit down when I tell you this. They never looked. This is a scandal ten times as outrageous as Watergate. They never looked. Let me repeat that. They never looked. 1992 was the first time the NAEP scores were analyzed by states. Are you with me? There was no pre-test. They assumed that things must have been much better in California before, but they never looked.

Jeff McQuillan looked. Jeff McQuillan wrote a book called "The Literacy Crisis: False Claims and Real Solutions." I refer to it as the "Pelican brief." Now, I’m not a completely disinterested observer. Jeff was my student and this was a revision of his dissertation. But, my goodness, what he found! Jeff looked. He looked at other tests given in California and the results are so important that I’ve put them on the handout. What you have on the flip side are CAP scores, comprehensive assessment of progress, in California, from 1984 to 1990, and you will see by glancing at these scores that there is no significant movement up or down anywhere in these scores. In other words, scores were low in California way before our committee met. We had nothing to do with test scores rising or falling. They were low well before. Believe me, we’ve tried to publicize this. McQuillan’s book has been selling fairly well for an academic book, certainly not as well as anything by Barbara Cartland, but, you know, it’s done O.K. I’ve written my usual letters to the editor on this. Some of them have gotten published. But we have not penetrated the public consciousness on this at all.

I’ve been writing my senators lately. I’ve been writing Diane Feinstein and Barbara Boxer. You know, I’d write a note like, "The current Bush administration plan is wrong and here’s why, blah, blah, blah. And the emphasis on testing is harming education." So, I keep getting replies like, "Thank you for your message. Senator Blah, Blah, Blah is also interested and we’re in favor of higher standards to standardized tests." So, then I write back saying, "Obviously you didn’t read my letter." Then you get something, "Thank you for your interest and blah, blah." It should be in Harpers, you know, you keep getting the same response.

Anyway, despite our efforts to publicize this, people still have the impression that test scores did indeed plummet in California. They didn’t. It’s an urban legend.

I want to briefly mention McQuillan’s analysis of the situation in California and why our scores are so low. He says it’s not because of lack of phonics. It’s because of lack of access to books. Let me give you a brief idea of what is going on in the altered state of California in terms of access to books. Let’s consider first school libraries. We know from research that I’m going to try to review tomorrow, go over some of it, that quality of school libraries is intimately connected to reading scores. Did you know that? The better the school library, the better the reading scores. That’s a very interesting result.

Well, let’s look at what’s going on in California. In the United States, the average elementary school library has 18 books per child. In California, in 1990—this is about the time our literacy, our supposed literacy crisis hit—we were dead last in the country with 13 books per child. That’s very low. Today, thanks to our two education governors, it is now 11 books per child. Los Angeles Unified? Six books per child. That is beyond outrageous.

School librarians. According to research, the presence of a school librarian is an independent predictor of reading scores. Keith Curry Lance's research. Very interesting. The United States has approximately in the average elementary school, or in general, in schools in general across the country, approximately 900 students to each school librarian. In California, before our literacy, when our literacy crisis hit, we had one for 5,000, by far the lowest. If we wanted to catch Mississippi, the next state up, it would have required a Marshall Plan. Mississippi was one for every 2,000. Today, thanks to our two education governors, it’s gotten worse. It’s now one for every 5,300. California spends exactly half of what other states spend on school libraries.

Public libraries. Also a predictor of reading scores. Our public libraries in California, no matter how you do it, are either in the bottom seven of the country, the bottom ten of the country. Library budgets have been slashed since 1987, since our committee met, approximately 30 percent. Children’s services have been hit the hardest.

Poverty. McQuillan’s data, he found that California ranks in the bottom ten of the country—or the top ten, you can call it—in terms of homes in which there are fewer than 25 books. This has gotten worse over the last few years. Our children in California are children in poverty. There is no question. We rank in the top, what, ten in the country? seven in the country, depending on how you do it, in terms of children in school who are in poverty, where poverty among other things, as you know, means fewer books.

Every single possible access to books that kids can have. In school, public libraries, at home, California is near the bottom of the country. In his book, McQuillan put together all these things into one figure and did a correlation of a composite measure, access to books, and how different states do on this NAEP examination. The correlation, for those of you who remember statistics, was over .8, which is sky-high for education. It means that just knowing the number of, the access of books that a child has in a state is two-thirds of the information you need to predict their NAEP score, which conforms as well to the rest of the research.

