POQ69(1).book(nfi001.fm) Public Opinion Quarterly, Vol. 69, No. 1, Spring 2005, pp. 2–29 doi:10.1093 / poq / nfi001 © The Author 2005. Published by Oxford University Press on behalf of the American Association for Public Opinion Research. All rights reserved. For permissions, please e-mail: journals.permissions@oupjournals.org. ELITE REVISIONISTS AND POPULAR BELIEFS CHRISTOPHER COLUMBUS, HERO OR VILLAIN? HOWARD SCHUMAN BARRY SCHWARTZ HANNAH D’ARCY Abstract According to revisionist historians and American Indian activists, Christopher Columbus deserves condemnation for having brought slavery, disease, and death to America’s indigenous peoples. We ask whether the general public’s beliefs about Columbus show signs of reflecting these critical accounts, which increased markedly as the 1992 Quincentenary approached. Our national surveys, using several different question wordings, indicate that most Americans continue to admire Columbus because, as tradition puts it, “he discovered America,” though only a small number of mainly older respondents speak of him in the heroic terms common in earlier years. At the same time, the percentage of Americans who reject traditional beliefs about Columbus is also small and is divided between those who simply acknowledge the priority of Indians as the “First Americans” and those who go further to view Columbus as a villain. The latter group of respondents, we find, show a critical stance toward modal American beliefs much more broadly. We also analyze American history school textbooks for evidence of influence from revisionist writings, and we consider representations of Columbus in the mass media as well. Revisionist history can be seen as one consequence of the “minority rights revolution” that began after World War II and has achieved considerable success, but the endurance of Columbus’s reputation—to a considerable extent even among the HOWARD SCHUMAN is a research scientist and professor emeritus at the University of Michigan. BARRY SCHWARTZ is a professor emeritus at the University of Georgia. HANNAH D’ARCY is an independent statistical consultant who previously worked for the University of Michigan’s Center for Statistical Consultation and Research. We thank Lawrence Bobo and Stanley Presser for stim- ulating us to do multiple validations, and we are grateful to the editor of Public Opinion Quarterly for recommendations that substantially improved our final presentation. In addition, we are much indebted for help during the course of the research to Virginia Hopcroft, Government Documents Librarian at Bowdoin College; Maria Krysan, University of Illinois at Chicago; Alyssa Miller, Evanston, Illinois; and Irina Poznansky, Departmental Librarian, Teachers College, Gottesman Libraries, Columbia University. Support for the research was drawn in part from a National Science Foundation grant (SES-0001844). Address correspondence to Howard Schuman; e-mail: hschuman@umich.edu. Elite Revisionists and Popular Beliefs 3 minorities who have the least reason to respect him—raises important questions about the inertia of tradition, the politics of collective memory, and the difference between elite and popular beliefs. The revolution in minority rights over the past half century has not only changed the attitudes of the American public regarding race, gender, and other social divisions, but has also spurred attempts to revise beliefs about important individuals and events from the past. For example, Abraham Lincoln is now viewed less as the savior of the Union—the emphasis during the Civil War and for more than half a century afterward (Blight 2001)—and much more for his actions in ending slavery (Schwartz and Schuman, forth- coming). A far more radical attempt to change American collective memory took place during the 1980s and 1990s: according to recent editions of the Columbia Encyclopedia,“the image of [Christopher Columbus] as a hero was tarnished by criticism from Native Americans and revisionist historians . . . [and] his voyages [came to] symbolize the more brutal aspects of European colonization and represent the beginning of the destruction of Native American people and culture” (“Columbus” 1993, p. 605; 2004a, p. 629). The assault on the meaning of Columbus’s landfall in 1492, especially as its 500th anniversary approached in 1992, was the starting point for our research on the relation between elite and minority revisionist ideas, on the one hand, and the col- lective memories preserved in the general public, on the other.1 The fact that the widely consulted Columbia Encyclopedia includes such words in the final paragraph of its entry on Columbus indicates that revision- ist efforts have made an impact well beyond a limited set of writers and American Indian activists. Indeed, revisionist ideas about Columbus are appearing even in a variety of writings for young children, with titles such as Encounter (Yolen 1992), Discovering Christopher Columbus: How History Is Invented (Pelta 1991); and Who Really Discovered America? (Hart 2001; Krensky 1987). It is also clear that attempts at the time of the 1992 Quincen- tenary to reinvigorate the traditional heroic view of Columbus were unsuc- cessful, as recounted in detail by two knowledgeable observers in the book Sinking Columbus (Summerhill and Williams 2000). Thus, it seemed possi- ble that the picture of America’s founding event, which developed over some two hundred years of commemorating Columbus’s 1492 voyage and landing, has been turned upside down for a large part of the public who had previ- ously been taught to think of both the man and the date in triumphal terms. 1. Publication in 1980 of the English version of Maurice Halbwachs’s Collective Memory provides a convenient way to date the start of a great wave of research on collective memory. Halbwachs (1980, 1992) believed that writers, artists, and institutional representatives inter- pret history for the masses, but he never asked whether their interpretative activities deter- mined or expressed the beliefs of the communities of which they were members. For a review of collective memory studies, see Olick and Robbins 1998; and also early chapters in Wertsch 2002. 4 Schuman, Schwartz, and d’Arcy Yet we know from a number of past studies that intense debates at the level of elites and political activists often do not stir the larger public. For example, Stouffer (1955) found that relatively few Americans in the early 1950s appeared greatly worried about either an internal communist menace or a McCarthy- inspired threat to civil liberties, even though Stouffer’s research itself had been funded by a major elite foundation because of just such concerns. More generally, Converse (1964) delineated sharp differences between elite and mass political beliefs in levels of knowledge, sophistication, and organization across a wide range of political ideas and issues. In another twist, during the Vietnam War, the growing opposition to American military involvement had a quite different basis in the general public than it did for activists on college campuses (Schuman 1972). Thus, one cannot simply assume that revisionist efforts and Indian protests have had much effect at all on public thinking about the meaning of 1492. Figure 1. © The New Yorker Collection 2003 Lee Lorenz, from www. cartoonbank.com. All rights reserved. www.cartoonbank.com www.cartoonbank.com Elite Revisionists and Popular Beliefs 5 After briefly describing both traditional and revisionist ideas about Columbus, we will focus on popular beliefs as measured in national surveys of Americans at the end of the 1990s. Our results show little sign of revisionist influence, but at the same time, the data do not support a traditional heroic picture of Columbus, and this calls for a distinct and unanticipated interpretation. Furthermore, even for the small number of Americans who reject the long- standing positive image of Columbus as “the discoverer of America,” it proves useful to analyze the difference in ideological outlook between those who simply recognize the priority of Indians as the first Americans and those who go further and characterize Columbus as a villain. Our next step involves a content analysis of school history textbooks in order to consider the trans- mission of revisionist ideas from the elite level to standard writings about American history that can produce actual or future change in the beliefs of the public. In addition, we consider evidence from newspapers, television, and films to obtain a sense of what was available to the public about Columbus from the mass media around the time of the 1992 Quincentenary. Our research bears on larger issues of stability and change in collective memories and of what has come to be called the politics of memory (Gillis 1994). Before the 1980s, many scholars had written about politics and memory, but in the 1980s they began to write systematically about the politics of memory. The more radical strand of this approach, exemplified by John Bodnar (1992), assumes “reputational entrepreneurship” (efforts to shape the image and renown of another) to be centralized and controlled by government and other elite authorities. A second strand that may seem more credible depicts history and commemoration as emerging out of a context of cross-cutting coalitions, net- works, and enterprises. The creation of the Vietnam Veterans Memorial (Wagner- Pacifici and Schwartz 1991), the Enola Gay exhibit at the Smithsonian Institu- tion (Zolberg 1998), the naming of streets (Alderman 1996), the artistic depiction of the winning of the West (Dubin 1999)—such events show how commemo- ration articulates and sustains decentralized power networks. Moreover, of the chain that connects power diffusion and memory in these “second strand” writ- ings, the strongest link is the study of “history from below.” Reactions against conventional historiography emphasize the positive qualities and contributions of marginalized peoples, based on the conviction that every minority is entitled to interpret the past in its own way—to create “countermemories”—without the interference of Eurocentric interpretation.2 By the 1970s, these movements had penetrated the academy and reshaped understandings of the past by giving the victims of history unprecedented attention (Schlesinger 1991; Taylor 1994). That historical texts and commemorative symbols change as elites and power distributions change has been well documented, but no case raises more sharply the issue of what effects such changes have on popular beliefs than does Columbus’s landfall in 1492, one of the most symbolically important 2. On countermemories, see Foucault (1977) and Zerubavel (1995). 