A Question for Early-Childhood Programs: English First orFamilies First?
Should a society push an educational policy in the face of evidence that the policy is harmful to children and their families? This question is at the heart of a controversy over the provision of English-immersion preschool programs for 3- to 5-year-old non-English-speaking children. The purpose of these programs is to transform language-minority children into English speakers as early as possible.
In her June 5, 1991, Commentary, Rosalie Pedalino Porter challenges the stand taken by the National Association of Bilingual Education against these programs ("The False Alarm Over Early English Acquisition"). Ms. Porter attempts to discredit as a "false alarm" the conclusion drawn from a study I recently completed in collaboration with over 300 members of the education community from across the nation. This study found substantial and alarming evidence that English-immersion preschool programs frequently lead to children losing their primary language, and, with that, the ability to communicate with their parents.
Is this a false alarm, or is it an early warning to the communities whose children are targeted for such programs? In our eagerness to promote English, are we undermining the fundamental relationship between parents and children? Could we actually be increasing the educational, social, and psychological risks that these children already face?
Ms. Porter identifies three issues in this controversy. The first is: "Which early-childhood interventions best help limited-English speakers overcome the language barrier to equal educational opportunity?"
Early childhood is hardly a period in which one can speak of "the language barrier"! We're off to a bad start if we believe that a child's linguistic accomplishments by this period constitute a barrier to otherwise available educational opportunities. The purpose of early education is the total development of the child--cognitive, social, emotional, and physical as well as linguistic. The foundations for intellectual and social functioning and for schooling are laid in the early years. Research in child development has shown that primary-language development and cognitive and social development occur in tandem and are inseparable. Research has also established that well-developed primary language is essential to future success in second-language learning. Three- to 5-year-old children are at a critical juncture in development--everything is in place, but not much is stable.
Can children learn a second language during this vulnerable period? They can--easily, in fact--but not without disruption to their developing sense of self, especially when the second language is promoted in a way that suggests that it is more socially desirable and valued than their primary language.
Can children lose a primary language? Ms. Porter asserts that they don't, but anyone who has studied children's second-lan5p6guage acquisition in depth, as I have, knows differently. English often does replace the primary language of minority-background children. And the younger they are when they learn English under the conditions Ms. Porter is recommending, the more likely they are to lose their first language. When young children undergo this process, much can go awry. What is lost when English displaces the primary language? If that's the only language the parents speak, children stand to lose much of what parents can teach them.
The early-education interventions that work best for language-minority children are the ones that work for all children. Children need the kind of early-education experiences, in a language they understand, that turn them into enthusiastic and independent learners. They need experiences that build on the linguistic and intellectual resources they already have.
What children don't need are early educational experiences that destroy their confidence in themselves and their families--experiences that tell them: You are unacceptable the way you are. If you want to belong, you must change--or be left out. These are messages that get communicated to language-minority children in English-immersion programs. No language barrier impedes these messages. They come across loud and clear. Children want to belong, so they learn English to gain acceptance. And many eventually reject the language spoken at home--the language that made them unacceptable in the first place.
Ms. Porter's second question is: "What are the main responsibilities of public education for these children, and to what extent should schools be concerned with the change of language use in the home?"
The responsibility of the society's schools is not to supplant parents, but to assist them in preparing their children for the adult roles they must eventually take on. Yes, American schools must help children develop the English-language skills that will allow them to take full advantage of educational opportunities. And schools must provide access to the curriculum that will give children the knowledge, skills, and capabilities needed for jobs and other adult responsibilities.
But children need more than what schools provide to become responsible and productive adults. They also need what their parents can and must teach them--notions of justice and fairness, ideas about personal responsibility and honor, attitudes about work and effort, and beliefs about what is good and what is not. It is the parents' job to socialize children in the values, beliefs, and practices that will enable them to become the kind of men and women they want them to be. All of this takes time, closeness, and communication.
What happens to this kind of teaching when children abandon the parental language? In our survey, 64.4 percent of the parents whose children attended English-immersion preschools reported negative changes in family communication patterns. Few of these parents spoke English, but their children were abandoning the parental language and communicating in English at home. These changes in communication patterns were clearly affectingel10lparental authority and family closeness. Parents reported difficulties in talking to their children and in giving them the guidance they needed.
We give a lot of lip-service to the role families should play in the education of their children. Do we really mean it? The partnership between schools and parents is a cornerstone in the President's new education agenda. The job of educating children for the world they will inherit is too big for the society's schools to handle alone. It can be done only in cooperation with the children's most important teachers--their parents. We mustn't undermine a family's ability to educate its own children. But that's precisely what we do when we put parents in the untenable position of accepting early education that may very well alienate their children from them. They seldom realize how dangerous these programs are until it's too late. How can parents support the educational development of their children when they don't speak the same language? When we adopt programs that damage the family's ability to do the job it is supposed to do, who picks up the slack? Who's going to offer children the moral support and guidance that they're supposed to be getting from home?