This urban legend didn’t happen. Yes, we were last in the country. There was no plummet. We were worst in California before, we were poor in California before. Things have not gotten better. McQuillan’s spin on this, his conclusion, I think is very interesting. Look at this. Test scores in California have remained flat, in fact they continue to remain flat. Since our phonics push, NAEP scores have not risen at all, something conveniently overlooked by the press. Alright? They’ve been flat for about the last 20 years, or as long as we’ve had data. During this time, access to books is low and is constantly getting worse. Still, our test scores have stayed the same. McQuillan’s conclusion, that means California’s teachers are doing a good job, doing the best they can under very difficult circumstances.

Well, now one that’s very close to the purpose of this organization. Is it true that when Proposition 227 passed in California test scores skyrocketed? Is it true that test scores have gone up in California? Yeah, it is, actually. Test scores have gone up. That’s undeniable. That we know. But does dropping bilingual education get the credit? No.

Let me talk a little about this, and before I do let me preface this by saying why this is such an important argument. And our, I think in fact, I don’t want to exaggerate, but let’s be serious. The future of this group, of bilingual education in New Jersey, rides on having an answer to this argument, because it is being used throughout the United States to eliminate bilingual education systematically in every single state. It’s going to happen here. Whether it’s done by initiative or by some other process, Unz is doing very well on his promise to eliminate bilingual education—he even wants to get rid of it in Rhode Island, where there’s only one school that does it. Alright? It’s that serious.

Prop 227 passed in California. Proposition 203 passed in Arizona, as you know. The recent polls from Colorado show the public is ready to eliminate bilingual ed in Colorado. And I’m afraid if the vote were today in Massachusetts we would lose by a bigger margin than in any other state. So, so far, he’s been doing an excellent job of distorting the data and convincing people that bilingual education is a bad thing.

Let’s look at this argument. Have test scores increased in California? Yes. Does 227 get the credit? No. Ron Unz must lead a charmed life, because something very interesting happened when Proposition 227 passed in 1998. At the same time, a new test was introduced in California. That test was called the SAT9.

Here’s what happens when you introduce a new test. And what I’m about to tell you has been pretty much backed up by research. Robert Lin’s research in the 1980s when they looked at what happens when you introduce a new test. The first time you give it, scores are low. The next time you give it, scores go up. And they go up, and they go up and they go up. Till after a few years you have to recalibrate the scores because the tests have gotten too high. This always happens. Test scores have gone up because test scores go up when you introduce a new test. The first time they gave the new test it was, in effect, it was considered a pre-test. It was considered a measure of the "old way," under bilingual ed. And now the first time administered, after one year of 227, two years, etc…

Well, if you just consider the data on test inflation, what we know about test inflation, you’ll already get a different interpretation. Test inflation, generally, is about one and a half to two points on each component. If you put this into the equation on test scores in California, in the elementary school, this accounts for half of the gains, for early grades. Middle school, it accounts for all of them. In high school, things have actually gotten worse.

Now, test inflation has been especially severe in California because of the intense pressure to raise test scores. Alfie Kohn would call this "bribes and punishments." Alright? If your scores go up, you get money in your pocket, not for your school, not for your scholarship fund, but for you that you can use to pay off your VISA bill. O.K.? And this can run into thousands of dollars. If your test scores go down, your school can be closed down. Because of this intensive pressure to raise test scores, districts in California and individual schools have resorted to extraordinary means to raise test scores. Am I about to accuse people of cheating? Yes. Absolutely. There are two ways of cheating. I love being an emeritus professor.

Now, this is Armageddon, folks. Our backs are to the wall, so I really have to tell you what I think on these things. One way of cheating is selective testing. My student Tom Gault explained this to me with a neat example. He says, "Here’s how you raise test scores. Find the students in your class who have infectious diseases. Have them sit next to the low scorers a few days before the test, and your scores will go up."

