Truth and Goodness: Does the Sociology of Inequality Promote Social Betterment?

The historical development of sociology is outlined, from (1) close ties with social reform in the early 20th century, to (2) the divorce of sociology from social reform in the 1930s through the 1960s, when the “pure science” model dominated, to (3) the domination of the “unmasking inequality” model since the 1970s. I argue for a “reducing inequality” model. A content analysis of articles on inequality in three leading journals shows a shift in the 1970s to a structuralist perspective linked to unmasking inequality. But research that unmasks inequality still ignores social reform because career advancement depends on following the pure science model. Case studies of 12 sociologists illustrate strategies for combining research with social betterment.


INTRODUCTION
Most sociologists probably wonder from time to time: What is the point of my work? When I publish a research paper, does this benefit anyone besides me and my career and a few interested colleagues? When I teach a course, does it help my students have better lives?
Sociologists who study inequality have developed two major positions to respond to these questions, which I will label "pure science" and "unmasking inequality." I argue here that neither one is very useful in linking sociology to social betterment; instead, we should move towards a third position: "reducing inequality." I will begin by describing these positions and some of their problems.
The pure science position is based on the ideal of value-free science and the distinction between pure and applied research. According to this approach, pure research pursues objective scientific truth while applied research uses the results of pure research to address the social problems of different groups. Teachers cover both types of sociology. Thus, the combination of pure and applied work contributes to the general good, and sociologists who want to contribute directly to social betterment should do applied work.
But this justification of pure research is inadequate in dealing with three issues: audience, values, and scientific method. The audience problem is that pure research aimed at academic audiences usually is irrelevant to applied researchers, policymakers, and activists. Most sociologists engaged in social change-from liberal applied researchers like Peter Rossi to radical activists-seem to agree that the results of academic studies rarely are useful to them. Only the general concepts of sociology are useful. What they need is research that is aimed at applied and activist audiences, that addresses their concerns, uses their language, and includes policy recommendations (DeMartini 1982;Fritz and Clark 1989;Short 1975;Rossi 1980).
For example, imagine a group of sociologists who are doing pure research on status attainment and are investigating the impact of parents' education on their children's occupational status. These sociologists might use the pure science position and argue: "We are doing basic research on the causes of occupational success and failure, and our findings will help applied researchers to improve society." I would answer that their research is unlikely to contribute to social betterment because applied researchers will not find it useful. Their project would be more likely to have a social impact if it investigated questions that concern applied and activist researchers-for example, the effectiveness of different types of education in getting a good job, compared across different subgroups.
I am not arguing that all pure research should cease, since some research directed at academic audiences does turn out to be useful. I am urging sociologists to change the balance between pure and activist/applied research and focus much more of our work on issues relevant to non-academics.
The second weakness of the pure science position stems from the ideal of valuefree science. The possibility of value-free science has been rejected by feminists, postmodemists, and other scholars, who have shown that "pure" theory and research often is grounded in particular values and political interests (Haraway 1989;Harding 1986 andBernstein 1983). Nonetheless, the value-free ideal continues to be used to devalue activist and applied research, the type of sociology that is most likely to contribute to social betterment. Because the pure science position dominates most sociology departments and journals, sociologists who do academic, "value-free" research get the greatest respect and rewards. In contrast, activist researchers are seen as too ideological and political and are marginalized, while applied researchers are seen as second-rate.
This devaluation of activist and applied researchers is based on deception, since so-called pure research also is embedded in particular values and political ideologies. For example, the status attainment framework is based on a liberal/ conservative ideology that views individual talents and training as a major cause of inequality.
The final problem with the pure science position is that it overemphasizes a narrow set of quantitative methods. Promotors of pure science tend to devalue ethnography and other methods that often are especially useful to activists (Gotfried forthcoming;Stoecker andBonacich 1992, 1993). This methodological rigidity also undermines scientific standards by discouraging researchers from selecting the methods that best fit the question they are examining.
The second answer to how sociology contibutes to social betterment-"reducing inequality"-developed in the 1970s and 1980s. Acording to this position, the sociology of inequality is useful to society because it clarifies the social conditions that produce racism, classism, sexism, and other kinds of domination. But this position also raises problems about audience, values, and science.
Most sociological research that unmasks inequality is directed at academic audiences and, therefore, is not directly useful for social change. The research may have indirect effects by influencing students, the media, or policymakers. A few very influential academic studies have changed our concepts and had a major social impact-for example, the studies that developed the concepts of "world system theory" or "blaming the victim." However, most academic research is of little use to social reform.