Ms. Porter's third issue is the validity of the nabe study. The study was conducted without funding by educational researchers, practitioners, university students, community workers, children's advocates, and parents--people who got involved because of their concern for the welfare of language-minority children and their families. This was deliberate. The problem we studied needed to be the concern not only of researchers but of the education community at large. We saw the study as an opportunity to shed at least "a thousand points of light" on the problem.
Over 300 volunteers turned out to interview a total of 1,100 families across the country. The families included Latinos, Asians, American Indians, Arabs, and Europeans. The families had all had children in early-education programs. Ms. Porter charges that the study was flawed because the children were "not randomly assigned" to their programmatic treatments. The study was a survey, not an experimental treatment calling for random assignment. In most cases, the programs the children attended were the only type available to them. Low-income families generally don't have the luxury of choice when it comes to educational opportunities.
The study we conducted can and should be replicated. What would other researchers find if they were to look at children who have been in English-immersion programs such as the one Ms. Porter touts as a model for early education? I would bet that if such an investigation were conducted by trained language researchers and they were to examine family communication patterns several years after children have completed such programs, they would find essentially what we found in our survey. Two out of three families would be experiencing negative changes in communication patterns. One out of two children would be losing their primary language.
Future studies ought to focus especially on families with teenagers who lost their primary language through their preschool experiences. Our study suggests that early communication problems can lead to dangerous breakdowns in parental authority when children are older. Ms. Porter asserts that "the inevitable clash between generations, the desire of children to behave in ways that are not traditional or acceptable to their par4ents" is a natural occurrence, whatever language they speak. This is a fallacious argument. It's true that inter-generational clashes occur, but they certainly aren't inevitable. For most of the groups in our study, such problems are strictly an American phenomenon. They occur when a society invalidates rather than supports the role families play in the rearing of children. In tumultuous and dangerous times like ours, strong family ties are needed to keep young people out of trouble. It is plain bad social policy to support educational programs that destroy those ties.
Breakdowns in family communication sometimes figure in tragic events. After our study was completed, a news story provided a terrifying epilogue to the situation we have sought to warn the public about. In Sacramento, Calif., four Vietnamese youths took over a store, holding customers and staff hostage for nearly eight hours. The police at first brought in a Thai interpreter, since the boys didn't appear to be speaking Chinese or Vietnamese. They were speaking their version of English.
The siege ended with six people killed, including three of the gunmen, and 11 wounded. Three of the gunmen were brothers. Their parents said in an interview later that they had not been able to communicate well with their sons in years. The boys spoke little Vietnamese--at the level of 3-year-olds, their mother said. They had abandoned Vietnamese years ago. The mother said her boys were in a "middle world"--caught between their family and the larger society. This was communicated through an interpreter since the parents speak little English.
Language-minority children must indeed learn English, and schools in this society must find appropriate and effective ways to help them do so. I have devoted several decades to investigating how best to support the linguistic and educational development of such children. From that vantage point, I can see clearly that English immersion for preschoolers is the worst possible solution to a complex problem. If we can't educate young children without harming them, let's leave them alone.
Lily Wong Fillmore is a professor of education at the University of California at Berkeley. Over the past 20 years, she has conducted four large-scale longitudinal studies of the learning of English as a second language by Asian and Hispanic children in U.S. schools. Before that, she was a volunteer teacher of migrant children in farm-labor camps in California.
Research and the Renewal of Education: A Critical Review
Volume 10, Issue 39, June 19, 1991, pp 44, 34
Copyright 1991, Editorial Projects in Education, Inc.
Research and the Renewal of Education: A Critical Review
By Myron Lieberman
On April 28, the National Academy of Education released a summary of its report entitled Research and the Renewal of Education. The nae is a group of 75 individuals characterized by its news release as "renowned educational leaders or researchers." Essentially, Research and Renewal is an effort to define the role of research in the education-reform movement. Its authors emphasize the crucial nature of research and the importance of relying on it in education "just as we do in the other vital endeavors that shape modern life, such as industrial technology and medical practice."
Inasmuch as I regard the report as an unmitigated fiasco, my interest in it should be clarified at the outset. My new book nearing completion includes a chapter on educational R&D. I regard the topic as extremely important and looked forward to an authoritative statement on certain issues discussed in the chapter. The nae report fulfilled this purpose, but in a way that calls for an immediate candid response.
