About the Author
Read an Excerpt
Making Sense of Census 2000
Forty-seven million US residents – nearly one in five – speak a language other than English at home, according to the 2000 census. This group more than doubled over the past two decades, while the population that speaks only English expanded by just a fraction (see Table 1). If current rates of growth continue, a majority of Americans will be minority language speakers by 2044. Quite a sea change in a country renowned for its monolingualism. The implications for a national language policy (or lack thereof) are enormous.
All this assumes, of course, that the census numbers can be believed, an issue that deserves special attention when it comes to language. These decennial snapshots are instructive, to be sure. But like all photographs taken from a single vantage point, they can be distorted and misleading. For example, the growth of non-English languages was so prominent in 2000 that it's easy to miss a powerful countertrend: the growth of English at the expense of heritage languages.
The latest figures show that fluent bilinguals – respondents who say they speak English 'very well' – account for more than half of minority language speakers. That is, they outnumber those with less than full proficiency in English. Both groups are increasing at roughly the same rates. This pattern is especially striking when you consider that 42% of the foreign-born population counted in 2000 arrived during the 1990s and had less than 10 years to learn the language. With so many newcomers speaking languages other than English, it is harder to appreciate how many who immigrated just a few years earlier have become fluent, and often dominant, in English. Close scrutiny of the new census data suggests that the pace of Anglicization in this country has never been faster.
Besides immigration, high birthrates among linguistic minorities, Latinos in particular, have a major impact on these trends. As more data from 2000 become available, other factors may be identified in accounting for the rising numbers who speak languages other than English. Nobody is suggesting, however, that masses of Anglo-Americans have suddenly shed their complacency about language learning. Linguistic diversity in this country remains a largely ethnic phenomenon.
Imagine my surprise, then, on viewing the 2000 census figures for my home state of Maryland. More than 11 persons in the state reported speaking Spanish at home for every 10 persons who claimed to be of Hispanic origin. Even if we hypothesize that almost all Latinos are maintaining their heritage language – quite a stretch – that would still mean that a lot of Anglos are speaking Spanish. ¿Que pasó?
Is there something special about Marylanders? No. The census documented similar patterns in at least a dozen other states.
Are substantial numbers of Spanish speakers lying about their ethnic backgrounds? I know of no evidence to support that hypothesis.
Are English speakers not only marrying Latinos but also adopting their language? In some cases, certainly, but hardly enough to explain this phenomenon.
Have Anglos suddenly changed their attitudes and behavior when it comes to bilingualism? Doubtful, because there's nothing new here. In previous censuses, up to 28% of self-reported Spanish speakers have reported non-Hispanic ethnicity (Veltman, 1988).
Exactly who are these people? What can explain a pattern that contradicts other available evidence? How reliable are the data? Thus far the Census Bureau has released insufficient information from 2000 to enable a thorough analysis. Yet some clues may be found in the rather sketchy questions it asks Americans about their language usage. These appear on the 'long form' mailed to one in six households:
Does this person [age 5 or older] speak a language other than English at home? [If so] What is this language? [and] How well does this person speak English – very well, well, not well, [or] not at all?
The 2000 census forms were translated into Spanish, Chinese, Korean, Tagalog, andVietnamese, and 'assistance guides' were provided in 49 other languages. No explanations of the language questions were provided, however, even in English. While subjectivism will never be eliminated from self-reported data, it can be minimized by a survey that is clear and specific. Unfortunately, there is little clarity or specificity here.
What does it mean to 'speak a language other than English at home?' How well, how often, and how exclusively must one use the language? Is it enough to utter a few words on special occasions? Or is it necessary to speak the language on a daily basis? Should skills be sufficient to converse with a native speaker or merely to order a meal in a restaurant? What answer would be appropriate for persons with a mother tongue other than English who live alone or with English-only speakers? How about students practicing a foreign language at home?
None of the answers is self-evident, given the lack of context. There is no explanation of why Americans are being asked about their language behavior, or of how the information will be used. Standards are lacking as well. The long form provides no gauge for measuring degrees of English-speaking ability.
How does one choose among the categories well, not well, and very well? Should literacy be considered? Conversational fluency? Knowledge of grammar and vocabulary? Foreign accents? Are native speakers of English (who are not asked this question) all assumed to speak the language very well, regardless of the errors that many make in standard English?