The unmasking inequality position also encounters problems with values and methods that are linked to "political correctness." Some left-liberal sociologists who are committed to unmasking inequality go too far in rejecting the traditional norms of science. They write essays that champion their values but fail to present supporting evidence or consider alternative perspectives. They ignore research that challenges their politics and marginalize colleagues who disagree with their values.
While some sociologists who are committed to unmasking inequality are methodologically sloppy, others are too rigid. In paticular, feminist sociologists tend to restrict themseslves to ethnographies and other qualitative methods and to devalue quantitative methodologies.
Because of these inadequacies in the pure science and unmasking inequality positions on how to link sociology and social betterment, I propose a third position-"reducing inequality." This position advocates doing more research and teaching that is directly related to resolving social problems and that is aimed at audiences involved in social change. Greater emphasis on applied and activist sociology is combined with a self-reflexive, tolerant stance on values and with rigorous but diverse scientific methods.
An example of research that reduces inequality is Lenore Weitzman's study, The Divorce Revolution (1985). Her work unmasks the inequity in divorce laws, particularly, how they put women at an economic disadvantage compared to men. But her book also suggests how laws should be changed to rectify the imbalance and it is written in accessible language. As a result, her study affected state legislators and promoted significant changes in divorce laws.
Sociology that reduces inequality, I will argue, should investigate social problems and remedies and, at the same time, be reflexive about values and committed to high scientific standards. Then we will have a better alternative for linking sociology and social betterment than doing pure science or unmasking inequality.
The following analysis is divided into two major parts. First, I examine the historical development of the pure science ideal in the 1930s and 1940s and the dramatic shift to unmasking inequality in the 1970s. Then, I present cases studies of individual sociologists and how they relate their work to benefiting society.

ASSUMPTIONS ON SCIENCE AND VALUES
Before turning to the history of sociology, let me clarify my assumptions about values and science. Like many contemporary sociologists, I assume that values and political interests are built into our teaching and research, and even our vocabulary (Gans 1991). Thus, conservatives talk about incentives and social disorganization, while radicals talk about exploitation and resistance. As Howard Becker argued long ago, sociologists cannot avoid taking sides (1967).
I also assume that values threaten scientific standards. Sociologists with passionate political beliefs tend to ignore evidence that challenges their position. Those with hidden values that are masked by detached, abstract language can be even more misleading, because their values are harder to see. There is a conflict between scientific standards and values, but we are stuck with it.
The "pure science" position encourages us to deny these problems-to pretend that we can exclude values by using scientific methods. A better approach is to be open about the values of different groups of sociologists and to grapple with conflicts between these groups on a collective level. On an individual level, we should be aware of the impact of our values and social position on our sociology, and tolerant about other people's values (on strong objectivity, see Harding 1991).
My assumptions about science build on Mary Hawkesworth's definition of the two basic norms of science (1989). First, according to Hawkesworth, scientific beliefs must be supported by good evidence. In my view, this norm means that we should use research methods that are appropriate for answering the question at hand and reject methodological rituals like the indiscriminate use of regression models. We should also criticize sloppy research, even if it supports our beliefs.
The second norm of science is that beliefs must be open to debate and criticism and that alternative interpretations must be considered. I assume that to ensure open debate, all major groups with distinctive values or interests must have an academic base that permits them to participate, providing that they conform to scientific norms. Once we admit that people's values and social positions affect their social science, then open debate becomes a sham if, for example, the standpoint of the poor or people of color is marginalized. As long as ruling groups have much more power than others to develop and communicate their sociological interpretations, scientific norms require sociologists to favor the perspectives of the relatively powerless and to promote greater diversity within the research establishment (Cancian 1992).
Building on these assumptions about values and science, I will examine the history of U.S. sociology and trace the development of the pure science and unmasking inequality positions.

EARLY AMERICAN SOCIOLOGY: THE LATE NINETEENTH CENTURY TO THE 1920s
In the late 19th century, most sociologists investigated society in order to promote social welfare and combat evil. Sociology included many social workers and reformminded Protestant clergymen and was strongly influenced by the Progressive Movement (Kuklick 1980;Mitchell 1968). Over the next hundred years, sociology became increasingly divorced from social betterment and from nonacademic audiences (Friedrichs 1970;McCarthy 1985;Obershall 1972).