Let me turn first to its main recommendation: "Nae recommends that overall research funding for education be increased from roughly one half a percent of total expenditures for U.S. educational institutions at all levels (about $150 million)."
The financial presuppositions and the logic of this recommendation are a mystery. The nea Estimates of School Statistics, 1990-91, prepared by the National Education Association, estimates current expenditures for public elementary/secondary education to be $198 billion. If capital outlay, interest, and current expenditures for other programs are included, the figure is $222 billion. Even this amount would be an enormous understatement of the costs of K-12 education. For example, 30 percent of all college freshmen were enrolled in a remedial course in 1989. The annualized value of school land and buildings is not included, nor are several costs of public education on the budgets of non-educational agencies. A realistic figure for public education would be well over $300 billion, but the nae report implies that $300 billion is the total spent for all levels of education. Obviously, this includes higher education.
In 1990, the National Center for Education Statistics estimated that expenditures for higher education would be $126.8 billion, $82.5 billion in public and $44.2 billion in private instiel10ltutions. The only arrangement of figures from both K-l2 and higher education that comes close to $300 billion is to combine the $198 billion for K-12 with the $82.5 billion for public higher education. Aside from the fact that these figures vastly understate the actual expenditures, the nae report does not assert the estimates are for public institutions only and there is no logical reason why they should. Furthermore, the report calls for an unspecified increase in "overall research funding." Is the increase supposed to include funding for research on higher education? The language of the report and the inclusion of higher education in the base imply that it is, but the report is otherwise oriented to K-12 education. In any case, it grossly understates the real costs of K-12 education, quite possibly by 100 percent or more.
These errors are modest compared to those for educational research. The report estimates that $150 million is spent for educational research. It also asserts that the federal government spends between $100 million to $200 million for it, depending on how research is defined. Another part of the report asserts that the 28 largest foundations contributed $36.8 million to educational research in 1989. These assertions imply that very little is spent for educational research outside of these sources.
In fact, the $150 million estimate leaves out more spending for research than it counts. Referring to sabbatical leaves, the report states: "The direct subsidy for education research given in this form is larger than any other source of research funding for the field of education in the United States."
This assertion is false by a wide margin. Institutions of higher education employ about 35,000 faculty members in education. Surveys from the Association of American Colleges of Teacher Education and other data suggest that, on the average, 14 percent of their load time is allocated to research. True, some of the 35,000 are not in fields that relate directly to elementary/secondary education, but they are probably outnumbered by academics in other fields who conduct educational research; the nae's membership includes several such individuals. On the basis of conservative assumptions about average compensation, and using a 50 percent overhead rate, the value of regular faculty time devoted to research is about $283 million. This estimate is based on the assumption that 10 percent of the faculty is on sabbatical leave at full salary at any given time. Under any reasonable set of assumptions, institutions of higher education are spending $300 million to $400 million a year for educational research. This does not include the value of student time devoted to more than 8,000 doctoral dissertations in education annually. If such time were valued at the average annual salary and fringe benefits for all teachers, $200 million would be a reasonable estimate of its total value.
Elsewhere, the report states that "sabbatical support and a portion of many regular faculty salaries earmarked for research, undoubtedly constitutes the largest single investment this nation makes in education research. But because it is so thoroughly mingled with teaching and administrative functions, the overall amount and its impact on research productivity remain unclear."
Why is there no estimate of the value of regular faculty time allocated to research? After all, none of the estimates in the report are precise. Why not a range, similar to the "$100 million to $200 million'' cited for federal expenditures for educational research? I suspect the reason has something to do with the adage: "Don't ask the question if you can't stand the answer." Any reasonably honest answer would have (1) demolished the argument that educational research is underfunded; and (2) raised some hard questions about our research leadership, which is so amply represented in the National Academy of Education.
In preparing the report, the nae surveyed 28 large foundations whose total grants amounted to $1.1 billion in 1989. Of this amount, $272 million went for education, including $36 million for educational research. The nae report cites this as evidence that "disciplined inquiry, and all the constellation of information-gathering activities around it, are but a small speck on the broad social vision of foundations."
It is possible to interpret the data differently. In 1989, the 7,581 foundations listed in the Foundation Directory made grants totaling approximately $7.7 billion. If the proportion going to educational research was comparable to the proportion shown by the 28 foundations in the nae survey, foundations contributed $257 million for educational research in 1989. This might not satisfy educational researchers, but it is surely more than a "small speck."
Furthermore, other significant expenditures for educational research are ignored in the nae report. For example, the 1990 annual report of the Educational Testing Service shows expenditures of $50.1 million for "research, development, advisory and instructional-service expenditures" for the year ending June 30, 1990. Ets itself put up $21.1 million of this amount. Granting that not all of it went for educational R&D, a significant amount surely was spent for this purpose. Yet ets is only one of the many private companies that fund educational R&D.