6 Schuman, Schwartz, and d’Arcy events in all of American history. Examination of what has happened to beliefs about Columbus within the general public poses questions about the results of revisionist efforts and illuminates the forces that sustain long-held collective memories when they are attacked.3 Collective Memories and Countermemories of Columbus The first important commemorations of the 1492 landfall occurred on its 300th anniversary in 1792, when the discovery of America was called “the greatest event in the history of mankind since the death of our Savior” (de Lancey 1893). Columbus was seen as “the solitary individual who challenged the unknown sea . . . [and] was ultimately betrayed by royal perfidy, [but] as a consequence of his vision and audacity, there was now a land free of kings, a vast continent for new beginnings” (Wilford 1991, p. 252). Although he had never reached the North American continent, nor indeed understood what it was he had come upon, the phrase “Columbus discovered America” increasingly merged the landing in the Bahamas in 1492 with the birth of the United States itself (Koch 1996). Columbus continued to be idealized through the nineteenth and early twentieth centuries, as signified by the installation in 1847 of the great John Vanderlyn painting, The Landing of Columbus on San Salvador, in the rotunda of the U.S. Capitol, and the placement of other commemorative symbols in towns and cities across the nation (Groseclose 1992). A multivolume biogra- phy by the celebrated author Washington Irving characterized Columbus in terms of “the grandeur of his views and the magnanimity of his spirit. . . . Instead of ravaging the new found countries . . . he sought . . . to civilize the natives” (Irving [1828] 1981, p. 565). The eminent nineteenth-century historian William F. Prescott wrote that it would be “difficult to point to a single blemish in [Columbus’s] moral character” (1874, vol. 3, p. 254), and the 1492 voyage reso- nated with divine purpose in Walt Whitman’s 1874 poem “Prayer of Columbus”: “a message from the Heavens . . . / These sped me on” ([1874] 1982). The 400th anniversary of the landfall in October 1892 was celebrated over a year-long period, starting with “a grand civic parade of more than 80,000 participants led by the President of the United States and including the entire Cabinet, the Supreme Court, and most of the Congress” (West and Kling 1989, pp. 56–57). The year culminated in the spectacular World’s Columbian Exposition in Chicago, which drew over 27 million visitors and “produced an 3. There are certainly other examples of radical change in the reputations of celebrated individu- als and events. Charles Lindbergh was treated to hero worship after his solo flight to Paris in 1927, but by the beginning of the 1940s his celebrity status had almost completely vanished (Boorstin 1962). Warren G. Harding was highly regarded while he was president, but after his death his prestige fell as his supporters disappeared and others condemned him for scandals that occurred under his watch (Fine 2001). Recent Israeli scholarship has transformed the mass suicide at Masada, once regarded by Israelis as an event worthy of solemn commemoration, into an object of indifference at best (Ben-Yehuda 1995). None of these examples, however, concerns a long- established collective memory under assault in quite the same way as is true for Columbus. Elite Revisionists and Popular Beliefs 7 unparalleled surge of creative energy that had an important influence . . . on the cultural values of the nation” (Columbia Encyclopedia 2000, p. 3108). The inaugural orator spoke of “the crowning gift to humanity from Columbus . . . in search of a great land”—identified with the United States—and the official history of the exposition declared Columbus “the greatest human benefactor of the human race” (Johnson 1897, vol. 1, p. 2). Having started as a symbol of American individualism and progress, Columbus then became an ethnic hero as well, with Italian Americans playing a major role in turning Columbus Day into a full federal holiday in 1968. But ironically, his enshrinement in the federal calendar occurred just as his reputation was caught “in a riptide of conflicting views of his life and his responsibility for almost every- thing” wrong that could be linked to 1492 (Wilford 1991, p. 247). In 1973 the geographer and historian Alfred Crosby wrote of the havoc produced by diseases that Europeans brought to Native Americans. A broader critical book in 1975 by historian Francis Jennings bore its thesis in its title: The Invasion of America. Probably the most widely read early attack appears in the opening chapter of Howard Zinn’s (1980) A People’s History of the United States, which depicts the events of 1492 from the standpoint of Indians and emphasizes their oppression by Columbus and his successors. Zinn’s book has sold more than a million copies (personal report by the author), is owned by approximately 4,000 libraries (World- cat database), and has been translated into Spanish and 11 other languages. Also important in terms of popular impact was James Loewen’s The Truth about Columbus, a detailed summary of revisionist thinking for students, published in the Quincentenary year of 1992. Much of the book then appeared three years later as a chapter in Loewen’s Lies My Teacher Told Me (1995), which has had sales of more than half a million copies (author’s report). Moreover, in addition to the major focus on injustice toward Indians, Columbus has been connected to the despoiling of the natural environment, “now threatening . . . the existence of the earth as we have known it and the greater proportion of the species, including the human” (Sale 1990, p. 4). There were counterattacks against revisionist critiques as well (e.g., Royal 1992), but even such defenses of Columbus’s reputation showed that he was no longer an undisputed hero.4 4. It is important to recognize that revisionist views of Columbus did not result mainly from the discovery of new facts, but from attention to and reappraisal of information already available. A hundred years earlier at the time of the 400th anniversary, Justin Winsor published a book that characterized Columbus quite negatively: for example, Columbus “had no pity for the misery of others . . . [consigning Indians] to the slave market” (Winsor 1891, pp. 505–6). Moreover, much of Winsor’s information came from a manuscript written in the early sixteenth century by a close observer of the Spanish colonization of the Americas, Bartolomé de las Casas (1974, 1992), who admired Columbus as a navigator but was highly critical of his and other Spaniards’ treatment of the Indians. (See Axtell 1995 and Phillips and Phillips 1992 on early criticisms of Spanish actions.) Even Samuel Eliot Morison’s (1942) widely acclaimed biography includes such nega- tive information, though the criticisms are overshadowed by the book’s focus on Columbus as a great mariner. We have not attempted to cover here the specialized scholarly writing stimulated by the Quincentenary, much of which is neither idealizing nor completely revisionist in tone and content. For comprehensive reviews, see Axtell 1992, 1995; Lunenfeld 1992. See also Zerubavel 1992 for a different perspective on the meaning of “Discovery of America.” 8 Schuman, Schwartz, and d’Arcy Even earlier than the main revisionist writings and at least as important was the influence of the civil rights movement during the 1960s. With the growing emphasis on black power and black identity, other minorities with long-held grievances against the white majority came to the fore (Rhea 1997; Skretney 2002). In particular, the rise of “red power” ideology at the end of the 1960s (Nagel 1995) challenged white views broadly, and one effect was to question the assumption that Columbus Day was an occasion for celebration. Instead, Columbus’s destruction of native peoples and culture was said to call for condemnation. Then, as Indian and scholarly critiques came together, rein- forced by the anticolonial sentiments that had developed in the wake of World War II, major organizations with a much wider reach began to express guilt over what Columbus represented. For example, the National Council of Churches, which includes 36 denominations with more than 50 million mem- bers, passed a lengthy resolution in 1990 that included, among other similar statements: “For the indigenous people of the Caribbean islands, Christopher Columbus’s invasion marked the beginning of slavery and their eventual genocide.” “For the Church,” the resolution announced, “this is not a time for celebration” (National Council of Churches 1990). Criticism of Columbus has also made its way into the mainstream media, for example, appearing promi- nently in an episode in 2003 of the popular television drama The Sopranos. The 500th anniversary in 1992 offered an opportunity for reviving a posi- tive image of Columbus. He could have been portrayed as the embodiment of individualistic and adventurous enterprise, as he had been in the nineteenth century (Phillips and Phillips 1992), perhaps now stressing a connection to the recent triumph of American capitalism over Soviet communism. But it was exactly in 1992 that revisionist criticism and the active protests of Native Americans peaked. Within elite groups that attempted to mount celebrations, there were conflicts, doubts, and trepidation over controversy, and few if any commemo- rations were successful (Summerhill and Williams 2000). A Smithsonian exhibit for the Quincentenary leaned over backward to present negative as well as positive views of Columbus and to focus less on the man than on diseases, foods, and other indirect effects of 1492—informative, but unlikely to inspire commemorative enthusiasm at all similar to that of 1892. “The most striking difference between the fourth and fifth Columbian centenaries,” one observer noted, “[was] that native Americans a century ago were relegated to the foot- notes while today they not only dominate the text but have begun to rewrite it” (Axtell 1992, p. 337). College campuses were major sites for protests against positive commemor- ation of Columbus. We located six college newspapers from October 1992: those at Bowdoin College, the University of Georgia, the University of Illinois in Chicago, the University of Michigan, the University of Oregon, and San Francisco State College. Each contained at least one article damning Columbus or reporting a local protest. Although hardly a random or large sample, the six are diverse enough to suggest that something similar probably occurred on Elite Revisionists and Popular Beliefs 9 many nationally known campuses at the time of the