In the first issue of The American Journal of Sociology (AIS), published in 1895, editor Albion Small defined the four goals of sociology as: (1)building theory to identify the principles of social relations, (2) relating abstract sociology to everyday life and the interests of leading citizens, (3) promoting the general welfare, and (4) restraining premature public opinion (Small 1895). "Knowledge needed for social purposes," he stated, "shall be combined and applied as a means of promoting welfare" (1895:7). Reflecting these goals, volume 1 of the AIS included articles on reforming city government in Chicago; several articles on Christian sociology, including one about Jesus Christ's social theory, based on the Bible; and an article by Jane Addams describing the working conditions of women domestic workers.
As I read these early sociologists, I was impressed by their ability to work with social reformers and citizens' groups to improve society. Their work mattered, in contrast to contemporary research which too often is trivial, careerist, and interesting only to other sociologists. But I was also appalled by the moral superiority of many sociologists, their smug assurance that social betterment was defined by their progressive and Christian beliefs. And their research methods were sometimes embarrassingly dilletantish. Much of their work confirmed the stereotype that social reformist sociology entails closed-minded moralizing and bad science. (This paper uses the term "social reform" to include both liberal reforms and radical change, although it would be interesting to examine them separately).
The emphasis on social reform and Christianity quickly diminished. By 1900, the AJS contained almost no Christian sociology (Shanas 1945), and many social workers and reformers left sociology during the 1905 split between the American Sociological Society and the reform-oriented Conference of Charities and Corrections (Mitchell 1968).
The 1920s and 1930s were dominated by sociologists at the University of Chicago, who continued the social reform tradition by studying social problems in Chicago-immigration, assimilation, gangs, and family disorganization. But there was a new emphasis on doing systematic, objective research and a conservative drift in their research agenda, which neglected issues of social class and stratification (Oberschall 1972:240-241;Kinloch 1988).
Then, in the 1930s, the pure science position became dominant and destroyed most of the connections between sociology and social reform. Because of several social changes, sociologists were focusing more on academia and less on outside social reform. Sociology departments were being established in a growing number of universities, and sociology's scientific respectability was becoming more important. In the wider society, the Progressive Movement had weakened and, in contrast to many European countries, there was no strong socialist party to promote progressive research. Thus, university colleagues became the sole reference group for many sociologists.
In this social context, there was a "sharp reaction against preaching, muckracking, and value judgements," and some leading academic sociologists tried to purge reformers and dilettantes from the profession, using quantitative methods as a weapon (Oberschall 1972:205). Thus, in his 1926 presidential address to the American Sociological Society (ASS, forerunner of the ASA), John Gil attacked the "enemies of sociology" whose work, he said, was popular but unscholarly. "Doubtless we shall have to put up with some of them longer, inasmuch as there is no sociological orthodoxy and no sociological inquisition or holy office by which these fellows can be eliminated. Emphasis upon rigidly scientific methods will attend to them" (Oberschall 1972:242). As Anthony Obershall concludes in his analysis of the 1920s and 1930s, "In their desire for collective professional mobility and recognition, sociologists hid behind the facade of value-free social science that spared them the risk of radical social criticism and unpopular suggestions for the remedy of social problems" (1972:243). Sociology flourished in the 1930s, 1940s, and 1950s, and Oberschall and other historians of sociology argue that it prospered partly because it abandoned the political interests and radical ideas of earlier generations (Mitchell 1968:216; see also Oberschall 1972;Gouldner 1970;Pease, Form, and Huber 1970).
The pure science ideal dominated sociology in the United States until at least the early 1970s, according to Oberschall (1972:244). It stressed "value-free sociology, the separation of the roles of scientist from that of the citizen and reformer," and "limited, specific, and quantitatively oriented sociological research" (Oberschall 1972:242). In support of this ideology, William Ogbum's 1929 presidential address to the ASS asserted that "sociology as a science is not interested in making the world a better place to live." Instead, sociologists must "crush out emotion ... [and] taboo ethics and values (except in choosing problems)" and build a science "by the accumulation of bits and pieces of new knowledge" verified by quantitative techniques (Oberschall 1972:243).
The pure science ideal destroyed most of the connections between sociology and social reform by purging reformists, condemning reformist values, and narrowing the audience of research to fellow social scientists. Ironically, it also undermined scientific standards by restricting debate and promoting rigid definitions of appropriate methods.
