The great Italian novelist, Alberto Moravia, is said to have remarked in dismay that in these times even the illiterates read and write. He was probably referring to his critics, of course, but his observation is perfectly correct. The history of literacy in the West is one of increased expectations and increased performance. Once expectations for high levels of literacy were held for only a small elite of the citizenry. After the Reformation expectations were held for all of society to achieve modest levels of literacy. Now we seek the previously unimaginable goal of high levels of critical literacy for all of our society (Calfee & Patrick, 1996; Resnick & Resnick, 1977). This is a worthy goal, no doubt, but perhaps one that fails to recognize the enormity of individual differences and the environmental consequences in our society of widespread poverty and large numbers of mentally undemanding jobs. So we have upped the standards by which we judge literacy, and against these newer standards-the highest the world has ever tried to reach for the general population-some find the U.S. public schools to be deficient. I am not of this group. I have argued repeatedly that too much is made of the alleged deficiencies of American public schools, while the genuine achievements of that system for most Americans are really quite remarkable (Berliner & Biddle, 1995). To me the data suggest that the U.S. public schools have done a good job of producing large numbers of literate Americans should that level of literacy be needed in their lives.

Test makers and test interpreters often seem to confuse typical with maximum literacy performance. The typical performance is found in one's daily encounters with written communication and with the tested versions of literacy that are divorced from the lives of the people being tested. Maximum literacy performance is what would be obtained were people motivated to perform to the best of their abilities in situations about which they have some background and familiarity. This distinction between typical performance and maximum performance in literacy tasks will be needed to interpret some of the literature associated with the first point of this essay, namely, that the extant data suggest that our public schools have not failed to develop high levels of literacy. Rather, in the aggregate, our public schools have taught many of their students well.

The Literacy of Adult Americans

Literacy in America can be examined through the well regarded National Adult Literacy Study (NALS) (Kirsch, Jungeblut, Jenkins, & Kolstad, 1993) conducted by the Educational Testing Service (ETS) for the National Center for Education Statistics (NCES). The press conference announcing the release of this report resulted in a front page story in the New York Times with a headline that read "Half of Adults in US Lack Reading and Math Ability" (Celis, 1993). The Washington Post reported the story also; its headline was "Literacy of 90 Million Is Deficient" (Jordon, 1993). The nation's weekly newsmagazines joined hundreds of other newspapers to decry the state of literacy in America and the failure of public education. This was frightening news from the newspapers and magazines of record.

Let us look more closely at the study, however. First, the tests of prose, document, and quantitative literacy were just that, formal tests, and therefore not part of the daily lives of the people involved. The literacy skills of this sample of 26,000 Americans were not evaluated in the contexts of their everyday lives. Yet cognitive anthropologists and others have informed us that the tested performance of individuals does not always fairly represent the skills those individuals possess when those skills are actually required in situations with which they are involved on a regular basis. The NALS test items were creative and intended to be relevant to the real life activities engaged in by the sample. Nevertheless, these items were still test items, and formal testing can be a strange activity for those out of school, as was true of most people in this sample. Literacy is best thought of as instrumental and cannot be understood when it is removed and abstracted from the environments in which it operates. The theorists Luria and Vygotsky, along with the empirical work of Lave (1988), warn us that literacy should never be thought about outside the sociohistorical contexts-the material contexts-in which literacy operates.

Second, the test from which we deduce the level of different literacy skills of the nation was 15 minutes in length (Kirsch, Jungeblut, Jenkins, & Kolstad, 1993, p. 5). Though item sampling is sensible from a psychometric standpoint, a 15-minute test cannot provide an accurate assessment of literacy skills for an individual who is not used to formal testing. Furthermore, the 15 minute exam of just a few items yielded scores between 0-500. Each respondent got only a few items which had been given a scale value, say 258 or 420, and the performance on just these few items was assessed. A national profile for different demographic groups could be determined, but data on the literacy skills of individuals were never possible. I would hypothesize that the combination of a 15-minute formal test and the lack of a measure of literacy in context led to a large underestimation of the true level of literacy of the individuals that comprised the sample.

