http://whorulesamerica.net/teaching/teaching_who_rules_america.html (retrieved May 22, 2024)

Teaching Who Rules America?

by Shelly Tenenbaum & Robert J. S. Ross

Clark University

From Teaching Sociology, Vol. 34, October 2006 (pp. 389-397).

At a small private liberal arts school such as Clark University, students are familiar with people from upper middle class backgrounds. Either they themselves are from upper middle class families or, if they are from working class or lower middle class backgrounds, they are acquainted with more affluent peers. Clark students have a clear image of how wealthy professionals, managers, and entrepreneurs live; they know what their homes look like, where they go for vacations, and are familiar with their patterns of material consumption. In contrast, the upper class -- the top 1 percent of the population who owns 51.4 percent of the privately held corporate stock -- is so far removed from the vast majority of our students' experiences that our required social stratification course, "Class, Status, and Power," is usually their first introduction to this social group (Kerbo 2003). In this article, we explore two assignments that help our students understand the formidable political and economic power of the "invisible" elite and the corporate structure that sustains it.

Fortunately professors have access to an array of sociological studies on the upper and corporate classes to use in their courses (Sernau 2000), including G. William Domhoff's Who Rules America? (2002), C. Wright Mills's The Power Elite (1956), Susan Ostrander's Women of the Upper Class (1984), Michael Useem's The Inner Circle (1984), and Richard Zweigenhaft and G. William Domhoff's Diversity in the Power Elite (1998). While reading texts provides students with one level of knowledge, applying concepts to concrete situations requires an even higher degree of understanding (Angelo 1993; Gamson and Chickering 1987; Gronlund and Linn 1990). As a result, instructors of social stratification have developed classroom projects to reinforce their course materials and to help students move beyond individualistic interpretations to sociological analyses of inequality (Davis 1992; Manning, Price, and Rich 1997; McCammon 1999; Misra 1997). Johnson and Steward (1997), for example, developed an exercise that focuses on the intersection between wealth and politics by asking students to research cabinet members' class backgrounds. Our course projects add to the repertoire by exploring the class backgrounds, political affiliations, and other organizational roles of university trustees and board members of major American corporations.

Since our department offers "Class, Status, and Power" every semester, we typically each teach this required course once per year. We both assign Domhoff's (2002) Who Rules America? and Harold Kerbo's (2003) Social Stratification and Inequality. To complement these texts, we have each developed a power structure research exercise to provide undergraduates with the opportunity to engage in original research and to make new discoveries. In one class, students focus on the boards of trustees of two American universities while in the other class, they investigate the board of directors of a major American corporation. Although students gain different institutional expertise depending on which project they complete, the exercises have identical goals: 1) to reinforce our students' understanding of the texts they read on the upper and corporate classes and 2) to teach our students the skills used by Domhoff and other sociologists so that they are able to conduct research on the corporate community. Since it is often the first time that students have been asked to conduct research that goes beyond synthesis of secondary sources, many students are palpably anxious when they first read the assignment. By the end, however, most report a sense of enormous accomplishment, and a large number get immersed in the project.

Sociological context

The upper class is characterized by old wealth and by a certain lifestyle. Through a network of expensive prep schools, exclusive social clubs, summer resorts, high status philanthropic and cultural organizations, elite universities, and the "right" fraternities and sororities, the upper class has developed a higher degree of class consciousness than any other class in the United States. Rituals such as debutante balls and leisure activities such as fox hunts, polo matches, and yachting facilitate the solidarity and close social interactions between members of the economic elite (Baltzell 1958; Domhoff 1967).

While the material base of the upper class is its ownership role in the primary means of production, a new part of the capitalist class, whose power lies in the control of corporate resources, has emerged alongside it at the top of the contemporary stratification system in the United States: the corporate class. Domhoff (2002) calls the combined upper class and corporate class the "corporate community. " The distinction between ownership versus control of productive property as a basis of power was a major issue in social theory from the era of Weber (1946) and Veblen ([1921] 1982) in the early twentieth century, through Berle and Means (1932) closer to the midcentury, and then Dahrendorf (1959) in the second half of the period. By now, however, sociological theory has "compromised" by accepting Domhoff's concept of a corporate "community" (by which Domhoff appears to designate what classical theory termed the "capitalist class") with different strata.