In the 1940s, 1950s, and early 1960s, the pure science ideal flourished as academic sociology grew enormously in size, resources, and influence. Harvard became a leading center, where Talcott Parsons and his followers championed grand, functionalist theory as a critical part of scientific sociology. Now, good research had to be theoretically relevant in addition to methodologically sophisticated.
Parsons and others extolled the value of "pure" as opposed to "applied" research. When I was a graduate student at Harvard in the late 1950s, I remember Parsons telling me with pride that sociology was now in a much better position because it was separated from social work, and I remember agreeing with him.
In his 1949 presidential address to the American Sociological Society, Parsons argued that if sociologists "put the overwhelming bulk of our resources ... into immediately practical problems, it would do some good, but I have no doubt that it would have to be at the expense of our greater usefulness to society in the future" (Larson 1986:116). The best way to promote social betterment, according to Parsons, was to advance basic science. Thus, Parsons presented functionalist sociology as basic science that would promote the general good-a spectacular case of how I.pure science" can mask values and political interests, since functionalism typically presents the perspective of the elite, not of society in general.
At Columbia, another leading center, survey research and quantitative analysis were emphasized in the 1950s and 1960s. Columbia also had an important center for applied research, which primarily served the government and corporations. But applied researchers were generally regarded as inferior, as they are to this day, and were excluded from top universities and journals (Gollin 1989;Rossi 1980).
Both applied and pure researchers were expected to exclude their values from their work as much as possible. Applied researchers were defined as social engineers or technocrats, who used the "pure science" developed by others or themselves to serve the values of their clients (Bernstein 1983;deMartini 1982). However, it was assumed that sociologists would conform to the implicit liberal/ humanitarian values of sociologists and would not, for example, do research for the Ku Klux Klan (Furfey 1954;Becker 1967).
Of course, in the 1950s, as in all decades, sociologists were diverse and the pure science position was widely challenged. Undergraduate classes and textbooks emphasized social reform and current social issues, as they had since the beginning of sociology (McCarthy 1985;Mitchell 1969:216). Students then, as now, were recruited to sociology by its social reform agenda, creating a chronic tension between the personal values of sociologists and dominant professional norms (Rossi 1980;Mitchell 1968;Gouldner 1970). Social problems-especially criminology, deviance, and race relations-continued to attract many researchers.' The Society for the Study of Social Problems and its joumal, Social Problems, were founded in the early 1950s to promote the study of social problems and provide "the knowledge for sound social action" (Burgess 1953:2; see also Skura 1976). Finally, a few influential sociologists such as C. Wright Mills and Alvin Gouldner criticized the dominant trends and argued that sociology should expose injustice and work for progressive social change (Mills 1959;Gouldner 1962Gouldner , 1970. But the pure science ideal dominated the high-status journals and universities.

THE 1970s AND 1980s: UNMASKING INEQUALfTY REPLACES PURE SCIENCE
A major challenge to the pure science, functionalist position began in the late 1960s with the antiwar movement, the revival of the left on university campuses, and the subsequent rise of feminism and multiculturalism. By the 1980s, this challenge had succeeded in overthrowing functionalism and changing the dominant political values in the field of inequality. The ideal of unmasking inequality replaced the belief in value-free sociology. But the sociology of inequality continued to have very weak ties to social reform, in part because many aspects of the pure science ideal remained in place.
Before the 1970s, research on stratification was dominated by conservative political values and neglected economic and political inequality, as John Pease, William Form, and Joan Huber Rytina demonstrate in their excellent analysis of the stratification literature (1970; see also Kinloch 1988). They show that, except for the Great Depression, the dominant perspective in stratification studies was "evolutionary liberalism," which "holds that amorphous classes emerge as a consequence of individual mobility" and explains inequality as resulting from "natural selection of those who are socially and biologically most 'fit"'-similar to the functionalist model. Sociologists ignored the alternative perspective, "structural realism," which "holds that distinct classes emerge as a consequence of socially created arrangements that maintain economic, political, and social inequality" (Pease, Form, and Rytina 1970:128). In the 1940s, they report, stratification studies emphasized social prestige and status and compared the values and style of life of different strata. In the 1950s and 1960s, some sociologists began to criticize the conservative bias of functionalism, but research continued to focus on upward mobility and occupational prestige, and the United States continued to be seen as a classless, or predominantly middle-class, society.