Quincentenary. In the years since 1992, campus demonstrations against Columbus Day have contin- ued, some led by American Indians and others by interested non-Indian students. A search of the Lexis-Nexis Academic Universe Campus News database identified 60 mentions of Columbus Day between 1997 and 2000, 55 of which dealt with campus protests of one sort or another. Thus, social scien- tists on campuses have been among those most fully exposed to revisionist criticisms and protests, a point we return to later when we consider the expect- ation that revisionist ideas are widely known and accepted. Collective Memory at the Individual Level To explore recent beliefs about Columbus, we asked a basic open-ended question to a national sample of 1,511 Americans in 1998, six years after the peak of criticism and protest during the 500th anniversary of the 1492 land- fall. In keeping with the assumption that collective memories are ordinarily passed from one generation to another, our question (after pilot work to check for its acceptance and comprehension) was: Suppose a nephew or niece about 14 years old had just heard some mention of Christopher Columbus and asked you to explain what Christopher Columbus had done. What would you say in just a few words? The indirect phrasing avoided having the question appear to be a threatening test of personal knowledge, and interviewers were instructed to reassure hesi- tant respondents that there were no right or wrong answers, “just whatever you would say to a young person to explain what Christopher Columbus had done.” Interviewers were to record the responses verbatim and to probe non- directively where clarification was needed.5 5. The question was asked toward the end of the University of Michigan Survey Research Center (SRC) Monthly Survey. Each month SRC uses random digit dial (RDD) sampling to obtain a tele- phone cross-section of approximately 300 new adults age 18 and older, with a then current response rate averaging 60 percent, using the response rate 4 calculation method described in American Association for Public Opinion Research (AAPOR) Standard Definitions (AAPOR 2000). The Monthly Survey also reinterviews approximately 200 cases from the RDD sample from six months earlier (all those who are willing to be reinterviewed, which averages 80 percent of the original samples). A description of the sample design can be found online at http:// www.umich.edu/~umsurvey (see also Curtin, Presser, and Singer [2000] on the lack of bias due to nonresponse). Our question was included in three successive months (September to November) to provide a total N of 1,511, using both RDD and reinterview respondents; the latter were also new to our inquiry. There are rarely differences found between the two components of the total sample, and our results hold throughout with controls for the RDD and reinterview samples, excepting one minor instance probably due to chance. Our data are not weighted for household selection or demographic census figures because weighting changes the percentages in table 1 by no more than 1 percent. (See also DuMouchel and Duncan 1983 on not weighting data for regres- sion analysis.) http://www.umich.edu/~umsurvey http://www.umich.edu/~umsurvey 10 Schuman, Schwartz, and d’Arcy After reviewing a sample of responses and taking into account our theoretical purposes, we developed the set of categories shown in table 1, along with the percentage of responses in each category.6 The categories are ordered from the most positive beliefs about Columbus (Heroic Traditional) to the most negative (Villainous Columbus):7 1. Heroic Traditional: These responses stated or implied that Columbus discovered America and also included something especially admirable about Columbus. For example: “[He] had the courage and enterprising spirit to go on uncharted territories.” 2. Simple Traditional: Most of these responses consisted essentially of the words “He discovered America,” and others were variants (e.g., added references to Columbus’s three ships). The responses differ from those coded “Heroic Traditional” in that they do not explicitly mention admir- able personal qualities of Columbus, though they are not at all negative. 3. Other Europeans: The remaining three categories in table 1 challenge the traditional view of Columbus as “the discoverer of America” in increasingly critical terms. The mildest and least novel category, 6. Omitted from table 1 are five nonsubstantive codes. The largest is Don’t Know (DK) (N = 74): respondents who said they did not know enough about Columbus to answer the question. A sec- ond omission (N = 60) consists of vague positive responses that were not codable into one of our main categories. The three other omitted categories are 54 respondents who gave uninterpretable responses or rejected the question; 12 who insisted they would send the youth to the library; and 6 who gave completely incorrect answers (e.g., “He was our president”). The mean education for each of these five categories is lower than for all the substantive categories in table 1, with the DK respondents the lowest of all (mean of 10.7 years, whereas the lowest educational level in table 1 is 13.8 for Simple Traditional responses). The total of the five nonsubstantive categories, 206, subtracted from 1,511, gives the base N of 1,305 in table 1. (That nearly 14 percent of the original sample seemed unable to give a substantively codable response reminds us that a small but non- trivial part of the population does not hold any meaningful beliefs about Columbus.) 7. Initial coding was done by one author, then all responses that were considered nontraditional, combined with a random selection of 300 traditional responses, were coded by a second person to assess inter-rater reliability, yielding Cohen’s Kappa of 0.86 (p < .001), which indicates very good agreement. Discrepancies typically involved borderline responses between adjacent catego- ries and were resolved by consensus. In addition, we repeated our main analysis using only responses for which there was complete agreement, and none of our conclusions changed. Table 1. Beliefs about Columbus Category Description N % 1 Heroic Traditional Columbus 81 6.2 2 Simple Traditional Columbus 1,105 84.7 3 Other Europeans 43 3.3 4 Indians Already Here 29 2.2 5 Villainous Columbus 47 3.6 Total 1,305 100.0 Elite Revisionists and Popular Beliefs 11 “Other Europeans,” refers to others who have been said to have reached the western hemisphere prior to Columbus. For example: “Actually the Vikings were here first.” Since high school texts going back at least to the 1930s noted evidence of pre-Columbian landings by Vikings, such responses do not reflect recent revisionist criticism of Columbus. For that reason, along with their small number, we will not focus on this category, though we continue to include it when reporting tabular results. 4. Indians Already Here: These responses provided a clear critique of the traditional view of Columbus with the assertion that he could not have “discovered” America because the people he named Indians were already here. For example: “They say he discovered America but he didn’t, the Native Americans did.” Such responses reject the Simple Traditional view of Columbus as “the discoverer of America,” but they offer no explicit criticism of him or of his treatment of Indians, which has been the main emphasis of the revi- sionist position. 5. Villainous Columbus: Finally, there were responses that not only rec- ognized the priority of the American Indians, but that also portrayed Columbus in ways consistent with attacks by revisionist historians and Indian activists. This was the most important category in terms of our theoretical focus on the effects of revisionist criticisms. For example: “[H]e met up with Native Americans and he slaughtered them.” (Occasional answers noted that Columbus was brave but then criticized him severely in ways that clearly fit the Villainous category. With such mixed responses, we coded in the more negative direction in order to obtain a maximum estimate of revisionist influence.) R E S U L T S Our first major finding was that 85 percent of this national sample gave Simple Traditional answers (category 2) that basically described Columbus as the “discoverer of America.” Only 6 percent were more laudatory and gave Heroic responses; at the other extreme, fewer than 4 percent character- ized Columbus in the Villainous terms advanced by revisionist writers and protestors, and another 2 percent acknowledged the priority of the Indians, for a total of just under 6 percent who held revisionist beliefs broadly defined. Because so few respondents were either clearly positive or clearly negative, we wondered if our survey question had failed to capture fuller sentiments about Columbus. We repeated the question in a new monthly survey (N = 126) 12 Schuman, Schwartz, and d’Arcy in August 2000, omitting the phrase “in just a few words,” in case those words had constrained answers unduly. But neither the overall distribution nor any of the categories changed significantly (p >.10) or even in tendency. It still seemed possible that our inquiry was deficient by not encouraging respon- dents to go beyond an initial remembered cliché from childhood about the “discovery of America.” Therefore, we repeated our initial question once more in October 2000 (N = 130) (again, without the “few words” phrase) and then pressed respondents harder with a mandatory follow-up inquiry: “Can you add anything else important about Columbus or about what he did?” To this additional question, more than half the respondents (N = 70) said they could add nothing more, but enough did respond further to change the table 1 distribution somewhat. It now had fewer Simple Traditional responses, though they still constituted 68 percent of the total. The largest increase was in the direction of Heroic responses (now 12 percent of the total); the main revision- ist category of Villainous Columbus increased to 8 percent; and the Indians Already Here code scarcely moved (now 3 percent). Thus, by pushing respon- dents quite hard to go beyond their initial response, we raised slightly the sev- eral nontraditional answers, but clearly positive (Heroic) or negative (Villainous) responses remained a very small minority. Since we were still not completely satisfied that our open-ended inquiry captured the full extent of revisionist influence on ordinary Americans, we decided to take a more direct approach aimed at clarifying the Simple Tradi- tional responses that constituted by far the largest category of answers and that seemed on their face positive, though not laudatory, about Columbus as the