Teaching about the intertwined corporate and upper classes is central to any course on social stratification. Together, these two elite groups dominate America's economic and political institutions. Of the top 20 industrial corporations, 54 percent of the board members were from the upper class and of the top 15 banks, 62 percent were upper class members (Domhoff 2002). Similarly, in his large study of institutional elites, Thomas Dye (1995) found that 44 percent of 3572 board members and top executive officers of the largest 201 corporations were members of the upper class.

Through campaign contributions, lobby organizations, the policy-formation process, and direct participation in government, economic elites yield considerable political influence. Of 205 cabinet members serving between 1897 and 1973, 66 percent belonged to the upper class before they obtained their government positions (Mintz 1975) as did 63 percent of the secretaries of defense, treasury, and state between 1932 and 1964 (Domhoff 1967). The most affluent Americans are disproportionately represented among the ranks of policy planning groups such as the Council of Foreign Relations (CFR) and the Committee on Economic Development (CED). Members of the corporate and upper classes come together in these groups to discuss policy, to publish and disseminate research, and to arrive at some consensus about what should be done about foreign policy and domestic issues. The 61 directors of the CED held an average of 4.1 corporate directorships and the 22 directors of the CPR held an average of 3.2 corporate directorships. Seventy-two percent of the CED directors belonged to upper class clubs, as did 64 percent of the CFR directors (Dye 1995).

What are the indicators for upper and corporate class membership? In his classic text, Who Rules America?, G. William Domhoff lists three major indicators for upper class membership: 1) a listing in one of the various blue books or the Social Register, 2) attendance in an elite prep school, and 3) belonging to an exclusive social club (2002:5-7). (Domhoff provides a listing of upper class prep schools and social clubs.) Members of the corporate class are chief executive officers (CEOs) or board members of a major corporation and at the same time board members of other corporations (Kerbo 2003). But among the corporate class, there is an inner circle that comprises an even more elite group. This inner group tends to have positions on multiple corporate boards, on boards of large corporations, and on boards of large banks. In addition, they are likely to belong to elite social clubs, to have worked their way up the corporate ladder to top positions rather than to have entered at the top, and to represent corporate interests in other institutions such as foundations, universities, and government (Useem 1984; Kerbo 2003).

Assignment #1: Who rules American universities?

This assignment asks students to address the following question: To what extent do members of the upper and corporate classes govern American universities? To help them explore this question, the instructor gives students the names, places of employment, and job titles of the Executive Committee of Clark University's Board of Trustees (ten members) and of the President and Fellows of Harvard University'S Corporation (seven people). All the information distributed in class is in the public domain. To show students the broad authority that these groups have in governing their respective institutions, the instructor provides descriptions of their responsibilities from university documents.

To determine which of the women and men are upper and corporate class, students consult the following sources: The Social Register; Who's Who in America; Who's Who in Finance and Industry; Standard and Poor's Register of Corporations, Directors and Executives; the Martindale-Hubbell Law Directory; and the website They Rule. In addition, they use the Clark University Library's web page to access Lexis-Nexis for business and corporate information and Academic Universe/Biographical Information for data on individuals. Through these sources, students try to determine who has attended one of Domhoff's (2002) exclusive prep school and/or social clubs. In addition, they gather data on corporate affiliations, government positions, membership in policy planning groups (e.g. Committee on Economic De velopment, Council on Foreign Relations, Trilateral Commission), and board memberships on cultural and philanthropic organizations. If an individual is an officer or board member of a corporation, students must find out the Fortune ranking of the corporation. The assignment requires students to discuss gender, race, ethnicity, and religion (Zweigenhaft and Domhoff 1998) and to determine if individuals are members of the corporate community's inner circle (Useem 1984).