By the late 1970s, the dominant theoretical framework in stratification shifted from functionalism to structuralism. Gradually, the label "stratification" yielded to the label "inequality," reflecting the new focus on economic and political inequality, world system theory, and the interlocking inequalities of race/class/gender.
How much did research on inequality change after 1970? To assess the changes, I did a content analysis of a random sample of 48 articles from the American Journal of Sociology, the American Sociological Review, and Social Problems from the years 1968, 1978, 1988, and 1992. I selected these journals because they are top-ranked; in addition, Social Problems has a distinctive applied orientation. In selecting years, I began with 1968, the most recent year that was still clearly dominated by functionalism; then I selected every tenth year following 1968, plus the most recent year at the time I began the analysis. I analyzed 12 articles each year, four from each journal. Twelve binary categories were used to analyze each article. Most of the categories were designed to measure the use of a liberal/functionalist framework versus a structural framework, as defined by Pease, Form, and Rytina. For example, did the article imply that inequality was caused by individual differences in talents and values or by structural arrangements? Was the existing economic and political order portrayed as good or bad? I also noted whether each article used quantitative and/or qualitative methods, mentioned social policy implications, and/or used the writing style of a detached, value-free scientist. Finally, I gave each article an overall rating as "functional" or "structural," summarizing the 12 categories (see Appendix for the category system and results of overall ratings).
The content analysis shows a clear change from a functional to a structural perspective in the AJS and the ASR In 1968, only one out of eight articles was structural; by 1992, five out of eight were structural. Social Problems was mostly structural in 1968 and remained that way.
To illustrate the change, my sample of articles from 1968 included an AJS article, "Class Identification in the US," that used survey data to show that class identification was based on friendship networks more than relations to the means of production. The authors concluded that the concept of class was not very useful, given the ambiguity of status differences in the United States (Hodge and Treiman 1968). In contrast, my 1992 sample included an article on "The Empowerment and Health Status of African-Americans" by Thomas LaVeist, a quantitative analysis showing that cities in which blacks controlled more city council seats (relative to the black population) had lower rates of black infant mortality (LaViest 1992). Whereas the earlier paper minimizes economic and political inequality, the 1992 paper links economic inequality to the death of children and shows that redistributing political power is an effective remedy.
With this dramatic shift from a functional to a structural framework, sociologists of inequality increasingly identified with the interests of the relatively powerless and poor, in contrast to many functionalists' identification with the elite or with "society as a whole." The ideal of value-free, pure science was replaced by the new ideal of taking sides with the relatively powerless and unmasking hidden structures of inequality.
However, despite all these changes, three aspects of the pure science ideal persisted, and they still prevent our work from reducing inequality. First, with the exception of Social Problems, articles in the high-prestige journals are still directed at a narrow, sociological audience and rarely even mention the implications of the research for policy or for the interests of different social groups (Henslin and Roesti 1976). Thus, Thomas LaVeist's article on the political empowerment of African-Americans and infant mortality does not present any recommendations for health policy or political mobilization, nor does he connect his research to broader issues of racism, political power, and public health. If he had done so, his article probably would not have been accepted by the AJS but it might have had a greater social impact.
Second, journal articles continue to be written in dispassionate, detached language that masks the authors' values and political agendas, even though their values are usually pretty obvious (Bazerman 1988). This writing style maintains the pretense of value-free sociology even though most sociologists probably do not believe in it anymore. It also makes our publications less interesting to social reformers and the general public. Moreover, this professionally imposed censorship of researchers' values and politics violates a major standard of good science and scholarship-that is, important assumptions should be clearly stated and analyzed (Harding 1991).
The third aspect of pure science that has persisted is the ritualized use of certain quantitative methods (Wilner 1985). Multivariate models seem to be almost a requirement for publication in the top journals, regardless of the question being asked (with the exception of a few comparative-historical papers). While these methods are an excellent choice in some research situations, they sometimes are inappropriate. For example, in 1992, 11 of the 12 randomly selected articles were quantitative studies; the single exception was a theoretical essay in Social Problems (Geschwender 1992). All four articles from the AJS in 1992 used multivariate regression or probit models to analyze data from large national surveys. This methodology was inappropriate for at least one study, which investigated the impact of gender role ideology on husbands' and wives' decisions to relocate when one spouse was offered a better job (Bielby and Bielby 1992). To obtain valid measures of gender role ideology, the researchers should have done open-ended interviews or used attitude scales. Instead, they used two questions on ideology that happened to have been included in a large survey on unemployment. This methodological choice produced a dubious measure of the main independent variable but it enabled the researchers to do a multivariate analysis of survey data and, thus, improve their chances of being published in the AJS.