discoverer of America. To test whether such responses were in fact unin- fluenced by recent criticisms of Columbus, we repeated our standard open inquiry in February 2002, then followed it with a blunt closed question to all those classified as Simple Traditional responders: “Do you think young Americans should admire Christopher Columbus?” This emphasis on the key word “admire” failed to reduce positive views of Columbus appreciably: of the 478 respondents who gave Simple Traditional responses to the initial question in this new sample, the great majority, 81 percent, said “yes,” young Americans should admire Columbus; and if we restrict our focus to white respondents (N = 433), the figure is 87 percent. In addition, the small proportion who said “no” to the admiration question were asked: “Why not?” Of the 13 percent of white respondents who had said either “no, should not admire” or simply “I don’t know,” more than half gave reasons that did not reflect primary revisionist criticisms, for example: “If he hadn’t done it, someone else would have.” Thus, despite a leading question that invited respondents to indicate whatever reservations they might have about Columbus, only a handful showed signs of revisionist influence. In the analysis that fol- lows, we work with the original large sample shown in table 1, treating the supplementary tests as not different enough to change analytic findings appreciably. Elite Revisionists and Popular Beliefs 13 In sum, beliefs of Americans about Columbus, assessed after the height of revisionist writing and protests, point to continued regard for him as “the discoverer of America,” which in turn is seen in positive terms. At the same time, since only a small percentage of Americans spontaneously attribute heroic qualities to Columbus in response to our basic open question, this also calls for explanation, the more so since revisionist criticism evidently had little impact on most people. Thus, we have two distinct issues to inves- tigate further: First, why is there so little evidence of effects on the American public of the many revisionist attacks on Columbus? Second, why is the continued positive belief in Columbus as discoverer of America not accompanied by more explicit characterizations of him in heroic terms? White Americans: Hypotheses, Findings, and Interpretations We will focus mainly on white Americans, since it is primarily for whites, not minorities, that revisionist critiques are written, though we also report briefly at a later point on minority respondents. Although most (86 percent) white Americans gave Simple Traditional answers that viewed Columbus simply as the discoverer of America, we expected the smaller categories in table 1 to show discernible relations to age or education or both. If there has been a decline in heroic characterizations of Columbus within American culture, the oldest respondents in our sample should con- tinue to express a more heroic view of Columbus that they had absorbed when growing up in an earlier era. Likewise, younger respondents should be more influenced by the recent revisionist attacks on Columbus, espe- cially if we take seriously Mannheim’s ([1928] 1952) emphasis on adoles- cence and early adulthood as critical ages for political learning. Greater education should also be associated with more critical views of Columbus, since awareness of revisionist ideas depends at least in part on reading or attending to serious media. The evidence for testing these hypotheses is provided in table 2, which reports results from a multinomial logistic regression of beliefs about Columbus on age and education. All five cate- gories from table 1 are included, with the numerous Simple Traditional responses as the reference category with which the other categories are compared.8 8. Age is treated as continuous, ages 18 and up. Education is treated as a five-level continuous variable: 0–11, 12, 13–15, 16, 17+ years of schooling. Race/ethnicity was self-reported and is used here, and also later in table 3, to identify white respondents. We neither hypothesized nor found significant associations involving gender or region and therefore do not report such data. 14 Schuman, Schwartz, and d’Arcy H E R O I C T R A D I T I O N A L R E S P O N S E S The most reliable finding in table 2 is that older cohorts are more likely than younger cohorts to hold a Heroic rather than a Simple Traditional view of Columbus (odds ratio > 1.25, p = .003). This provides evidence of a decline over time in honorific characterizations of Columbus. (An alternative inter- pretation in terms of “aging” lacks plausibility here, and the aging-conservatism assumption has been challenged more generally; e.g., Danigelis and Cutler 1991; Glenn 1974; Riley, Foner, and Waring 1988; see also Alwin 1997.) There is no relation of Heroic responses to education, nor any sign of an interaction of age with education in a separate analysis; thus, only cohort experience appears to be involved in this decline in the glorification of Columbus. Presumably, the relation would be even stronger if our sample included respondents who had grown up around the time of the 400th anni- versary in 1892. However, given the paucity of Villainous responses, the decline in Heroic responses is unlikely to be due to revisionist efforts, and a different explanation is called for. Schwartz (1998) has documented an erosion of historic reputations that affects collective memories of past U.S. leaders generally. In addition to a loss of trust in government that grows out of failures in the Vietnam and Water- gate periods, there appears to be a continuing decrease in historic reputations, perhaps promoted by television and now the Internet, which expose America’s historical narrative to more widely held skepticism than in earlier days (e.g., Stewart, Karlin, and Javerbaum 2004). The waning of spontaneous heroic characterizations of Columbus—despite a continuing belief in his being the Table 2. White Beliefs about Columbus by Age and Education NOTE.—This table is based on multinomial logistic regression, with Simple Traditional responses treated as the reference category. OR = odds ratio; CI = confidence interval. Odds ratios for age pertain to an increase of 10 years. Education is measured using five categories, as explained in the text; odds ratios pertain to an increase of one education level. The confidence intervals apply to the odds ratios, having been computed initially for the beta values and then exponentiated. Significant effects are highlighted by boldface. Overall table p-values for age and education are .004 and .042, respectively. Response Predictor OR 95% CI p Heroic Traditional Columbus Age 1.25 1.08, 1.45 .003 Education 1.14 0.92, 1.41 .238 Other Europeans Age 1.06 0.86, 1.31 .602 Education 1.09 0.80, 1.48 .591 Indians Already Here Age 0.74 0.51, 1.08 .121 Education 1.76 1.09, 2.82 .020 Villainous Columbus Age 0.82 0.65, 1.03 .093 Education 1.31 0.97, 1.77 .079 Elite Revisionists and Popular Beliefs 15 discoverer of America—fits well with a general diminution of past heroic rep- utations in the eyes of the larger public. R E S P O N S E S C R I T I C A L O F C O L U M B U S Since revisionist attacks on Columbus date from the 1970s and especially from the years leading up to the Quincentenary, it should be younger respondents who show the two more critical answers, Indians Already Here and, especially, Villainous Columbus. Both categories do reveal nonsignificant trends to be given more by younger cohorts, but contrary to our expectation, the relation appears slightly weaker rather than stronger for the more critical type, the Villainous category. Since the two categories are conceptually similar insofar as both assume that Indians were the first Americans, and since both show the same trends in relation to cohort, it is reasonable to conclude that younger cohorts are more likely to emphasize Indians as the first Americans (if the com- bined two categories are regressed on age, p = .03). But cohort does not account for the additional negative content of the Villainous responses. Similarly, greater education is related to giving both types of nontraditional responses, but slightly more so for Indians Already Here than for the more extreme Villainous answer.9 The problem of accounting for the highly negative nature of Villainous responses thus remains. Because of their extremely critical content, their tiny proportion in the total population, and their lack of a distinctive connection to cohort, education, and other demographic variables, we hypothesized that such answers draw on a more general negative attitude toward conventional American verities. We tested this hypothesis by using a standard question asking respondents their religious preference: Protestant, Catholic, Jewish, Other, or None.10 Those who say “none” are individuals who reject the widely accepted norm in the United States of claiming an attachment to some religious faith. Dwight Eisenhower famously stated the norm a month before his inauguration as the 34th U.S. president: “our form of government has no sense unless it is founded in a deeply religious faith, and I don’t care what it 9. We explored the cohort effect further by plotting both the Heroic category and the Indians Already Here/Villainous combination against the full cohort variable (ages 18 to 97), with educa- tion controlled. With very few exceptions, cohorts born before 1948 are more likely to give Heroic responses than those born after 1948, and those born after 1948 are more likely than those born before 1948 to give the combination of Indians Already Here and Villainous responses. Fur- ther, within neither grouping (pre-1948 or post-1948) do we detect any additional association with birth year, nor does changing the dividing point from 1948 by a year or two in either direction alter these conclusions. To interpret the pattern, note that a respondent who was born around 1948 reached early adulthood in the late 1960s—years that saw considerable political disenchantment and the development of a larger counterculture. It was a time when traditional symbols were chal- lenged, and also a time of growing attention to the concerns of minorities. 10. The religious preference question had been asked only of the RDD component of our sample: N = 857, of whom 625 are both white and included in the present analysis, with N = 72 for no preference and N = 553 for some stated preference. 