For simplicity I will discuss here only the prose literacy scale, though almost identical results were found for the document and quantitative literacy scales. In each area to be assessed, five levels of ability were created. Level 1, scoring 0-225, and Level 2, scoring 226-275 were deemed inadequate by the test developers and those that scored at these levels were deemed deficient in literacy by the press. By extrapolation from the sample to the general population, almost 50% of the population of American adults were then declared deficient in literacy. This was the basis for the broad assertions made by the media.

A literacy Level of 3 was considered to be adequate, but only 3% of the American population were found to be highly literate (Level 5). So, according to the study, we are a nation with hardly anybody really good at understanding complex prose. That is, around 97% of the adults in America are unable to "interpret a brief phrase from a lengthy news article" (Kirsch, Jungeblut, Jenkins, & Kolstad, 1993, p. 11). Figures like these should have rated a warning. Commenting on the report, Secretary Riley (U.S. Department of Education, 1993) said that the schools had to do better because illiteracy was causing poverty. This is a possible distortion of the data, to be cornmented on below.

What do we know of those in Levels 1 and 2? They are much more frequently poor and minorities, and they are not as often white, although almost half of the whites were in Levels 1 and 2 (p. 113). Scores on the test also revealed that around half the 16- to 54-year-olds scored at Level 3 or over; but the 55- to 64-year-olds in the nation, scored on average much lower. Fifty-seven percent of these people scored at Level 1 or 2, implying, I think, that only a little over 40% of the 55- to 64-year-olds in the nation have the ability to make sense of the documents they read. This sounds suspicious to me. But they are not America's worst problem because fully 76% of those over 65 years of age were classified as severely deficient in literacy skills (p. 116).

A little common sense would have gone a long way in dismissing the charges of widespread literacy problems in the U.S. If journalists had looked closely at the report, they would have discovered that among those in Levels 1 and 2 were: 67% of all those in the entire sample who were physically or mentally impaired; 80% of those in the entire sample who were visually impaired; 66% of those in the entire sample who had hearing difficulties; 80% of those in the entire sample who had leaming disabilities; 72% of those in the entire sample who had a mental or emotional condition; 90% of those in the entire sample who were mentally retarded; 79% of those in the entire sample who had speech disabilities; 74% of those in the entire sample who had physical disabilities; and 70% of all those in the sample who had long-term illnesses of six months or more (p. 135). Moreover, 25% of those who scored at the lowest level of prose literacy were not born in this country, presumably not native speakers of English (p. 119). The upshot of all this is that the two lowest levels of tested prose literacy, particularly Level 1, consisted of people who have overwhelming health problems or who are disproportionately recent immigrants. While the health of these people should be a concern for the nation, their performance should not be used to judge how well public school graduates are doing in terms of ordinary and reasonably healthy young men and woman, that is, those who do not suffer from the devastating effects of illness, poverty, and the possible stress of immigration.

One inference about how America's public schools may be doing can be obtained if we compare recent immigrants in the sample with those of similar background who were born in the U.S. For example, we find that those Americans born in Central or South America scored an average of 187, a very low Level 1 score, indicating a very low level of literacy. But those with Central or South American parentage who were born in the U.S. scored an average of 281-a score at Level 3, demonstrating adequate levels of literacy in English, according to the researchers. This difference in scores between generations provides a very different picture of literacy and schooling in America than the picture suggested by the news media and by some of our government officials. Obviously, the larger the number of recent immigrants, the larger the number of people who are likely not to do well in a formal test of English literacy. In recent years this country has had a large surge of immigration, primarily from non-English speaking origins. Pooled data containing such individuals are likely, therefore, to underestimate the literacy levels of others in the pool, particularly those born and educated in the U.S. In fact, the possibility that underestimates of American literacy might have taken place was not commented on when the NALS report was released by the goverriment.

