Notes by Steve Krashen on the Unz Attack

Monday February 9, 1998


Ron Unz put an interesting paper on his website today, attacking me and Jim Cummins for various things. Here is my response, which includes Unz' entire paper. Comments appreciated!

In Ron Unz' recent position paper, there is at least one error in every sentence. I reproduce this paper here, pausing at points to illustrate the mistatements, distortions, and factual errors:

UNZ: Supporters of bilingual education rely on the discredited social theories of politically-correct academics such as Steve Krashen and Jim Cummings. Their claims- -endlessly repeated by proponents of bilingual education---include the following:

ERRORS:

1) My work is not a "social" theory but a theory of language acquisition.

2) My theoretical work has been criticized (of course), but is hardly discredited. It remains the most widely cited theory in the field of language acquisition.

3) Unz misspells Jim Cummins' name.

UNZ: The older you are, the quicker and easier it is to learn another language.Young children have the hardest time learning a new language; adults have the easiest time.

ERRORS:

4) Unz is apparently referring to my published work of the late 1970's, a review of research studies that show that in early stages, older children move more quickly in second language acquisition than younger children and that adults are typically faster than children. This only holds for early stages. In the long run, those who begin second language acquisition as children do better than those who begin as adults.

5) None of the research cited in the review dealt with "ease" of language acquisition.

UNZ: Young children require 5-7 years to fully learn a new language (and the latest study by UC Riverside bilingual academics claims the correct figure is actually 10 years).

ERROR:

6) There is variation depending on the age of the child, but it can take five to seven years to reach grade level. Because the demands of the curriculum are much greater for the older child, it typically takes them longer to reach grade level. Older children have more to learn. One cannot shorten the process by removing children from bilingual education: Without bilingual education, it takes longer.

UNZ: Children learning English should be exposed to minimal English until 9 or 10 years old.

ERROR:

7) My position has always been that children should be exposed to maximum comprehensible input in English from the beginning. I have never supported delaying English for any reason. Mr. Unz should know that this is my position. I told him this on the telephone, and discussed it on KCRW, in response to his statements on the same program.

The programs we have designed provide comprehensible input in English directly from the beginning, in the form of ESL, and move the child into subject matter teaching in English as soon as it can be made comprehensible. There is no point in exposing children to incomprehensible input, to noise.

Providing the child with subject matter instruction in the first language before they are instructed in English is a wonderful investment, because it makes instruction in English more comprehensible. A child who knows math, thanks to instruction in the first language, will understand more in an English-language math class than a child who doesn't know math. The one who knows math will get more math, and will acquire more English, because he will
understand more.

UNZ: These academic dogmas have been produced by individuals with no background in the hard sciences.

ERROR:

8) This ad hominum argument is irrelevant, but happens to be false. If neuropsychology is considered hard science, I qualify as a hard scientist. My doctoral dissertation was in the area of neurolinguistics. I have published in Brain and Language, The Journal of the Acoustical Society, Biological Psychiatry, and The New York Academy of Sciences and am co-author of The Human Brain (Prentice-Hall). I also held a post-doctoral fellowship at the UCLA Neuropsychiatric Institute.

In addition, I have a Ph.D. in Linguistics from UCLA. I have over 200 publications, including ten books. My book, Second Language Acquisition and Second Language Learning, won the Mildenberger Award for the best published book in Second Language Acquisition in 1981. I was co-winner of the Pimsleur Award, given by the American Council of Foreign Language Teachers in 1985 (for a paper on sheltered subject matter teaching), my paper "Lateralization, language learning, and the critical period" was selected as a "citation classic" by Current Contents,
and I was given the Distinguished Presentation Award by the editors of the School Library Media Annual in 1993. I have presented keynote or plenary speeches at the International Congress of Applied Linguistics, TESOL (Teachers of English to Speakers of Other Languages), NABE (National Association for Bilingual Education), and the Georgetown University Roundtable on Languages and Linguistics.

What has Mr. Unz published on these subjects?

UNZ: Besides contradicting common sense and the personal experiences of nearly every immigrant family, these ideas have no support whatsoever among researchers in brain neurobiology or related areas of science.

ERROR:

9) We have documented numerous case histories of immigrants consistent with our position, namely cases in which those who arrive with more education in the primary language do better in English language acquisition. These case histories are consistent with empirical studies.

By contrast, a standard scientific work entitled Speech and Brain-Mechanisms by leading neurobiologists Wilder Penfield and Lamar Roberts (Princeton, 1959) devotes its last chapter (pp. 235-257) to language acquisition, concluding that:

UNZ: Prior to the age of approximately 10, a child's brain is especially suited to rapid language acquisition through the indirect method, i.e. simple immersion in the new language.

ERROR:

10) This point is irrelevant to my position on bilingual education, as I fully support early exposure to comprehensible English. But, for the record, Penfield and Roberts' work was on mapping areas of the brain through stimulation of the cortex, to determine which areas were responsible for language so they could be avoided in surgery if possible. Their chapter on language acquisition is pure speculation and has been supplanted by the research of the last 30 years.

UNZ: For young children, the process of learning a second language through immersion is identical to the process of learning their own primary language, and is as trivial and easy. When adults learn a second language through the direct method, i.e. formal instruction, the process is far more difficult, lengthy, and incomplete.

ERROR:

11) This point is irrelevant to my position on bilingual education, as I fully support early exposure to comprehensible English. But, for the record, we have documented that adults go through the same process in language acquisition as children do, and that "comprehensible input" (language that the acquirer understands) is responsible for language acqusition in the adult as well as the child. The problem most adults have is that they have rarely had a chance to get comprehensible input in another language. Traditional language classes give you very little.

UNZ: This scientific understanding of human language-acquisition has long been settled. Furthermore, this view has recently been reinforced by experiments which use magnetic resonance imaging techniques to examine brain activity in bilingual adults. A paper published in Nature (7/10/97, pp. 171-4), a prestigious international scientific journal, concluded: Individuals who learn a second language as young children store that language in the same area of their brain as they do their primary language. Those who learn a second language as young adults or later, store that language in a different portion of their brain. This indicates that learning a second language as a child is functionally similar to learning a first language; for adults, the process is quite different.

ERROR:

12) Once again, this point is irrelevant to my position on bilingual education, because I support early exposure to comprehensible input in English. But for the record, the researchers in Nature found far more similarity than difference between languages acquired by children and adults. For both, the classical language areas were involved, and for both, representation in Wernicke's area was identical. There were very small differences in Broca's area. In my view, this study supports the idea that child and adult language acquisition are similar.

UNZ: Based on these scientific facts, our current system of "transitional bilingual education," in which young non-English-speaking children are not heavily exposed to English until they are in 4th or 5th grades, past the age at which they can easily learn it, appears utterly bizarre and nonsensical.

ERROR:

13) In good bilingual programs, young children are exposed to a great deal of comprehensible English quite early. The UC Riverside study that Mr. Unz cited above reported that by the time children in bilingual education in the Santa Ana district were in grade three, they were doing about 75% of their math, science, and social studies in English.