In contrast to the methodological rigidity of these journals, there is more diversity in books and in some less prestigious journals (Wilner 1985). They often are aimed at a broader audience and include open political positions, policy recommendations, and a wide array of methods. But the top journals continue to define excellence in research and to strongly influence decisions about hiring and promotions.
In sum, much of our practice of sociology continues to follow the pure science ideal-the belief that sociology contributes to social betterment by producing objective knowledge which is then applied by others to address social problems. But most of us no longer believe in this ideal, so we feel cynical or disillusioned about our work. Many of us justify our sociology as challenging inequality but, under the influence of the pure science ideal, we do not communicate our research to nonacademic audiences nor do we suggest concrete actions to improve the situation. Thus, our research unmasks inequality but is unlikely to reduce it. The sociologists I interviewed illustrate different ways of negotiating these problems and provide examples of sociology that does effect social change.
CONTEMPORARY SOCIOLOGISTS: SOME CASES I interviewed 18 sociologists about how their teaching and research on social inequality related to social change. The interviews show how individuals combine elements of the pure science, unmasking inequality, and reducing inequality positions in justifying their work. Several cases also show the overriding importance of the pure science ideal in career advancement, because it has become institutionalized as the basis for hiring, promotion, and publication in academic journals.
The respondents were selected nonrandomly but were diverse in political orientation, ethnicity, and location in research or teaching institutions. Most interviews were face-to-face and lasted about an hour; a few shorter interviews were conducted by phone (see Cancian 1993 for a discussion of some of these interviews). Three types of sociologists emerged from my anaysis: researchers whose work unmasks inequality but is not directly related to social change; several activist sociologists who are heavily involved in social change; and one applied researcher who also is directly involved in social change.
The first sociologist who sees her work as unmasking inequality is Laura Stevenson, recently retired from a research university (I have changed names and a few details to provide anonymity). She is a feminist committed to unmasking male domination, but she also supports elements of the pure science ideal, including the separation of sociology and social reform.
Laura's father taught sociology at a small college and had previously been a Protestant preacher. Sociology, to him, meant promoting "moral uplift" and combating racism and other social problems. When Laura was asked, as a child, to define sociology, she reportedly answered: "I think it's about how to be good." Entering Harvard as a graduate student in the 1940s, she recoiled from her father's moralizing, Christian perspective and welcomed Harvard's "lack of interest in social problems" and "dispassionate" theoretical analysis.
Through her writings in feminist theory, Laura hopes to encourage social change through having her work be widely read and using it to help people recognize male dominance. But her writings do not suggest models of change or support specific reforms, and she seems uncomfortable with the ideological fervor of politically engaged sociologists. Sociology, in her view, is "very removed from social change" and should focus on finding out truth. Truth in itself promotes social betterment. Her position is a contradictory mixture of pure science and feminism. She rejects moralizing, reformist sociology, yet her own research has the reformist goal of combatting sexism.
Another sociologist who tries to unmask inequality is John Barker, recently tenured at a research university. He is more strongly committed to reducing inequality than Laura, and less attracted to the pure science ideal. He would like to connect his work on international political economy to social change, but he is not able to do so in an academic setting where rewards are based on the pure science ideal.
When I asked him whether his work related to social change, he confidently responded "almost everything I do is social change oriented" and then described a series of academic studies he had done that analyzed economic and political inequality. But his doubts and confficts surfaced when I pressed him for details on how his research effected social change. He explained that only his earliest work was policy oriented. "You end up getting pulled into more theoretical and methodological stuff," he said. Especially "when you're trying to get tenure there's a lot of pressure to publish" and do "compartmentalized" research that "doesn't have much to do with changing things." On the other hand, he believes that sociologists have an "obligation to try to figure out how to apply our research." "I'm interested in seeing a better world," he added, but "it's hard to figure out how to use knowledge, (or) who to transmit it to." If you do research for policymakers, he commented, they may well use it to serve the elite. He finds that his teaching is much easier to link to social change, but teaching receives few rewards in his university.