Without guidance on such matters, respondents are less likely to provide accurate answers about their language usage. Yet US census officials offer little help, unlike their counterparts in other countries. Canada's 1996 census, for example, came with a comprehensive guide. It also featured more detailed and less ambiguous questions:
Can this person speak English or French well enough to carry on a conversation? What language(s), other than English or French, can this person speak well enough to carry on a conversation? What language does this person speak most often at home? What is the language that this person first learned at home in childhood and still understands ?' [emphasis in original]
If the objective is to count speakers of minority languages, rate their oral skills in the nation's dominant language(s), and plot trends of linguistic assimilation, these are useful questions. Such a survey is unlikely to elicit large numbers of unintended responses, for example, from people who claim to speak Spanish at home when they are doing homework for Spanish class or giving simple instructions to Latino gardeners.
The US census questions, by contrast, tend to invite exaggerated assessments of non-English proficiency and usage. These probably account for a sizable percentage of the estimated 7.9 million non-Latinos nationwide who reported themselves as Spanish speakers in 2000.
No doubt the minority language count is further inflated by responses shaped more by ethnic identification than by actual behavior. This pattern is especially evident among Native Americans. Many tribal members still report speaking tribal tongues that are, by other measures, eroding at alarming rates. A Navajo Nation study, for example, found that barely 32% of kindergartners on the reservation were proficient in Navajo, while 73% were proficient in English – a reflection of language usage in their families (Holm, 1993). Yet in the 2000 census, three out of four reservation residents reported speaking Navajo at home. On the neighboring Hopi reservation, where language loss is considerably more advanced, six in ten said they spoke Hopi.
For many Native Americans, saving endangered languages from extinction is seen as a sacred duty. It is not hard to understand how such feelings might encourage respondents to overstate their usage of Navajo or Hopi. Ironically, such assessments are certain to diminish the magnitude of this crisis in official statistics and thus in the minds of policymakers.
A more precise questionnaire, less open to subjective responses, could surely elicit more accurate information. It is disappointing that the census has no plans to revise the current questions on language, which have been used since 1980.
'It's a problem of real estate,' explains Rosalind Bruno, the census official who oversees the language survey. The long form is so crowded with queries on various subjects – 53 headings in all – that space is at a premium. But this is also a question of priorities. Is an accurate language count less important than, say, surveying the hour at which Americans leave for work or the amount they pay in condominium fees? Members of Congress, who exercise tight control over census operations, seem to think so.
Nevertheless, Congress relies increasingly on language data in funding federal programs. The most recent example is the No Child Left Behind Act (2002), which created a formula grant system to subsidize the instruction of English language learners. Allocations for states and school districts (totaling $665 million this year) will depend on their populations of English language learners (ELLs) aged 5–17, as reported by the census. This is problematic for several reasons:
Respondents to the census questionnaire, generally the parents of these students, have no training in language assessment or, as noted above, any official guidance whatsoever.
The English skills that children need to succeed in school include not just speaking, but also listening, reading, and writing.
In the past, there have been large and unexplained disparities between the number of ELLs served by schools and the number counted by the census. In 1990, Florida schools identified 83,937 ELL students, while the census reported 113,441. By contrast, Michigan schools identified 37,112 while the census reported 27,815 (for complete figures, see Crawford, 1997). Relying on census data in distributing Federal funds could have the perverse effect of rewarding states that underserve ELLs while penalizing those that do a more thorough job.
To get an accurate count of these students, self-reports need to be aligned with more objective assessments. This is no simple matter. Based on a 1982 study, the English Language Proficiency Survey, the Census Bureau concluded that a majority of respondents exaggerate their English proficiency. Therefore, in estimating the size of the limited-English-proficient (LEP) population, it decided to include all those who say they speak English 'well,' 'not well,' or 'not at all.' Only those who answer 'very well' are deemed to be fluent in English (MacArthur, 1993). The validity of this correlation was questioned in the 1980s (Waggoner, 1986) and, considering the more diverse immigrant population today, its application needs to be re-examined. Certainly, it is counterintuitive to believe that students who say they speak English well have 'difficulty with English.' That, however, is the reigning assumption.
In a time of anxiety about the assimilation of immigrants, perhaps it is no accident that the census tends to overstate the number of minority language speakers while understating their proficiency in English. These findings feed the perception that English is 'threatened' and needs 'legal protection' (US English, 1987). It is obvious that some members of Congress like to foster such views. Anglo-paranoia might be alleviated, however, if the census looked at the other side of the coin: the loss of heritage languages not only among Native Americans but also among immigrants.
The extent of that loss is illustrated by the Children of Immigrants Longitudinal Study, a decade-long research project involving about 5,000 students of 77 nationalities in South Florida and Southern California. The subjects were 'second generation' immigrants, defined as US-born with at least one foreign-born parent, or foreign-born with at least five years' residence in the United States. The children were 8th or 9th graders when surveyed, with a mean age of 14. Self-reports of English proficiency were confirmed by objective assessments, and the news was encouraging – in part. More than nine out of ten students said they could speak, understand, read, and write English 'well'; nearly two out of three said they knew the language 'very well' (see Table 2).