16 Schuman, Schwartz, and d’Arcy is.”11 The hypothesis is clearly supported when we add self-reported religious preference to the regression reported in table 2, which also controls for age and education. Responses classified as Villainous Columbus show a very large and highly significant association in the predicted direction for the dichotomy of No Religious Preference versus all others (Protestant, Catholic, Jewish, Other combined): odds ratio = 7.75, p < .0001. Moreover, there is no relation of that dichotomy to any of the other Columbus categories in table 2 (p > .40 in all comparisons), so the impact of rejecting a religious preference is solely on the Villainous category. These findings provide strong evidence that among white respondents, the characterization of Columbus as villainous draws on a larger receptivity to non-normative beliefs generally, presumably in a liberal or radical direction. Additional evidence for both the liberal and unconventional meaning of expressing “no religious preference” comes from the General Social Survey (GSS), which regularly asks for religious preference and also includes various questions about social and political issues (Davis, Smith, and Marsden 2001). We located four GSS questions where only a small proportion of respondents (15 percent or fewer) deviate from the modal response in what would usually be considered a liberal, radical, or at least unconventional direction (GSS mnemonics in parentheses): 1. protesting a government action can legitimately involve occupying government offices to stop work there for several days (PROTEST4); 2. the right of a person to end his/her life if tired of living and ready to die (SUICIDE4); 3. not very proud or not proud at all of American history (PROUDHIS); and 4. adultery is wrong only sometimes or not wrong at all (XMARSEX). On each of these questions, No Religious Preference respondents answer in the deviant direction shown here, with the difference in each case highly significant (p < .001).12 T H E V I E W S O F A M E R I C A N M I N O R I T I E S Indians should be the single group most likely to perceive Columbus in villainous terms (e.g., Deloria 1969), and despite their very small number in a national sample, that is clearly the case shown in table 3: 42 percent give Villainous 11. Emphasis added. Remarks made at the Freedom Foundation, New York City, December 22, 1952, as obtained from the Eisenhower Presidential Library, Abilene, KS. 12. These results are based on whatever years a particular attitude question had been asked by the GSS, with comparisons limited to white respondents to match the analysis of the Columbus data. In addition, on a 0 to 100 point scale on which respondents rate their feelings about “liberals,” No Religious Preference respondents (9 percent of the total cumulative sample) average 7 points higher than Some Preference respondents (Protestant, Catholic, and Jewish combined). The oppo- site holds for ratings of “conservatives,” with No Preference respondents averaging 10 points lower than those with Some Preference. For both comparisons, p < .001. Elite Revisionists and Popular Beliefs 17 responses, as compared to less than 4 percent of any of the other three groups (most importantly, Indians versus whites: p < .001). In addition, African-Ameri- cans seemed unlikely to identify with white heroes, and indeed they are less likely than whites to give Traditional descriptions of Columbus (p < .001), though their distinctive choice emphasizes Indians as the first Americans. Both of these differ- ences fit expectation, but it is important to note that half the Indian sample and some 84 percent of the black sample fall into the first two traditional positive cate- gories; thus, even for the two racial/ethnic groups most likely to be receptive to revisionist ideas, the impact of such ideas has been far from complete. Consideration of Hispanic responses led in a different direction, for there was considerable Hispanic involvement in planning the Quincentenary celebration because of its connection to Spain. Furthermore, both de la Garza, Falcon, and Garcia (1996) and Rosenzweig and Thelen (1998) report that Mexican-Americans (the only Hispanic group they studied) are at least as conventionally patriotic as whites and are quite different in this respect from blacks and American Indians. Indeed, in table 3 we see Hispanic respondents’ holding traditional rather than critical views of Columbus, and they even show a nonsignificant trend (p = .13) toward a higher percentage of Heroic responses than whites.13 The Transmission of Revisionist Influence to the Public In order to explore further the connection between elite revisionist ideas and our individual-level findings, and at the same time to estimate the likely influ- ence of such ideas in the future, we examined a major institutional vehicle through which new ideas reach the general public: American history text- books. If we accept Yerushalmi’s (1982, p. xv) thesis that “collective memory 13. Omitted from table 3 are 26 respondents categorized as “Asian,” since we had no hypothesis about their answers. Their distribution turns out to be statistically indistinguishable from the white distribution in table 3. (They are included in table 1, but not in later analyses.) Table 3. Beliefs about Columbus by Race/Ethnicity White (%) Hispanic (%) Black (%) Indian (%) 1. Heroic Traditional Columbus 6.4 11.0 1.8 0.0 2. Simple Traditional View 86.1 78.1 81.8 50.0 3. Other Europeans 3.0 4.1 3.6 8.3 4. Indians Already Here 1.4 4.1 9.1 0.0 5. Villainous Columbus 3.2 2.7 3.6 41.7 Total 100 100 100 100 (N) (1,069) (73) (110) (12) 18 Schuman, Schwartz, and d’Arcy is not a metaphor but a social reality transmitted and sustained through the conscious efforts and institutions of the group,” then such texts are probably the single most important medium through which a society transmits and legitimizes what to believe about the past. We collected the pages on Christopher Columbus from all high school his- tory textbooks dating from the mid-1940s through the 1990s (N = 55) that could be located in the Teachers College Library at Columbia University, which maintains the most complete collection of textbooks in the country.14 We used the index in each book to find all pages dealing with Columbus and then identified two kinds of statements: evaluations of Columbus or his actions and evaluations of the American Indians he met. The statements were copied into a table and then coded as follows:15 Pos: Columbus/Indians characterized positively, as would be viewed today (e.g., Columbus described as brave or as the source of positive Ameri- can development; Indian culture described positively or Indians seen as the “first Americans”). Neg: Columbus/Indians characterized negatively, as would be viewed today (e.g., Columbus takes captured Indians to Spain; Indians called child- like or savage). Unclear: No clear characterization of Columbus/Indians is provided. The percentages of positive and negative codes for Columbus by decade are shown in table 4. For example, of the eleven books from the late 1940s and 1950s combined, ten or 91 percent were coded as having a positive statement about Columbus, three or 27 percent, a negative statement, and one or 9 per- 14. For a general assessment of American history textbooks, see FitzGerald (1979) and Loewen (1995); the former is organized chronologically and the latter topically, though neither uses sys- tematic coding to examine trends over time as we do here. Our analysis covers a span that begins well before what FitzGerald considers the most important period for change in content—the middle and late 1960s—and extends for more than two decades beyond the date of her book. 15. Overall agreement between two independent coders was 93 percent. Readers can obtain the full table of quotations and codes from the authors. Table 4. Results of Content Analysis of High School Textbooks: Evaluations of Columbus by Decade NOTE.—Since a book could be coded for both positive and negative evaluations, the percent- ages in a column may add to more than 100 percent. Year 1944–59 1960s 1970s 1980s 1990s Positive 91% 83% 17% 40% 80% Negative 27% 17% 42% 50% 50% Unclear 9% 17% 58% 30% 0% (N of books) (11) (12) (12) (10) (10) Elite Revisionists and Popular Beliefs 19 cent had no characterization. (We use the number of books as the base N in all calculations; percentages need not add to 100 since both positive and negative characterizations could occur in the same book.) The overall pattern for Columbus in table 4 appears curvilinear, starting off as predominantly posi- tive, moving in the 1970s to much more negative characterizations, and then recovering a more positive view in the 1980s and 1990s, though these two decades continue to include negative statements as well. Despite the small number of decades and limited number of books per decade, a logistic regres- sion of positively coded books (versus all others) on decade yields a highly significant quadratic effect (b = .82, p = .002), consistent with the observed curvilinearity. (We also examined the codes with five-year intervals and reached the same conclusions as with decades.) Thus, Columbus’s reputation suffered considerably in the texts published in the 1970s, but then moved to a more “balanced” portrayal in both positive and negative terms during the last decades of the twentieth century. The pattern for characterizations of American Indians is simpler, and since there are no books that are coded both positively and negatively, the percent- ages in table 5 add to 100. Positive characterizations increased from zero before 1960 to above 50 percent in the 1970s, and they have remained close to that level. Negative characterizations began at a middling level—no higher because Indians were rarely mentioned at all in the earlier decades—then disappeared completely in the 1980s and 1990s. The trend is generally con- sistent and is also significant for positive responses versus all others, linear b = .56, p = .02. Qualitative review of the quotations supports the quantitative results. Passages appearing in the 1940s and 1950s, for example, speak of Columbus in terms of vision and sacrifice, describe him as a hero, and refer to his “great service to civilization.” In those early decades, when Indians appear at all, they are described as “half naked savages” or as “childlike.” By the 1970s, different Indian tribes begin to be discussed in geocultural terms rather than as merely a welcoming party for Columbus. Increasingly, the texts state that the Indians were really the “first Americans.” The 1970s also saw more negative pictures of Columbus, as in one book’s characterization of European explorers Table 5. Results of Content Analysis of High School Textbooks: Evaluations of American Indians by Decade Year 1944–59 1960s 1970s 1980s 1990s Positive 0 25% 58% 40% 50% Negative 36% 42% 8% 0% 0 Unclear 64% 33% 33% 60% 50% (N of