Students must limit their papers to 12 double-spaced pages and organize them into the following four sections: 1) Introduction (thesis statement and brief literature review based on Domhoff [2002] and Kerbo [2003]), 2) Methods (description of the social indicators for upper and corporate class membership, strengths and limitations of these indicators, description of sources), 3) Discussion of Findings, and 4) Conclusion. While the students have two weeks to complete the research project, they are required to bring a copy of their Introduction and Methods sections to class a week after the assignment is distributed. This helps to ensure that they understand the assignment, have formulated a thesis, and that they are basing their research on correct indicators. This project is labor intensive and it would be frustrating for students to devote considerable time and effort only to find out in the end that their indicators were incorrect. Since an indicator of corporate class membership is being a CEO or board member of a major corporation, together we discuss criteria for what constitutes a "major" corporation. Students usually decide that if a corporation is a Fortune 500 company, then it is major. During our discussion of their research methods, it becomes clear that some students confuse a listing in one of the Who's Who books with an indicator of upper class membership. We distinguish between sources that yield evidence of indicators and a text such as the Social Register, which is an indicator. We devote an entire class session to a discussion that includes peer review of the Introduction and Methods sections. Students exchange papers and comment on classmates' drafts in reference to a set of questions that explore thesis development, content, clarity, paragraph construction, proper citations, and writing mechanics.

Since Clark and Harvard are two very different educational institutions in terms of academic prestige and wealth (Clark's endowment is $202 million compared to Harvard's $23 billion), it is not surprising that students fmd vast differences between the Harvard and Clark trustees in terms of socio-economic class. While all the Harvard people appear in the sources and a majority qualifies as corporate class, students uncover relatively little information about the Clark members. Students found that none of the Clark trustees appeared in any of the Who's Who books or in Standard and Poor's. One of the Clark members met one of Domhoff's (2002) upper class indicators by having attended an exclusive prep school. But since he was neither affiliated with any social clubs nor had corporate ties, students suspected that he might be a false positive. The students concluded that the Clark trustees are largely upper middle class people whose power and influence rests on the local level.

In contrast, students could see the strong economic and political influence that the Harvard people commanded on the national and even global scene. Three members of Harvard's Corporation sat on multiple Fortune 100 companies, with a fourth on multiple Fortune 500 companies. The corporations included JP Morgan Chase, New York Life Insurance, Exxon Mobil, Metropolitan Life, Corning, Citigroup, and Ford. Two were former cabinet members and five belonged to the Council on Foreign Relations. The Brookings Institution and the Trilateral Commission often surfaced during the course of the students' research. Several members sat together on other boards including boards of art museums and television stations. Only one person, however, met an indicator for upper class membership by belonging to three elite social clubs. As Domhoff (2002) warned in Who Rules America?, private schools were rarely reported in the reference books and few people chose to be listed in the Social Register.

As they do their research, many students complain of frustration and boredom when they cannot fmd data about the Clark people. They come to learn, however, that silence is a significant fmding and they have the chance to experience the lows, as well as the highs, of the research process. Some students become protective of the Clark board members because they think that the Clark trustees do not look good in comparison with the Harvard members. They interpret fmding evidence that someone is upper or corporate class as the goal and they want the representatives of their university to meet the benchmark. We critically discuss this assumption that the Clark board members are somehow deficient because they do not appear in the sources. A few students actually end up deviating from their indicators just so they can say that some of the Clark people are upper or corporate class. We discuss this impulse to manipulate data to fit a researcher's expectations and the importance of being conscious of and resisting biases. Including the students' own university in the project, then, is both positive and negative. On the one hand, doing research on their own institution inspires curiosity, committnent, and an intrinsic motive for completing the project. On the other hand, this vested interest can lead to bias.

As the students use Domhoff's (2002) indicators of upper class membership, they become aware of the strengths and weaknesses of each indicator. Before working on the research project, they, like most readers, generally do not question the author's research methods. After they began conducting their own research, however, students quickly realized that the Social Register was not useful since even the member of the Harvard Corporation who was clearly upper class was not listed in this source. In addition, students found little information on prep schools and they became critical of this indicator since some knew people who attended schools on Domhoff's list who were not upper class. These methodological problems, however, are a boon to teaching. They force us to acknowledge the limits of indicators while also emphasizing the need for indicators in power structure research. Domhoff's discussion of factors that might lead to false positives and to false negatives only left an impression after students implemented his methods. Becoming critical readers of social science research is one of the most important benefits of participation in this project.

Assignment #2: The boards of directors of Wal-Mart/General Electric/Halliburton

In this assignment students select one of three corporations and research the background of each board member. In fall 2004 their choices included Wal-Mart (14 board members), General Electric (GE) (17 board members), and Halliburton (11 directors). The initial assignment asked students to determine the number and proportion of upper class membership of each board, and to make comparisons to those reported for other boards by Domhoff (2002) and Kerbo (2003). Students relied on the criteria given by Domhoff and summarized by Kerbo. As an alternative, students could choose to determine the inner circle membership of the boards as defined by Useem (1984) or to find the partisan pattern of contributions to presidential campaigns in the 2000 and 2004 election cycles made by board members of any of the three corporations.