It was a different story with the heritage language. Fewer than half of the students said they knew their parents' vernacular well and only 16% rated themselves as fluent. There were variations by ethnicity, with language loss more pronounced among Asians and Pacific Islanders than among Latinos. But all ethnic groups expressed a strong preference for English, which helps to explain why relatively few of the youths were fluent in both languages (Portes & Hao, 1998).
This pattern of assimilation is cause for concern, according to a follow-up phase of the study conducted three years later. Second-generation students who became fluent bilinguals reported better relations with their families, greater self-esteem, and higher educational aspirations than those who became English monolinguals (Portes & Hao, 2002). But fluent bilinguals remained the exception, not the rule. One reason is that very few US schools are encouraging immigrants to retain their heritage language. The researchers conclude:
While popular with the public at large, educational policies that promote complete linguistic assimilation contain hidden costs for these children, depriving them of a key social resource at a critical juncture in their lives. Family relations and personality development suffer accordingly. ... Cut these moorings and children are cast adrift in a uniform monolingual world. They, their families, and eventually the communities where they settle will have to pay the price. (Portes & Hao, 2002: 294)
Considering the prevalence and prestige of bilingualism in Miami, where many of the students grew up, their rate of monolingualism is remarkable. It should also be troubling to anyone who values the linguistic resources that immigrants could contribute to this country. Until the census sharpens its focus on language, most Americans won't even know what we are losing.
Monolingual and Proud of It
'If you live in America, you need to speak English.' According to a Los Angeles Times Poll (1998a), that was how three out of four voters explained their support for Proposition 227, the ballot initiative that dismantled most bilingual education programs in California. Many Arizonans cited the same reason for passing a similar measure last year (Proposition 203, 2000).
Ambiguous as it is, this rationale offers some clues about the way Americans think about language. No doubt, for some English-only enthusiasts, the statement has a patriotic subtext: one flag, one language. Rejecting bilingual education was away to 'send a message' that, in the United States, English and only English is appropriate for use in the public square.
Other voters merely seemed intent on restating the obvious. English is so dominant in this country that non-English speakers are at a huge disadvantage. Thus schools must not fail to teach English to children from minority language backgrounds. Students' life chances will depend to a large extent on the level of English literacy skills they achieve.
Immigrants have generally understood these truths more keenly than anyone, and have behaved accordingly. As the linguist Einar Haugen (1972) once observed, 'America's profusion of tongues has made her a modern Babel, but a Babel in reverse.' By their third generation in the United States, newcomers have typically adopted English as their usual language and abandoned their mother tongue.
There is no reason to think the historic pattern has changed. Although the number of minority language speakers has grown dramatically in recent years, owing primarily to a liberalization of immigration laws in 1965, so too has their rate of acculturation. Census figures confirm the paradox. While a language other than English is now spoken at home by nearly one in five US residents, bilingualism is also on the rise. A century ago, the proportion of non-English speakers was 4.5 times as large as it is today, and in certain states the disparity was considerably larger (see Table 1). As the US population becomes increasingly diverse, linguistic assimilation seems to be progressing rapidly by historical standards.
Excerpted from "Advocating for English Learners"
Copyright © 2008 James Crawford.
Excerpted by permission of Multilingual Matters.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents1. Introduction2. Making Sense of Census 20003. Monolingual and Proud of It4. Heritage Languages in America: Tapping a “Hidden” Resource5. Plus ca change...6. Concern or Intolerance: What’s Driving the Anti-Bilingual Campaign?7. The Bilingual Education Story: Why Can’t the News Media Get It Right?8. Ten Common Fallacies about Bilingual Education9. Agenda for Inaction10. “Accountability” vs. Science in the Bilingual Education Debate11. Hard Sell: Why Is Bilingual Education So Unpopular with the American Public?12. Has Two-Way Been Oversold?13. Surviving the English Only Assault14. Official English Legislation: Bad for Civil Rights, Bad for America’s Interests, and Even Bad for English15. The Bilingual Education Act, 1968-2002: An Obituary16. No Child Left Behind: Misguided Approach to School Accountability for English Language Learners17. A Diminished Vision of Civil Rights18. The Decline of Bilingual Education: How To Reverse a Troubling Trend?19. Loose Ends in a Tattered Fabric: The Inconsistency of Language Rights in the USA