Now, a very alert reporter in the San Francisco Chronicle has very good evidence that, or very suggestive evidence that this is precisely what’s been going on in California. Her name is Nanette Asimov. Yes, she’s related to Isaac. She is Isaac Asimov’s niece. Now, she works for the San Francisco Chronicle and you can still find her article in the archives. It was about, I think, 2000, October, 2000, is when her first paper, her article on this came out. She looked at districts in the Bay area. She found where test scores went up, the number of students tested decreased. When test scores stayed the same, the number of students tested stayed the same. Are there any questions? Get it? This very suggestive evidence for selective testing in California.

Another way of increasing test scores? Test preparation. In fact, many districts now are hiring professionals to come in and do test preparation. Now, if ever there was a sign of pathology, this is it. This is outrageous. Test prep sometimes leads to more knowledge, sometimes leads to more learning—it’s supposed to. But sometimes it’s just devices to cause higher test scores without kids learning anything. The most obvious? If you know that the test does not penalize for guessing, tell your kids to guess. Which, you know, they’re not used to doing on your classroom tests, you know—that doesn’t help you understand where they are. A child has five items left. He doesn’t have time. He doesn’t know the answer. And it’s a four-choice possibility? Mark in B or C on all of them and you’re bound to get one or two of them right. This will raise test scores.

Doing things like this is like claiming you raised the temperature in the room when all you did was put a match under the thermometer. Now, let me, I just attacked the test. I’ve told you they’re invalid. I think they are, but this is not a sour grapes argument.

Let’s pretend for the moment—in fact, the public thinks this kind of argument I just gave you, the public interprets as whining, alright? That, ah, you’re just saying the test is wrong. Let’s assume the tests are right. Well, Proposition 227 still doesn’t deserve credit. Here’s why. Let me give you some interesting research on this.

When you’re finished looking up the National Reading Panel on your search engine tonight, type in the name Kenji Hakuta, H-A-K-U-T-A. And you will find his Web site with no trouble at all. Kenji Hakuta is a professor at Stanford University, a person with a sterling reputation in the field, and he’s done analysis on SAT9 results for, over several years. Let me give you his conclusion. SAT9 scores have gone up in districts that never did bilingual education. They’ve gone up in districts that have kept bilingual education through special waivers. They’ve gone up for everybody. And they’ve gone up just about the same for everybody.

The most recent result has come from Arizona State University, this very good paper that just came out by Thompson, DiCerbo, Mahoney and MacSwan. Anyway, they did a very careful analysis and compared the gains made by LEP students, L-E-P students, and fluent speakers of English. No difference. The kids made exactly the same gains.

Alright, now, I’ve pretty well taken care of the solid arguments. But there’s a special case I need to address. Ron Unz has cleverly directed nearly everyone’s attention on one district. And that one district is Oceanside. That’s what the articles have been about in the New York Times, etc… People always point to Oceanside, and in fact it’s the only clear case where even any kind of an argument for Prop 227 getting credit.

Let me give you the story. The superintendent at Oceanside, I won’t mention his name, Ken Noonan, N-O-O-N-A-N, took over at Oceanside at about the time 227 became law. Noonan—believe it or not, it’s a great story—was not only a supporter of bilingual ed, he was one of the founders of the California Association for Bilingual Education, and its first president. Isn’t this wonderful?

Reluctantly, he said, you know, I didn’t vote for 227, I worked against it, but the law is the law. And he systematically eliminated all traces of Spanish from the district. By the way, Oceanside is not all that far from Mexico. It’s just a little bit north of San Diego. Well, what happened? Test scores skyrocketed. That’s why I call this the skyrocket legend. Test scores went up dramatically. Noonan said, I’ve got to admit it, I was against this, but now I think English immersion is the way to go.

The result. Noonan was lionized. He was on all the talk shows. He was interviewed. Special articles about him. I’m surprised he wasn’t made national secretary of education instead of Ron Page(sp).

Well, let me tell you what really went on at Oceanside. We’ve all taken a very good look at Oceanside. First of all, the first SAT9 scores for Oceanside, looking at their LED population, the first time it was given, which is really a report card on bilingual ed, were incredibly low. The state average was 18, for LEP kids. Their average was 12, which is just about at the chance level. It’s hard to get it lower.