books) (11) (12) (12) (10) (10) 20 Schuman, Schwartz, and d’Arcy (including Columbus) as “shameful,” and another’s tongue-in-cheek account of an American Indian who steps off an airplane and “discovers” Columbus’s birthplace of Italy. In more recent years, Columbus himself continues to be seen as courageous and skillful and thus deserving of praise, but there is also recognition of the destruction that he and his successors (not always clearly distinguished) perpetrated on Indians and their cultures.16 Stimulated by Frisch’s (1989) observation that beliefs about the past may be laid down at early ages, we also examined a set of 18 “social studies” texts intended for elementary-level schools. Within the limits posed by their mis- cellaneous character and smaller sample size, they present much the same trends as the high school texts, and indeed there were two books from the mid- 1960s that already showed signs of change. Like the upper-level history text- books, they present an increasingly positive picture of Indians and a mixed, positive/negative picture of Columbus. We cannot connect survey respondents directly to the textbooks and related instruction by teachers that they received, but in the aggregate we expect the part of the American population that went through middle and high schools from the 1970s onward to have somewhat different views of both Indians and Columbus than older cohorts. In an approximate way this fits the findings in the survey data that indicate among younger cohorts some erosion of heroic coloring to Columbus and some increased belief in Indians as the “first Americans,” though the evidence for these changes was modest. M A S S M E D I A We looked for evidence in the mass media for both traditional and revisionist views of Columbus, with virtually all of it prompted by the 500th anniversary in 1992. The Readers’ Guide to Periodical Literature yielded 62 relevant articles in 1992, far more than in the previous ten years. Fourteen of the articles were positive with regard to Columbus’s reputation, nine were negative, nine were mixed positive/negative, and the rest simply used the Quincentenary as a peg for writing that had little or nothing to do with the controversy, for example, astron- omy (Scientific American), food (Gourmet), and travel (National Geographic). We also searched the nation’s leading newspaper, the New York Times, and found that it published 50 stories in the 1980s and 1990s that referred to the Quincentenary: 18 dealt with or at least mentioned revisionist or Indian criti- cisms, but the rest were entirely positive (for example, describing replicas of Columbus’s ships that were due to arrive in the New York harbor). The Washington Post had some 24 stories, half entirely positive, half with at least 16. Viking voyages to the western hemisphere are mentioned in 31 of the 55 books, with no rela- tion to publication date. This acknowledgment of pre-Columbian discoverers could have been widely known to those who had a high school education more than a half century ago. The rela- tively small number of mentions of “Other Europeans” in our national survey suggests that such information did not make a great impression on the general population. Elite Revisionists and Popular Beliefs 21 some mention of negative concerns, while somewhat surprisingly the Chicago Tribune (1985–1996) had 12 of its 14 stories at least mention criticisms of Columbus. Drawing on the Lexis-Nexis archive of newspapers from some twenty other American cities (for example, Houston Chronicle, Denver Post), we searched ten of these using the keywords “Quincentennial” and “Quincentenary.” Overall, 66 articles touched on Columbus; the largest number simply noted a local plan for a “Columbus Day Celebration” without further comment. Some local protests were reported, though they were seldom given major play. The Vanderbilt Television News Archive yielded eleven brief stories from the three national news networks in 1992. Four were positive accounts of the arrival of replicas of Columbus’s ships in New York harbor. The others concerned the controversy over Quincentenary celebrations, though most of these included replies to criticisms of Columbus. A series of seven one-hour public television documentaries, entitled Columbus and the Age of Discovery, was carefully balanced to present both exciting positive aspects of Columbus’s voyages and negative consequences for Indians. The programs were widely promoted and had 4 to 5 million households watching on average over the series, according to the WGBH research office. Two films about Columbus were released for the 500th anniversary: Christopher Columbus: The Discovery and 1492: Conquest of Paradise. Both were primarily action films that pre- sented Columbus in a generally favorable light; only the former showed cruel treatment of Indians, but it attributed such treatment mainly to Spaniards who were working to undermine Columbus.17 In sum, the mass media gave the public some exposure to revisionist ideas during the Quincentenary period, but the exposure was not great and, of course, would have reached only that frac- tion of the public that attends to such accounts. It is also useful to learn what government officials say about Columbus. All recent presidents from Franklin D. Roosevelt to George W. Bush have issued glorifying proclamations for Columbus Day (for example, Columbus was “brave, determined, open to new ideas and new experiences,” according to Bill Clinton). The 1992 Congressional Record has 144 entries for “quincente- nary” or “quincentennial”: 128 were entirely favorable, just 4 presented revision- ist ideas, 2 defended against revisionist ideas, and ten were positive about Indians but contained no criticism of Columbus. The U.S. Postal Service (2001) issued four stamps in 1992 in honor of the 500th anniversary: one shows Columbus seeking Queen Isabella’s support, two picture his first voy- age, and one depicts his 1492 landing, but with no sign of Indians present. It was left to the more elite Smithsonian Institution to take fuller account of revi- sionist ideas, but to do so along with positive views of 1492, thus presenting a “balanced” picture to those who were interested enough to visit the museums or to read a related special issue of Newsweek (1991). 17. Total attendance figures are hard to obtain, but a writer for the Hollywood Reporter, a trade paper, indicated that 3 to 4 million Americans saw each of the films; more would have seen the videos, but such figures were not released at all. 22 Schuman, Schwartz, and d’Arcy The Inertia of Memory Revisionist ideas that may have reached the public through textbooks and occasional treatment in the mass media have had to face the “inertia” of Columbus’s long-established reputation as the intrepid discoverer of America. One force that supports that tradition is the reappearance each year of Columbus Day, especially its institutionalized recognition by schools (see Schudson 1989b on cultural calendars). Informal conversations with primary school teachers indicate that teachers continue to use the day to speak of Columbus, his three ships, and the obstacles he overcame to reach American shores, even while they provide a much more positive view of Indians than would have been expressed in earlier years. The Columbus story also remains in elementary school books, including in a widely used series to teach first to third graders how to read, so that children absorb both reading lessons and stories about Columbus simultaneously (Krensky 1991). The text of the series is careful to speak of Columbus not as “discovering America” but as finding “a new world that no one in Europe knew” (p. 48); yet most of the words and the colorful pictures convey a traditional image of Columbus’s perilous first voyage and its successful outcome. In addition to the forces that shape historical consciousness in early childhood, there is later reinforcement as well: the many paintings, statues, and other commemorative symbols that maintain Columbus’s visibility, along with more than a thousand schools, natural sites, and other places named after him. “Once commemoration gets under way,” observes Schudson (1989a, p. 108), “it picks up steam, it operates by a logic . . . of its own. Not only are records kept, diaries saved, and news accounts written, but statues are built, museums endowed, brass plaques are engraved.” The importance of this symbolism is evident in the intensity with which critics attack it. “Beginning with Columbus,” declares an Indian activist, “we are insisting on the removal of statues, street names, public parks, and any other public object that seeks to celebrate or honor devastators of Indian peoples” (News from Indian Country, October 15, 1992, p. 2). Indian activists would feel no need to remove these symbols if they were not seen as legitimations of Columbus. Language underscores the inertia of reputation. There are the familiar rhymes that almost everyone learns (“In fourteen hundred and ninety-two . . . .”), but Columbus’s name itself is used as a symbol of individual boldness in exploring new ideas. A recent book refers to Albert Einstein as “an intellectual Columbus” who sailed beyond “the safe anchorage of established doctrines” (Cropper 2001, p. 203). A New York Times book reviewer calls Mikhail Gorbachev “a sort of political Columbus—setting out with high ideals to find one thing and achieving something better by discarding them. He is a hero of our times” (Figes 2002). The favorable connotation of Columbus’s name is also perpetuated by a myriad of magazine articles and Web sites that connect it to foods, travel, navigation, and much else unrelated to controversy. Elite Revisionists and Popular Beliefs 23 Of course, the inertia of Columbus’s positive reputation resists revisionist attacks that define his arrival as the beginning of a European invasion that brought slavery, disease, and death to indigenous peoples. But we should not assume that the public absorbs new ideas without emphatic and repeated use of school instruction and the mass media to spread new beliefs. Careful studies of the factual knowledge of Americans indicate that a good deal of the informa- tion that most social scientists take for granted is known by less than half— sometimes much less than half—of the general public (Delli Carpini and Keeter 1996). Although revisionist ideas were available to ordinary individuals who were interested, they were not so frequently or forcefully communicated as to offset the traditional image of Columbus as the courageous discoverer of America. “Inertia is a crucially important historical force,” Le Goff (1985, p. 170) observes; “Mentalities change slower than anything