Like many left-leaning sociologists, John would like to link his research to social action but is unclear about how to do so. He continues to produce the type of mainstream research that is published in top journals. I would guess that he avoids activist research both because it is difficult for sociologists to connect with nonacademics and because doing activist research would threaten his career advancement. John Barker, like Laura Stevenson, is committed to unmasking inequality, but his research has little or no direct impact on social change, although it may have an indirect impact by affecting teaching or even the national discourse on public policy. Now, I would like to describe some activist sociologists who have a strong direct involvement in social change. They do research focused on the concerns of unions or community organizations. They are in close contact with nonacademic groups and address some of their publications to community groups, social reformers and policymakers. Many of the activists use some elements of participatory research; that is, community groups often help determine the research agenda and sometimes participate in carrying out and disseminating the research (Stoecker andBonacich 1992, 1993). These nonacademic contacts andpublications require considerable time and energy, and many activist sociologists have two careers at once-one directed to their social change projects and one focusing on academic contacts and publications.
Besides the burden of having two careers, activist sociologists face substantial threats to their career advancement because their work is judged by the pure science ideal and seen as ideological (as opposed to value-free). They are charged with doing research that lacks "theoretical significance" and methodological rigor, since their work addresses specific social problems and often uses qualitative methods.
Mark Chesler, who has published an account of his experiences, suffered major career setbacks at the University of Michigan because his work was seen as too political and ideological. In the 1970s, Chesler helped create the Educational Change Team, a research and social action organization focused on racism in high schools which was part of the Institute for Social Research at Michigan. "The team received substantial federal and foundation funding for research and action projects ... [but] leaders of the Institute for Social Research argued that these projects contained an inappropriate balance of social research and social action, and 'invited' the team to locate its action work outside of the institute" (Hasenfeld and Chesler 1989:503). Several years later, "colleagues in the Sociology Department decided not to promote me to full profesor status. The principal reasons involved ... the appropriate balance of scholarship and action in my work, and the general relevance of my work for advancing the intellectual frontiers of sociology as a discipline " (1989:504). Eventually, Chesler published some more mainstream research; he was promoted "and the disrespect and stigma now seem much diminished " (1989:505).
Another sociologist, Julia Vasquez, described the barriers to publishing activist research. Julia is a newly tenured sociologist at a research university who studies racism in medical schools and other topics that are important to the Mexican-American community. "All of my research has been a response to being part of that community" and "having an obligation" to it, she said. But for many years, she had trouble getting her social-change-oriented research published in respected journals. The "theoretically significant" questions that journals prefer "are so narrowly defined that your work can't have any impact" on social change, she commented. The experiences of Mark Chessler and Julia Vasquez illustrate how the pure science ideal persists in research universities and academic journals, deterring us from doing research that promotes social change.
My final example of an activist, Barbara Bloch, is a tenured professor at an urban state university. She does not care very much about publishing in top journals and, therefore, is much freer to do sociology that will benefit her students and the feminist and left-wing activists whom she studies and supports. Barbara was originally drawn to sociology by C. Wright Mills' "critique of inequality" and his "vision of a possible alternative." In the late 1960s, her fellow graduate students at the University of Wisconsin gave her a thorough grounding in Marxist sociology, and in the early 1970s, she became a feminist and found ways of integrating sociology with being of use to students and to the political movements that she supports.
She described her classes in stratification with great enthusiasm. "You have to teach it so it is empowering," she said, and "give them some sense of the possibility of change ... not just critique." She tries to increase feelings of self-worth among her students, most of whom are the first generation in their family to go to college. "They deserve to know they are struggling against tremendous odds," given the inequality of the United States. When her students write papers analyzing their own class backgrounds, many of them "go through incredible transformations ... it's a major turning point in their lives." Besides teaching, Barbara does research on feminist movements that is useful to activist organizations in planning and evaluating their strategies. She also is involved as an activist herself, and encourages her students to get involved. Activist teaching and research, like Barbara's, is a model of sociology that reduces inequality, as long as it avoids the two pitfalls of reformist sociology: sloppy research methods and closed-minded moralizing.