Next year, after Prop 227, it went up nearly to the state average for LEP kids, above 20. O.K., big jump. Well, I got very interested in that and I called the district. I said, "Can you tell me what kind of a bilingual program you had?" Maybe it was one of these models that we know doesn’t work, like concurrent translation. How did you manage to get such low scores? Well, the guy on the phone said, "Oh, we’ll let you know, we’ll fax you the results." This was four years ago when we faxed things. Right? We don’t fax things anymore. It’s all e-mail. So, the fax came in. It was their press release on how wonderful English immersion was. I then found out that it wasn’t the newspapers coming in and finding out—it was Ken Noonan publicizing, you know, the big growth under English immersion. I called them up again and said, no this wasn’t what I wanted. I wanted to know what kind of a bilingual program… This time I got no response.

I then found myself on the radio at an interview show with Ken Noonan. What an opportunity. A show called "Which Way, L.A.?" It was me, Noonan, Lord Voltemort and .. Norm Gold(?), a friend of mine, California department of ed, and I had a chance to talk to Ken Noonan on the radio. So, I spoke to him in our mutual Southern California dialect, trying to be real comprehensible. I said, "Ken, dude, talk to me. What up, G?" How did you get such low scores? He stalled, refused to answer the question and it was time for commercial. I then, on the advice of my lawyer and therapist, decided to write him an e-mail directly. I said, "Please tell me." Got no response. It’s called stonewalling. Noonan has a good reputation for being open to the press. He wasn’t open to me, but I didn’t have to pursue it much further. He then wrote an article that came out in the Washington Post that described what happened at Oceanside. And it’s all there.

The so-called "bilingual program" at Oceanside, Noonan said was a terrible program. They delayed English for four to five years. This isn’t a bilingual program. This is a monolingual Spanish program. I know very few experienced professionals in bilingual education who would support such a program. What we do in bilingual education as I think everyone in this room knows is we use the first language in a way that facilitates and accelerates second language development. We know the name of the game. We are committed to English language development, and the good bilingual models that we all agree with are ones that introduce ESL right from the very first day. True? Say yes. Yes. Please reassure me. And that introduce subject matter in English as soon as it can be made comprehensible. This is what we do.

So, this was a lousy bilingual program. Why didn’t he just tell me that.

I also found articles in the San Diego Union Tribune, if you can believe your newspapers, about what bilingual ed was like. One person said, for example, in talking about the Laurel school, which had been a bilingual school before 227, "When 227 came in, now we have books." In other words, the schools must have been terribly undersupplied. Since 227, there has been a phenomenal emphasis on test preparation in the Oceanside district. Teacher professional days have been reduced—in some cases, eliminated. More focus on English and mathematics, because that’s what’s tested. And you have seen this all through. We’ve seen this in every school.

In light of this, let me recommend a new book by Susan Ohanian. The title tells everything. "What Ever Happened to Recess?" We see this all over, where things have been cut back, cut back, to have more test prep.

The final data on the Oceanside miracle. Hakuta has analyzed their results. The gains at Oceanside, according to Hakuta, are, "not remarkable, the same that we find throughout the state."

Well, I’ve given you two urban legends, both of them extremely dangerous. They continue to cause a great deal of suffering and they continue to run virtually unopposed in the national media. They’ve both become part of the national fabric. Test scores did not plummet in California when whole language was introduced. When bilingual education was dismantled in California, test scores went up, but bilingual education had nothing to do with it.

I’m going to end here. I’ll just make a summary comment and then we’ll do the business tomorrow. The part that I’m going to do tomorrow is to show that the conclusions I’ve come to tonight—that whole language is not to blame and that bilingual education is not to blame—I will back up tomorrow by giving you additional arguments, independent arguments showing that whole language is a very, very good way to teach children to read. I remain unrepentant. I’m still a strong supporter of whole language. And I’ll give you the rest of the evidence tomorrow why bilingual education remains an excellent way of helping English language development. O.K. Film at 11:30.

This talk was taped on a Panasonic microcassette recorder at the NJTESOL/NJBE First Annual Awards Dinner, on Thursday, May 16, 2002, at the Doubletree Hotel in Somerset, New Jersey. The transcript is provided as a service for readers of the ESL MiniConference Online newsletter. Stephen Krashen's e-mail address is: krashen@usc.edu.

ESL MiniConference Online