else.” Conclusions Once we recognize that collective memory can be different in its different forms, our results become more complex than those from studies that draw on only a single type of evidence. Had we focused our research entirely on revisionist writings and protests, we would conclude with Summerhill and Williams (2000, p. 1) that by the time of the Quincentenary, the “reputation of Christopher Columbus [was] turned upside down as fully as if the Admiral had indeed found monsters swimming in the Ocean Sea.” However, our surveys of the American public, using several different question forms and wordings, pro- duced little evidence of an impact from revisionist ideas: the predominant public belief is the traditional one that Columbus merits admiration as the “discoverer of America.” At the same time, we also found little evidence among Americans, especially younger cohorts, of the heroic image of Columbus that may have been widespread at the 400th anniversary in 1892 and in the early twentieth century. This absence of glorification can best be explained, however, not as a result of revisionist attacks and Indian protests, but as part of a broader erosion of idealizations of past American leaders. Furthermore, among the small number of Americans who reject the traditional belief that Columbus discovered America, it proved illuminating to distinguish between two different positions: the simple recognition that Indians had “discovered America” long before Columbus and additional characterizations of Columbus in villainous terms. The latter characterizations turned out to be linked to a critical attitude toward conventional American beliefs much more broadly. When we turned from individual beliefs to what is transmitted at the cul- tural level, we found the content of history textbooks to show a clear trend in positive treatment of Indians and a more complex negative/positive trajectory for Columbus. Both are consistent with some effect of revisionist beliefs on younger and better-educated Americans. It is interesting to note, however, that 24 Schuman, Schwartz, and d’Arcy changes in textbooks appear to have begun by the late 1960s, and thus prior to the main revisionist writings and even prior to what Nagel (1995, pp. 957–58) calls “the resurgence of American Indian ethnic identity in the 1970s and 1980s.” Quite likely the shift was a consequence of the still earlier post– World War II increase in sensitivity to minorities and their viewpoints, ini- tially with regard to Jews because of the discovery of the Nazi death camps (Fredrickson 2002; Stember 1966), which affected attitudes toward African- Americans as well (Hyman and Sheatsley 1956; Schuman et al. 1997). Similar sensitivity spread to other groups that had been relegated to the margins of the society, especially as the groups themselves protested their disadvantaged status. Postwar decolonization of large parts of the world previously under European rule also won support in the United States, with implications for reappraising the treatment of Native Americans by early European explorers and colonists. Thus, it seems likely that revisionist writings on Columbus were more an effect than a cause of the transformation of attitudes toward Indians and other minorities. Three statements now seem warranted. First, criticisms of Columbus usually reached the larger public in a much attenuated form, without the full negative force found in revisionist writings and Indian protests. Second, strong countervailing forces of inertia sustained Columbus’s reputation even in the face of revisionist attacks. There is a third, more positive factor as well, we believe, that supports Columbus’s reputation and explains why the 1492 landfall inspired commemoration in the first place and continues to do so: the significance of “firstness” or “priority.” Frisch (1989) reports that when stu- dents in his college history courses are asked to write ten names in “American history from its beginning through the end of the Civil War, excluding presi- dents and generals,” the list invariably includes Betsy Ross, the apocryphal creator of the American flag, and Columbus as well. These free associations tap memories that are absorbed at an early stage of life, Frisch believes, and they “stick” because of the lasting importance of “creation stories.” Columbus is still believed to have performed the key role in America’s founding moment. It is true that we may be living in a transitional era in which revisions of Columbus’s story have begun to appear in many textbooks and in other accounts of the American past, including books for small children. It is possi- ble that Columbus’s prestige will diminish further in the years to come. Yet some public commentators have attempted to counter revisionism at the elite level (Fox-Genovese and Lasch-Quinn 1999; Himmelfarb 1987; Schlesinger 1991), and reservations also appear in the online edition of the Encyclopedia Britannica (“Columbus” 2004b), which attempts to strike a “balanced” por- trayal of Columbus: The word “encounter” is now preferred to “discovery” when describing the con- tacts between Europe and the Americas, and more attention has been paid to the fate of indigenous Americans and to the perspectives of non-Christians. . . . The pendulum may, however, have swung too far. Columbus has been blamed for Elite Revisionists and Popular Beliefs 25 events far beyond his own reach or knowledge, and too little attention has been paid to the historical circumstances that conditioned him. . . . Columbus’s towering stature as a seaman and navigator, the sheer power of his religious convictions (self-delusory as they sometimes were), his personal magnetism, his courage, his endurance, his determination, and, above all, his achievements as an explorer, should continue to be recognized. References to revisionist criticism by the Encyclopedia Britannica suggest possible longer-term effects that have not been captured by our surveys, but the article includes considerable praise for Columbus as well. What will happen in the future as the controversy sparked by the Quincentenary dies down is hard to predict, especially if re-revisionism occurs, as the pendulum metaphor in this latest edition of the Encyclopedia Britannica suggests. Columbus’s was not the only contested reputation (Fine 2001) of the late twentieth century, but his case sets in relief the core of the theory of the poli- tics of memory: the tension between elite and popular beliefs about the past. The strength of politics of memory theory comes into play when we try to explain Columbus’s reputational decline during the twentieth century, for con- trol over conceptualizing the national past has been a major element of the minority rights movement. The theory’s weakness is apparent when we attempt to account for the limits of that decline: how Columbus’s reputation has withstood attacks and remained attached to the “discovery of America,” despite losing some of its earlier heroic color. We have considered several types of evidence in this article, with the two most important—revisionist accounts of 1492 and survey data on popular beliefs—leading to opposite conclusions. Readers who are influenced by the revisionist accounts may be skeptical of the survey results because they provide scant evidence of revisionist effects on the public. Such skepticism may be due to the mistake of extrapolating to the general public one’s personal experience with revisionist ideas and college campus protests. In addi- tion, those who find criticisms of Columbus personally persuasive may tend toward “looking-glass perceptions”: beliefs that seem reasonable to oneself are often assumed to be widely held by others (Fields and Schuman 1976–77; Ross, Greene, and House 1977). At the same time, readers com- mitted to the importance of national survey data may dismiss revisionist and Indian criticisms of Columbus as of little significance if they cannot be shown to have shaped popular beliefs. However, signs of revisionist influ- ence in textbooks, encyclopedias, and other intermediaries between elite levels and the larger public may well point to future change, especially through cohort replacement. In the end, the wider the range of evidence available about collective memory and about popular beliefs more gener- ally, the deeper our understanding will be about both the present and the future. 26 Schuman, Schwartz, and d’Arcy References Alderman, Derek H. 1996. “Creating a New Geography of Memory in the South: (Re)Naming of Streets in Honor of Martin Luther King, Jr.” Southeastern Geographer 36:51–69. Alwin, Duane F. 1997. “Aging, Social Change, and Conservatism: The Link between Historical and Biographical Time in the Study of Political Identities.” In Studying Aging and Social Change: Conceptual and Methodological Issues, ed. Melissa A. Hardy, pp. 164–90. Thousand Oaks, CA: Sage. American Association for Public Opinion Research (AAPOR). 2000. Standard Definitions: Final Disposition of Case Codes and Outcome Rates for Surveys. Lenexa, KS: AAPOR. Available online at http://www.aapor.org (accessed December 2004). Axtell, James. 1992. “Columbian Encounters: Beyond 1992.” William and Mary Quarterly 49:335–60. ———. 1995. “Columbian Encounters: 1992–1995.” William and Mary Quarterly 52: 649–96. Ben-Yehuda, Nachman. 1995. The Masada Myth: Collective Memory and Mythmaking in Israel. Madison: University of Wisconsin Press. Blight, David W. 2001. Race and Reunion: The Civil War in American Memory. Cambridge, MA: Harvard University Press. Bodnar, John. 1992. Remaking America: Public Memory, Commemoration, and Patriotism in the Twentieth Century. Princeton, NJ: Princeton University Press. Boorstin, Daniel J. 1962. The Image or What Happened to the American Dream. New York: Atheneum. “Columbus, Christopher.” 1993. Columbia Encylopedia. 5th ed. New York: Columbia University Press. ———. 2004a. Columbia Encyclopedia. 6th ed. New York: Columbia University Press. ———. 2004b. Encyclopedia Britannica from Encyclopedia Britannica Online, at http:// search.eb.com/eb/article?tocId=9109621 (accessed July 27, 2004). Converse, Philip E. 1964. “The Nature of Belief Systems in the Mass Public.” In Ideology and Discontent, ed. David E. Apter, pp. 206–61. New York: Free Press. Cropper, William H. 2001. Great Physicists: The Life and Times of Leading Physicists from Galileo to Hawking. New York: Oxford University Press. Crosby, Alfred W., Jr. 1973. The Columbian Exchange: Biological and Cultural Consequences of 1492. Westport, CT: Greenwood Press. Curtin, Richard, Stanley Presser, and Eleanor Singer. 2000. “The Effects of Response Rate Changes on the Index of Consumer Sentiment.” Public Opinion Quarterly 64:413–28. Danigelis, Nicholas L., and Stephen J. Cutler. 