The question is not whether our nation is getting an adequate return on these expenditures; of course it isn't. The point is that research policy should take into account all the resources devoted to educational research. If the return is inadequate, we might expect recommendations on how to get more from them, or (Heaven forbid!) perhaps institutions of higher education should stop giving load credit for unproductive research time.
In any event, a report urging more funding for research that is itself based on inaccurate and seriously incomplete data is hardly persuasive.
The nae report emphasizes the need to "restructure schools" but nothing in it suggests that interest-group opposition, not lack of "research," may be what prevents "restructuring" education. As Mikhail Gorbachev and others are finding out, real restructuring requires that some interest groups get hurt. If every interest group ends up with the same powers and privileges that existed before perestroika, restructuring is a myth. And so it is in education, despite the neglect of the issue in the nae report. Reading it, one would assume that a lack of information, remediable by "research," is holding up restructuring. The possibility that interest-group opposition, not lack of information, is the obstacle, is not considered. For instance, teacher unions did not just oppose experiments on contracting out instruction; they literally sabotaged the experiments. This is par for the course when "restructuring" is more than an academic buzz word. I would not expect Albert Shanker, an nae member, to emphasize this point, but neither would I expect the nae to be such a patsy for him.
This may have been naivete on my part. After all, without support from the two national teachers' unions, federal funding for educational research would shrink even more than it has. In this connection, the research success stories cited in the report raise some troublesome questions. One such story is supposedly "cooperative learning," a teaching strategy whereby students are di4vided into groups within classes and supposedly teach each other. This is the current fix to solve the problems emerging from the opposition of the research establishment to tracking. Such opposition implies an extremely higher degree of classroom heterogeneity, an obvious problem for classroom teachers.
The rest of us might have reason to be perplexed. If students can learn from each other, perhaps they can learn from teachers who aren't certified. Unfortunately, the possibility is off limits to educational research. Its political base will not tolerate it.
The research on school finance is cited as another research success story. We are told that this research led to and supported legal and legislative action to reduce disparities in educational spending between rich and poor school districts. To this observer, citing research on school finance as a success story is incredible. True, the research has been instrumental in getting the states to provide a higher proportion of school revenues. Unfortunately, no one has paid much attention to who pays the states. State taxes are highly regressive; the states get almost half of their revenues from sales and excise taxes, and their income-tax structure is not very progressive. Thus, as a higher proportion of school revenues comes from state sources, we can expect a higher proportion to be paid by low- and middle-income taxpayers. If this is an example of the efficacy of educational research, perhaps we can be less concerned about its decline. For that matter, neither the public nor policymakers have any reasonably accurate idea of the total costs of public education or the share of these costs absorbed by taxpayers at various income levels. These issues are not even addressed, let alone answered by our experts in school finance.
Again appealing to popular themes, the report states that "new incentives are needed to draw talented young people into the field of education research, including scholars from disadvantaged and ... minority groups." No evidence is presented on why more researchers8from the disadvantaged and minority groups is a problem. Since the educational-research community does include minority members, is the implication that ethnic groups should be proportionally represented? If not, what proportions of educational researchers should be black, Hispanic, or American Indian?
The report acknowleges that pork-barrel politics and political concerns play a significant role in what federal research funds are spent for and who gets to spend them. Its solution appears to be "further study" of the viability and value of a national panel of reviewers to "advise the federal R&D effort, proposing consensus on what is known and recommending new studies to close gaps in the research base." Is anyone really expected to take this proposal seriously? The federal government is free to get advice from anyone. If the "research community'' agrees on something, federal officials will know about it. The notion that some research group will put aside all of its interests and hobby horses and advise otherwise confused federal officials on what needs to be done, solely in the public interest, is about as anti-research as a proposal can possibly be. If the nae report is a sample of what to expect, further study of the idea should not be necessary.
In view of the fact that the president of nae (Lee Shulman) and a co-director of the report (Michael Kirst) are prominent Stanford University professors, the nae report understandably ignores the problem of excessive overhead rates for research. Nevertheless, one might expect a report on new directions for federally sponsored research to acknowledge that some financial reforms are needed on the recipient side. Be that as it may, the nae report has none to suggest. It is a conventional special-interest plea for more federal funds, prepared and disseminated by parties who would benefit but risk nothing if their recommendations turn out to be disastrous. The one useful purpose served by the report is to underscore the probability that more federal expenditures for educational research will not accomplish anything.
Myron Lieberman most recently served as visiting scholar at the Social Philosophy and Policy Center, Bowling Green State University. This analysis is based on the final draft of the report, made available by the National Academy of Education.