The final respondent that I want to discuss, Sam Whitman, is a retired applied sociologist. I have chosen to end with his story because his experience reminds us that there is intolerance on the left as well as the right and that this intolerance discourages excellent applied research. I hope that his story will lead us to be more self-reflexive and open-minded as we pursue a sociology that reduces inequality. Sam Whitman is a very successful sociologist who combined applied research on race relations and other issues with major contributions to academic sociology. Entering graduate school in the late 1930s, he embraced Parsonian functionalism and then, in the 1940s and 1950s, researched black-white relations in several U.S. cities. In addition to his highly rewarded academic research, he did extensive consulting and political work to promote racial integration, including working with the NAACP, local clergy, and community organizations; giving public speeches; and consulting with government commissions. "I never accepted the idea of pure versus applied research," he told me at the beginning of our interview. You "test ideas by acting on them." He feels that his community work and consulting enriched his academic work and "kept me in touch with things I ought to know" in teaching and research. Much of his work is a model of applied research to reduce inequality.
As the field of inequality shifted from a functionalist to a structural perspective in the 1970s, Sam was criticized and marginalized by young colleagues, some of whom saw him as a patronizing or even racist defender of the status quo. He eventually shifted his research focus away from U.S. race relations.
He, in turn, criticizes the closed-mindedness of the sociologists who attacked functionalism. C. Wright Mills and the young radicals were "not into advancing sociology," he said. Instead, they promoted a new kind of activism that "started with values" and critiqued the existing society. They have "trashed positivism" and replaced scientific research with political ideology. As a result, "some fields of research disappeared," such as military sociology, because they "became a sign of being a fascist." And other fields became taboo for many researchers. For example, it is no longer acceptable for white sociologists to study blacks.
Sam Whitman justifiably resents being marginalized by radical sociologists, but he seems unaware that radicals and activists are similarly marginalized by academic sociology. Like many liberal or conservative sociologists, he perceives the explicit political agenda of many leftists as dangerous ideology, but to him the hidden agenda of the right is invisible, or merely innocuous.

CONCLUSION
Sociologists of inequality have succeeded in replacing functionalism with a new commitment to identifying with poor and powerless groups and unmasking inequality. But they have failed to connect their work to social change. Ideological conflict is one barrier to linking sociology with social change and promoting applied and activist work. Mainstream liberal sociologists discourage activists by denying them access to top journals and blocking their career development. Left-leaning sociologists, though less powerful, discourage applied researchers like Sam Whitman by condemning and marginalizing them. This ideological conflict partly stems from the pure science ideal. Since the 1920s, many promotors of pure science have been intolerant and arrogant in relation to their reformist colleagues and have attempted to drive radicals and reformers out of sociology. These attacks from the right have elicited intolerant and arrogant counterattacks from the left.
A much greater barrier to activist and applied research is the persistent definition of high status sociology as separate from social change and nonacademic audiences. Here, again, the pure science ideal as it is institutionalized in academia is largely responsible. The explicit goal of promotors of pure science was to separate sociology from social reform, and they have succeeded.
I suggest that we abandon the pure science ideal. Instead, we should commit ourselves to reducing inequality. We should reunify sociology and social reform and, at the same time, strengthen scientific norms of open debate and good evidence and practice reflexivity and tolerance about our values and politics. I am not suggesting that sociologists abandon our autonomy or our more theoretical or arcane interests. I am suggesting a change in emphasis, so that research and teaching on inequality, as a whole field, would be more concerned with applied and activist projects.
To reconnect sociology with social reform, we need to address nonacademic audiences and their concerns. This means doing more applied research for government agencies and corporations and doing more activist research for community organizations. It also means changing our graduate curricula to train students to do such research and giving more rewards to undergraduate teaching that emphasizes social change. Standards of publication also need to be changed so that articles on social practice appear in leading journals, and standards for hiring and promotions need to give more recognition to activist and applied work. Other social sciences such as economics and psychology provide more training and recognition for social practice than sociology, even though their top journals are dominated by theoretical, academic work. Sociology as well can give more legitimacy to applied and activist research without threatening its academic standing.
Sociologists must take some position on how our research and teaching contributes to social betterment. We have to justify our work to ourselves and to the people that pay us and support our departments. In the sociology of inequality, reducing inequality is a better justification than pure science or unmasking inequality. By promoting activist and applied sociology, we can both strengthen the sociology of inequality and move towards reconnecting truth and goodness. or was balanced between the two positions, no code was assigned. Four articles were randomly selected from each volume by picking the first two articles on inequality (e.g., class, poverty, mobility, race, or gender) in issue number 1, and the last two in issue number 3, then moving to other issues if necessary. If several articles on inequality were grouped under a subheading, I selected only one.
Categories (iberallfunctionalist position listed first)