Students could choose to work in groups, with a recommended limit of three members. Most took this option. The grading paradigm for group projects and individual projects was the same and the instructor did not, as a rule, intervene in issues of sharing workload or unbalanced contributions. For those who chose to work together the instructor described the project as a "real-life" simulation of workgroups that students were apt to encounter in future employment. The instructor gave a brief introduction to group work, explaining that although groups are able to accomplish more work, the members must invest in coordination and depend on the productivity of others. Most groups seemed to work out their problems fairly well, developing a division of labor in which one student looked up biographies, another prepared charts, and another wrote prose. Some groups divided up names for their searches.

One of the meta-goals of these assignments is to build students' confidence that they can perform original empirical work and to demonstrate the ease with which it can be accomplished in the age of online databases. The instructor devoted most of two class sessions to explaining and demonstrating the corporation exercise by projecting the online search procedures, explicating the corporate filing requirements of the Securities and Exchange Commission, including proxy statements and Form 10s (annual reports) which are major sources of initial biographical and compensation information. During one class session the instructor demonstrated the use of the SEC inquiry machine, EDGAR, and various online services that extract data from it (http://www.sec.gov/edgar.shtml) and demonstrated the online biographical databases, including access to Who's Who, subscribed to by the University library, as well as the use of Lexis-Nexis for finding news stories and filling in biographical facts. He also gave a brief introduction to the Federal Election Commission database on financial contributions to campaign committees and political action committees (http://www.fec.gov/finance_reports.html). Shortly after this exercise began, students showed off their new skills by informing the instructor of a website they had discovered http://www.opensecrets.org/, a very convenient extraction of FEC data.

As in Assignment #1, students soon discovered that prep school attendance and private social club membership data were extremely difficult to find-despite the cooperation of Professor DomhOff who answered emails and phone calls from some of the students. The problem of potential false negatives frustrated many of the students leading some of them to slide over to the inner circle question which was easier to answer. The criterion for inner circle membership is relatively simple: membership on more than one Fortune 500 board. Domhoff (2002) reports that 15-20 percent of directors sit on more than one board (he does not limit his determination to Fortune 500 interlocks).

At the same time that researching WalMart proved conceptually challenging, it was the most popular subject of both social class and political research. On one hand, few of Wal-Mart's board members could be shown to be upper class, including the scions of Sam Walton. On the other hand, 43 percent of the board members (6 of 14) met the criterion for inner circle membership; these links included nine seats on other Fortune 500 boards. The titan of Bentonville, Arkansas is well connected. That this was also true for GE and Halliburton surprised few class members. The political research was unambiguous: Wal-Mart's board members, like the board members of other corporations, gave preponderantly to Republican presidential candidates.

A diverse, global, and profitable conglomerate, General Electric is well embedded in the overall network of corporate interlocks. In comparison to Domhoff's finding that 15-20 percent of directors sit on more than one board (2002), the students found that 59 percent of the GE directors were inner circle members. Of the 17 members of the board of America's fifth ranked corporation among the Fortune 500, ten sat on the board of at least one other Fortune 500 firm, making them members of what Domhoff calls the "general leadership" of the corporate community (p. 24). Two board members were African American and three were female. Together the ten inner circle members had 22 seats on other Fortune 500 Boards.

In contrast to these two very visible firms, the Halliburton corporation has no retail presence and is only discussed in the news in the context of logistical contracts in Iraq and the Vice-President's recent position as Halliburton CEO. As most students do not read newspapers, very few knew about the firm. (This was before the release of Michael Moore's documentary film Fahrenheit 9/11.) Eight of the eleven Halliburton directors (73%) were inner circle directors who had 16 ties to other big firms.


To assess whether or not we had met our two goals-to reinforce our students' understanding of Domhoff's (2002) Who Rules America? and to teach power structure research skills-at the end of the fall 2004 and spring 2005 semesters, we each asked our students to anonymously evaluate their assigned research project. Seventeen undergraduates completed a survey about the uni versity exercise and 19 assessed the corporate board assignment. All but one of the total number of respondents in both classes either agreed or strongly agreed that the research project reinforced their understanding of Domhoff's text. The vast majority (94%) agreed or strongly agreed that the assignment helped them to gain understanding of how the upper and corporate classes dominate American institutions and 80 percent of the respondents in each class asserted that the project allowed them to read Domhoff more knowledgeably.