Though the report was interpreted to mean that 90 million, or half the adults in America have limited reading ability, 54% of those at Level 1 read a newspaper every day or a few times a week, and 83% of those at Level 2 read a newspaper every day or a few times a week. The first meaning of literacy in the dictionary is "the ability to read and write."

Thus one should suspect a problem when over half of one of the lower level groups and about four-fifths of the other lower level group are reading newspapers on a regular basis. The survey that accompanied the 15-minute test revealed that about 93% of those with limited reading ability who read newspapers read the editorials regularly, while about 77% regularly read the home and fashion sections. It is therefore quite possible that many of the individuals at Level 1 or 2 read above their alleged level of literacy. Moreover, while the researchers declared these people to be greatly impaired, and the press declared them deficient in literacy, over two-thirds of those at Level 1 and over 95% of those at Level 2 reported that they read and write well. Something was clearly amiss.

Summarizing the NALS data

This analysis of the nation's largest contemporary study of literacy leads me to several conclusions. I think Secretary Riley's interpretation of the NALS data is wrong and therefore misdirects our attention from what I think are the real problems of America. It may not be illiteracy that causes poverty, but rather that poverty causes illiteracy. And if that is so, schooling is not a solution to the problem. Jobs and the economic viability of families are. I also think the NALS data demonstrate vividly that the individuals scoring at Level 2 are really quite competent to perform any employment task they are familiar with or trained in. They certainly were revealed to be Moravian illiterates, the kind that read and write and are comfortable doing so. In fact, if the skills of the Level 2 people are vastly underestimated, as I think they are, then they should be reclassified as literate enough for the needs they and their employers have.

The danger in studies like this one is the suggestion that the measures of literacy represent a fixed trait. That is not the case at all. A measure of literacy tells us where a person is on a continuum between decoding with minimum comprehension and the critical literacy skills expected of those who successfully live in a world dominated by prose literacy--college professors, lawyers, scientists, and so forth. A measure of literacy is a measure of a developing skill, not a measure of a relatively fixed characteristic. It is more like a measure of the ability to use a computer than it is like a measure of visual acuity. It is not a stable trait but a developing one. Thus many people at Level 1 and 2 could acquire higher level skills should their life circumstances require it. If that is so, then the category of Americans with adequate literacy could constitute an additional 27%--all those in Level 2-and comprise about 80% of the representative national sample (p. 113). By my standards that would not then be a bad record of achievement for an industrialized nation with a high rate of poverty and immigration.

My interpretation of NALS is that somewhere around 80% of Americas' adults are literate enough for the vast majority of the tasks that they must act thoughtfully about in their environments. We must remember that this is a nation where the economy has become primarily service oriented. It is an economy where the hospitality industry employs more people than any other segment of the market; where a retailer such as Wal-Mart is one of the largest employers in the nation; where there is a desperate need for more home care aides and janitors, and there is no shortage of mathematicians, scientists, and engineers. This is not a nation that requires the highest levels of critical literacy from all its citizens as they perform their daily work. Rather, we seem to require increasing amounts of domain and job-specific prose, document, quantitative, and technological literacy, the kinds of literacies that are taught on the job and not in schools.

Literacy and Schooling

But what about the schools and literacy? How have they done? The critics of the public school curriculum with respect to academic achievement love to point to the poor performance of our youth on many measures, though as Berliner and Biddle (1995) have made clear, most of the critics have little data on which to base their criticisms. The National Assessment of Educational Progress (NAEP) tests provide data relevant to the issues of literacy on which we focus. The NAEP tests of reading and writing are among the best the nation has produced for measuring the performance of our 9-, 13- and 17-year-olds over time. NAEP designers have scaled the developing performance of our youth in the way the NALS designers did. There are five levels in the reading test and the middle level, corresponding to a score of 250, is described as follows:

Readers at this level use intermediate skills and strategies to search for, locate, and organize the information they find in relatively lengthy passages and can recogniize paraphrases of what they have read. They can also make inferences and reach generalizations about main ideas and author purpose from passages dealing with literature, science, and social studies. Performance at this level suggests the ability to understand specific or sequentially related information. (Condition of Education, 1994, p. 195)

Readers at this level can understand complicated literary and informational passages, including material about topics they study at school.... Performance at this level suggests the ability to find, understand, summarize and explain relatively complicated information. (p. 195)

.... can locate and interpret facts from simple paragraphs, stories and news articles. In addition they can combine ideas and make inferences based on short uncomplicated passages. Performance at this level suggests the ability to understand specific or sequentially related information. (p. 196)

Writing scores on the NAEP are scaled, too (Condition of Education, 1994, p. 201). Level 250 is described as follows: "Writing at this level tended to be more focused and clear, containing enough development and detail likely to accomplish the assigned task successfully." Level 300 is described as writing that "tended to be complete and to contain sufficient information to accomplish the basic [writing] task." How did we do? Seventy-five percent of eighth grade students met or exceeded Level 250, and 25% of the eighth graders met or exceeded Level 300. Almost 90% of our 11th graders met or exceeded the 250 level of writing, showing ability to write focused, clear responses. Moreover, about 36% met or exceeded the 300 level-an advanced level of written communication. Furthermore, when the essays of the eleventh graders were examined, only 5% of the students had run-on sentences, only 2% showed misspellings of words, only 1% made errors of the wrong word choice, and only 1% made capitalization errors. For a public educational system, with great economic disparities in school funding and family wealth, these are admirable results. From my perspective, we should not bother to set as a national goal the overall improvement of reading and writing literacy, for it is already quite high. I believe our first priority should be to equalize educational resources so that all students have the same opportunities to acquire reading and writing skills.

Conclusion

Because our expectations are now so high, we forget that nowadays, even those classified with minimal reading and writing skills can read and write. In fact, according the NALS and NAEP data, this nation is blessed with youth and adults who demonstrate remarkably high levels of literacy. Our schools seem to be doing a terrific job, in aggregate. But literacy skills are not equally distributed across racial and class lines. And that is harmful not only to the individuals affected but to the nation as well. This is America's continuing problem. And the solution to low literacy rates among some groups of individuals may lie outside of educational institutions. Attainment of high levels of literacy is likely to be related to the nature of employment and the chances for a family to live in dignity in a given society. I believe low rates of literacy among segments of the population are likely to be more indicative of failed social and economic programs than of poor schooling.

References

Berliner, D.C., & Biddle, B.J. (1995). The manufactured crisis. Reading, MA: Addison Wesley.

Calfee, R . C., & Patrick, C.L. (1996). Teach our children well. Stanford, CA: Stanford University Alumni Association.

Celis, W. (1993, September 9). Study says half of adults in U.S. lack reading and math abilities. The New York Times, pp. Al, A22.

Jordon, H. (1993, September 8, 1993). Literacy of 90 million is deficient. Washington Post, pp. Al, A15.

Kirsch, I.S., Jungeblut, A., Jenkins, L., & Kolstad, A. (1993). Adult literacy in America. Princeton, NJ: Educational Testing Service.

Lave, J. (1988). Cognition in practice: Mind, mathematics, and culture in everyday life. New York: Cambridge University Press.

Organization for Economic Cooperation & Development. (1995). Education at a glance: OECD indicators. Paris: Author.

Resnick, D.P., & Resnick, L.B. (1977). The nature of literacy: An historical exploration. Harvard Education Review, 47, 370-385.

U.S. Department of Education. (1993, September 8). The national adult literacy study [Press briefing]. Washington, DC: Author.

U.S. Department of Education. (1994). The condition of education, 1994. Washington, DC: National Center for Education Statistics.