1991. “Cohort Trends in Attitudes about Law and Order: Who’s Leading the Conservative Wave?” Public Opinion Quarterly 55:24–49. Davis, James A., Tom W. Smith, and Peter Marsden. 2001. General Social Surveys: 1972–2000. Cumulative Codebook. Chicago: National Opinion Research Center. de la Garza, Rudolfo O., Angelo Falcon, and F. Chris Garcia. 1996. “Will the Real American Please Stand Up: Anglo and Mexican-American Support of Core American Values.” American Journal of Political Science 40:335–51. de Lancey, E. 1893. “Columbian Celebration of 1792: The First in the United States.” Magazine of American History 39:1–18. Delli Carpini, Michael X., and Scott Keeter. 1996. What Americans Know about Politics and Why It Matters. New Haven, CT: Yale University Press. Deloria, Vine. 1969. Custer Died for Your Sins: An Indian Manifesto. New York: Macmillan. DuMouchel, William H., and Greg J. Duncan. 1983. “Using Sample Survey Weights in Multiple Regression Analyses of Stratified Samples.” Journal of the American Statistical Association 78:535–43. Dubin, Steven C. 1999. Displays of Power: Memory and Amnesia in the American Museum. New York: New York University Press. Fields, James, and Howard Schuman. 1976–77. “Public Beliefs about the Beliefs of the Public.” Public Opinion Quarterly 40:427–48. Figes, Orlando. 2002. “Who Lost the Soviet Union?” New York Times Book Review, January 20, p. 11. Fine, Gary. 2001. Difficult Reputations: Collective Memories of the Evil, Inept, and Controver- sial. Chicago: University of Chicago Press. http://www.aapor.org http://search.eb.com/eb/article?tocId=9109621 http://search.eb.com/eb/article?tocId=9109621 Elite Revisionists and Popular Beliefs 27 FitzGerald, Frances. 1979. America Revised: History Schoolbooks in the Twentieth Century. Boston: Little, Brown. Foucault, Michel. 1977. Language, Counter-Memory, Practice: Selected Essays and Interviews, trans. D. F. Bouchard and Sherry Simon. Ithaca, NY: Cornell University Press. Fox-Genovese, Elizabeth, and Elisabeth Lasch-Quinn, eds. 1999. Reconstructing History: The Emergence of a New Historical Society. New York: Routledge. Fredrickson, George M. 2002. Racism: A Short History. Princeton, NJ: Princeton University Press. Frisch, Michael. 1989. “American History and the Structures of Collective Memory: A Modest Exercise in Empirical Iconography.” Journal of American History 75:1130–55. Gillis, John R., ed. 1994. Commemorations: The Politics of National Identity. Princeton, NJ: Princeton University Press. Glenn, Norval D. 1974. “Aging and Conservatism.” Annals of the American Academy of Political and Social Science 415:176–86. Groseclose, Barbara. 1992. “Monuments and Memorials.” In The Christopher Columbus Ency- clopedia, ed. Silvio A. Bedino, pp. 475–85. New York: Simon and Schuster. Halbwachs, Maurice. (1950) 1980. The Collective Memory, ed. Mary Douglas, trans. Francis J. Ditter and Vida Yazdi Ditter. New York: Harper and Row. ———. (1941) 1992. “The Legendary Topography of the Gospels.” In On Collective Memory, ed. Lewis Coser, pp. 193–235. Chicago: University of Chicago Press. Hart, Avery. 2001. Who Really Discovered America? Unraveling the Mystery and Solving the Puzzle. Charlotte, VT: Williamson Publishing Company. Himmelfarb, Gertrude. 1987. The New History and the Old: Critical Essays and Reappraisals. Cambridge, MA: Harvard University Press. Hyman, Herbert H., and Paul B. Sheatsley. 1956. “Attitudes toward Desegregation.” Scientific American 195:35–39. Irving, Washington. (1828) 1981. The Life and Voyages of Christopher Columbus, ed. John Harmon McElroy. Boston: Twayne Publishers. Jennings, Francis. 1975. The Invasion of America: Indians, Colonialism, and the Cant of Conquest. Chapel Hill: University of North Carolina Press. Johnson, Rossiter, ed. 1897. A History of the World’s Columbian Exposition. 4 vols. New York: D. Appleton and Company. Koch, Cynthia M. 1996. “Teaching Patriotism: Private Virtue for the Public Good in the Early Republic.” In Bonds of Affection: Americans Define Their Patriotism, ed. John Bodnar, pp. 19–52. Princeton, NJ: Princeton University Press. Krensky, Stephen. 1987. Who Really Discovered America? New York: Hastings House. ———. 1991. Christopher Columbus. New York: Random House. Las Casas, Bartolomé de. (1965) 1974. The Devastation of the Indies: A Brief Account. New York: Seabury Press. ———. 1992. Witness: Writings of Bartolomé, ed. George Sanderlin. Maryknoll, NY: Orbis Books. Le Goff, Jacques. 1985. “Mentalities: A History of Ambiguities.” In Constructing the Past: Essays in Historical Methodology, ed. Jacques Le Goff and Pierre Nora, pp. 166–80. Cambridge, UK: Cambridge University Press. Loewen, James W. 1992. The Truth about Columbus. New York: New Press. ———. 1995. Lies My Teacher Told Me: Everything Your American History Textbook Got Wrong. New York: Simon and Schuster. Lunenfeld, Marvin. 1992. “Adrift in a Sargasso Sea: Recent Books on Christopher Columbus.” History Teacher 26:15–21. Mannheim, Karl. (1928) 1952. “The Problem of Generations.” In Essays on the Sociology of Knowledge, by Karl Mannheim, pp. 276–322. London: Routledge and Kegan Paul. Morison, Samuel Eliot. 1942. Admiral of the Ocean Sea: A Life of Christopher Columbus. Boston: Little, Brown. Nagel, Joane. 1995. “American Indian Ethnic Renewal: Politics and the Resurgence of Identity.” American Sociological Review 60:947–65. National Council of Churches of Christ in the USA. 1990. A Faithful Response to the 500th Anni- versary of the Arrival of Christopher Columbus: Resolution, as adopted by the Governing Board, May 17, 1990. Available online at http://www.indians.org/welker/faithful.htm (accessed January 25, 2005). http://www.indians.org/welker/faithful.htm 28 Schuman, Schwartz, and d’Arcy Newsweek. 1991. “When Worlds Collide: A Joint Project with the Smithsonian’s Natural History Exhibit ‘Seeds of Change.’” Columbus Special Issue, Fall/Winter. Olick, Jeffrey K., and Joyce Robbins. 1998. “Social Memory Studies: From Collective Memory to the Historical Sociology of Mnemonic Practices.” Annual Review of Sociology 22:105–40. Pelta, Kathy. 1991. Discovering Christopher Columbus: How History Is Invented. Minneapolis, MN: Lerner Publications. Phillips, William D., Jr., and Carla Rahn Phillips. 1992. The Worlds of Christopher Columbus. New York: Cambridge University Press. Prescott, William F. 1874. History of the Reign of Ferdinand and Isabella the Catholic. 3 vols. Philadelphia: J. B. Lippincott. Rhea, Joseph T. 1997. Race Pride and the American Identity. Cambridge, MA: Harvard Univer- sity Press. Riley, Matilda White, Anne Foner, and Joan Waring. 1988. “Sociology of Age.” In Handbook of Sociology, ed. Neil J. Smelser, pp. 243–90. Newbury Park, CA: Sage. Rosenzweig, Roy, and David Thelen. 1998. The Presence of the Past: Popular Uses of History in American Life. New York: Columbia University Press. Ross, Lee D., David Greene, and Pamela House. 1977. “The False Consensus Effect: An Egocentric Bias in Social Perception and Attribution Processes.” Journal of Experimental Social Psychology 13:279–301. Royal, Robert. 1992. 1492 and All That: Political Manipulations of History. Washington, DC: Ethics and Public Policy Center. Sale, Kirkpatrick. 1990. The Conquest of Paradise: Christopher Columbus and the Columbian Legacy. New York: Alfred A. Knopf. Schlesinger, Arthur M., Jr. 1991. The Disuniting of America: Reflections on a Multicultural Society. New York: W. W. Norton. Schudson, Michael. 1989a. “The Present in the Past versus the Past in the Present.” Communica- tion 11:105–13. ———. 1989b. “How Culture Works.” Theory and Society 18:153–80. Schuman, Howard. 1972. “Two Sources of Antiwar Sentiment in America.” American Journal of Sociology 78:513–36. Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan. 1997. Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Harvard University Press. Schwartz, Barry. 1998. “Postmodernity and Historical Reputation: Abraham Lincoln in Late Twentieth-Century American Memory.” Social Forces 77:63–103. Schwartz, Barry, and Howard Schuman. Forthcoming. “History, Commemoration, and Belief: Reformulating the Concept of Collective Memory.” American Sociological Review. Skretney, John D. 2002. The Minority Rights Revolution. Cambridge, MA: Harvard University Press. Stember, Charles Herbert. 1966. “Recent History of Public Attitudes” In Jews in the Mind of America, ed. Charles Herbert Stember and others., pp. 31–234. New York: Basic Books. Stewart, Jon, Ben Karlin, and David Javerbaum. 2004. America (The Book): A Citizen’s Guide to Democracy Inaction. New York: Wamer Books. Stouffer, Samuel A. 1955. Communism, Conformity, and Civil Liberties. Garden City, NY: Doubleday. Summerhill, Stephen J., and John Alexander Williams. 2000. Sinking Columbus: Contested His- tory, Cultural Politics, and Mythmaking during the Quincentenary. Gainesville: University Press of Florida. Taylor, Charles. 1994. “The Politics of Recognition.” In Multiculturalism and “the Politics of Recognition,” ed. Amy Gutmann, pp. 25–74. Princeton, NJ: Princeton University Press. U.S. Postal Service. 2001. The Postal Service Guide to U.S. Stamps. New York: Harper Collins. Wagner-Pacifici, Robin, and Barry Schwartz. 1991. “The Vietnam Veterans Memorial: Commemorating a Difficult Past.” American Journal of Sociology 97:376–420. Wertsch, James V. 2002. Voices of Collective Remembering. New York: Cambridge University Press. West, Delno C., and August Kling. 1989. “Columbus and Columbia: A Brief Survey of the Early Creation of the Columbus Symbol in American History.” Studies in Popular Culture 12:45–60. Whitman, Walt. [1874] 1982. “Prayer of Columbus.” In Complete Poetry and Collected Prose, pp. 540–2. New York: Library of America. Wilford, John Noble. 1991. The Mysterious History of Columbus: An Exploration of the Man, the Myth, the Legacy. New York: Vintage Books. Elite Revisionists and Popular Beliefs 29 Winsor, Justin. 1891. Christopher Columbus and How He Received and Imparted the Spirit of Discovery. Boston: Houghton, Mifflin, and Company. Yerushalmi, Yosef Hayim. 1982. Zakhor: Jewish History and Jewish Memory. Seattle: University of Washington Press. Yolen, Jane. 1992. Encounter. Orlando, FL: Harcourt Brace. Zerubavel, Eviatar. 1992. Terra Cognita: The Mental Discovery of America. New Brunswick, NJ: Rutgers University Press. Zerubavel, Yael. 1995. Recovered Roots: Collective Memory and the Making of Israeli National Tradition. Chicago: University of Chicago Press. Zinn, Howard. 1980. A People’s History of the United States. New York: Harper and Row. Zolberg, Vera L. 1998. “Contested Remembrance: The Hiroshima Exhibit Controversy.” Theory and Society 27:565–90.