On the two questions that we asked about learning how to do research, however, students evaluated the corporate boards assignment more highly than they did the university boards project. More of the students who wrote about Wal-Mart, GE, or Halliburton (95%) agreed or strongly agreed that they learned how to do research on the corporate community than those who completed the university assignment (70%); the mean difference in responses was significant at the 95% probability level in a t-test. Similarly the university board project scored significantly lower than the corporate research assignment in building confidence in students' ability to do corporate research (59% vs. 84%). The t-test for the average score on the confidence question was significant at the 98 percent probability level. The difference in scores on these two questions may be related to the amount of time students invested in the respective projects. The Wal-Mart, GE, or Halliburton exercise was incorporated into a longer research paper that comprised a much higher proportion of the total course grade. That the instructor devoted an entire class session to a workshop on online data sources may have also contributed to their ability and confidence. However, over 80 percent of both groups (82% in one class and 84% in the other) recommended that the professors assign the project in future classes.

When we asked students to provide a brief reflection of their experience working on the research assignment, they highlighted the benefits that they gained. Here are some sample statements that are representative of their positive assessments:

It was a good way to apply our readings to current day reality.

The projection [sic] was one of the most interesting that I have done here at Clark.

It really reinforced Domhoff's writings. The feeling of discovery was emphasized by the initial sense of confusion.

Was effective in teaching students how to make connections between the upper and corporate classes.

Didn't mind all the research because it was such an exciting and revealing topic.

The research went very slowly. I dedicated at least 20-30 hrs (maybe more) to just research .... It was worth the effort though.

Even students who complained about the amount of time needed for the data collection portion of their project ultimately thought that they learned from the research experience. In addition to concerns about time demands, students were frustrated by their inability to find evidence for upper class membership.


Both assignments are suitable for any social stratification course. The first assignment on universities can be easily adapted so that students can either compare their own institution to an elite university or research the board members of a single Ivy League school such as one of the Big Three: Harvard, Princeton, or Yale. The second assignment can be either duplicated as is or modified if an instructor wants to substitute different major corporations for Wal-Mart, GE, and Halliburton. The sources listed in the two project descriptions are useful irrespective of the institution selected for the assignment.

By gaining the tools and skills that will permit them to investigate the personnel at the top of firms and organizations, undergraduates are able to reproduce the kind of work about which they are reading in class. While the research projects corroborated the existence of a corporate community including an inner circle of "corporate ambassadors" that ties it together, given the limitations of time and resources, they did not support Domhoff's (2002) major finding of upper class dominance of corporate boards. Our students found few upper class members of the university and corporate boards that they researched. We discuss with our students the reason for the discrepancy between their results and those of Domhoff; Domhoff's research depended on information from social club membership and school alumni lists not typically found in standard available sources such as the Who's Who books. Therefore, students are likely to confront what Domhoff called "false negatives." By recognizing and actually experiencing the strengths and weaknesses of a particular methodology, our students become critical readers of social science. By using Domhoff's methods, students come to appreciate the need for indicators and the limitations of some of his indicators. Furthermore, by understanding the methodological challenges and the intensive labor required to write a book, we hope that our students develop a new respect for the research they read. The hours of work that they devote to any of these projects would not comprise more than a paragraph in any of their assigned readings.

Producing research helps students to become informed consumers of social scientific studies. Most students of undergraduate sociology courses will not become professional researchers. Yet, in their work in human services or business or government, and as readers of newspapers and other forms of media, they will need to be able to investigate social science research. Many think of original research as so technical that they either avoid reading it or are intimidated by the idea of doing it. Even worse, they are reluctant to criticize what they read because they believe the research process is so esoteric that they have no legitimate angle of criticism. These exercises, we hope, strip from original empirical work its mysterious, untouchable quality. As Mills, the great power structure research pioneer, wrote in 1956, in a model democratic society citizens are supposed to compose publics-they should discuss issues and come to reflective judgment. Finding and understanding information put before these publics about those who make up what Mills called "The Higher Circles" is thus part of the method of democracy. We hope these exercises empower democratic citizens.


Angelo, Thomas Anthony. 1993. "A 'Teacher's Dozen': Fourteen General, Research-Based Principles for Improving Higher Learning in Our Classrooms." AAHE Bulletin 45:3-13.

Baltzell, Digby E. 1958. Philadelphia Gentlemen: The Making of a National Upper Class. New York: Free Press.

Berle, Adolph and Gardiner Means. 1932. The Modern Corporation and Private Property. New York: Macmillan.

Dahrendorf, Ralf. 1959. Class and Class Conflict in Industrial Society. Stanford, CA: Stanford University Press.

Davis, Nancy J. 1992. "Teaching about Inequality: Student Resistance, Paralysis, and Rage." Teaching Sociology 20:232-8.

Domhoff, G. William. 1967. Who Rules America? Englewood Cliffs, NJ: Prentice-Hall.

_____. 2002. Who Rules America?: Power and Politics. 4th ed. New York: McGraw-Hill.

Dye, Thomas R. 1995. Who's Running America?: The Clinton Years. 6th ed. Englewood Cliffs, NJ: Prentice-Hall.

Gamson, Zelda and Arthur Chickering. 1987. "Seven Principles for Good Practice in Undergraduate Education." AAHE Bulletin 39:5-10.

Gronlund, Norman and Robert Linn. 1990. Measurement and Evaluation in Teaching. 6th ed. New York: Macmillan.

Johnson, Margaret and Gary Steward Jr. 1997. "Integrating Research Methods into Substantive Courses: A Class Project to Identify Social Backgrounds of Political Elites." Teaching Sociology 25: 168-75.

Kerbo, Harold R. 2003. Social Stratification and Inequality: Class Conflict in Historical, Comparative. and Global Perspective. 5th ed. Boston, MA: McGraw-Hill.

Manning, Robert, Derek Price, and Henry Rich. 1997. "Concierge or Information Desk: Teaching Social Stratification through the Malling of America." Teaching Sociology 25: 15-28.

McCammon, Lucy. 1999. "Introducing Social Stratification and Inequality: An Active Learning Technique." Teaching Sociology 27:44-54.

Mills, C. Wright. 1956. The Power Elite. New York: Oxford.

Mintz, Beth. 1975. "The President's Cabinet, 1897-1972: A Contribution to the Power Structure Debate." Insurgent Sociologist 5:131-48.

Misra, Joya. 1997. "Teaching Stratification: Stimulating Interest and Critical Thinking through Research Projects." Teaching Sociology 25:278-91.

Ostrander, Susan. 1984. Women of the Upper Class. Philadelphia, PA: Temple University Press.

Sernau, Scott, ed. 2000. Social Stratification Courses: Syllabi and Instructional Materials. 4th ed. Washington, DC: American Sociological Association.

Useem, Michael. 1984. The Inner Circle: Large Corporations and the Rise of Business Political Activity in the U.S. and U.K. New York: Oxford University Press.

Veblen, Thorstein. [1921] 1982. The Engineers and the Price System. Reprint, New York: Augustus M. Kelley.

Weber, Max. 1946. "Bureaucracy." pp. 196-244 in From Max Weber: Essays in Sociology, edited and translated by Hans H. Gerth and C. Wright Mills. New York: Oxford University Press.

Zweigenhaft, Richard L. and G. William Domhoff. 1998. Diversity in the Power Elite: Have Women and Minorities Reached the Top? New Haven, CT: Yale University Press.

About the authors

Shelly Tenenbaum is associate professor of sociology and director of the Holocaust and Genocide Studies concentration at Clark University. Her research on ethnic enterprise, mutual aid, gender, education, and identity intersect the broad areas of sociology of American Jews and historical sociology. Her article "It's In My Genes: Biological Discourse and Essentialist Views of Identity Among Contemporary American Jews" (with Lynn Davidman) will appear in The Sociological Quanerly (2007).

Robert J .S. Ross is professor of sociology and director of the International Studies Stream at Clark University, and outgoing Chair of the Section on Political Economy of the World System of the American Sociological Association. His research interests are in globalization, labor, and urban affairs, and his most recent book is Slaves to Fashion: Poverty and Abuse in the New Sweatshops (University of Michigan Press, 2004).