In the target article, Duarte et al. argue that social psychology has become more politically lopsided over the years, with liberals and liberalism all but dominating a field that was once much more politically diverse. They go on to suggest a number of ways that the science of social psychology would benefit from increased representation of conservatives and conservative ideology in our departments, leading journals, and academic discourse. Our goals in this brief commentary are threefold.
First, we question the authors' conclusions about increasing liberal homogeneity in social psychology, and we come to a somewhat different conclusion from our own historical analysis of the field: Social psychology has been (and continues to be) a politically liberal social scientific discipline since at least the 1930s. Second, we consider how liberalism within the field can be both a weakness and a potential strength. And finally, we suggest that increasing political diversity, while important, may be less important for the health of the field than increasing other types of diversity, especially racial and ethnic diversity.
The starting point for our historical analysis of the field is the foundation of the Society for the Psychological Study of Social Issues (SPSSI) in 1936. From the outset, SPSSI members focused on liberal concerns such as racial prejudice, class conflict, and war. Prominent psychologists in this camp included Otto Klineberg, Gordon Allport, David Krech, Ralph White, Dan Katz, Nevitt Sanford, Daniel Levinson, Else Frenkel-Brunswik, Theodore Newcomb, Brewster Smith, and, later, Tom Pettigrew, Herb Kelman, Irving Janis, and Bob Abelson. Most contemporary researchers studying stereotypes, prejudice, and intergroup relations more broadly and, increasingly, environmental and health psychologists, are the intellectual heirs of the SPSSI researchers from the 1930–1960s eras.
However, the field of social psychology during much of this era was comprised of more than just the SPSSI social-justice tradition. A second tradition developed in the 1950s, which centered more on rigorous scientific experimentation and distinguished itself from the “softer” side of social psychology that focused on social issues. Prominent in this group were Leon Festinger, Carl Hovland, Hal Kelley, Don Campbell, Stanley Schachter, John Thibaut, and Bill McGuire. Probably most of these now-deceased researchers also at least leaned toward political liberalism, even though their research was not as clearly connected to politics. Duarte et al. seem less concerned about liberalism in these less politically relevant branches of social psychology.
Duarte and colleagues rely on a survey by McClintock et al. (Reference McClintock, Spaulding and Turner1965) to suggest that psychology was more politically heterogeneous a half century ago than it is today. However, it is difficult to make inferences about the politics of social psychologists from those data. The difficulty arises for two main reasons: (1) The survey lumped the then-small field of social psychology in with the larger other specialties in academic psychology, such as learning, sensation and perception, and physiological psychology, whose subject matters were far from social issues; and (2) as Duarte et al. note, party identification was not as highly correlated with political ideology in the American public then as now: r=.00 in 1956; r=.62 in 2004 (Lewis-Beck et al. Reference Lewis-Beck, Jacoby, Norpoth and Weisberg2008). The “sorting” of political parties into distinct ideological camps is a well-documented but fairly recent phenomenon (LevenduskyReference Levendusky2009), and there is no reason to think that academic psychologists were immune to such broader societal trends. As such, no doubt numerous psychologists who identified as Republicans in the McClintock survey were liberal and some who identified as Democrats were conservative. As noted, we think it is a good bet that most social psychologists in those days were at least somewhat politically liberal, regardless of their party identification.
One of us has long argued that scientists with differing theoretical or ideological priors have two competing models available to them in presenting their findings: the adversarial model and the inquisitorial model (Sears Reference Sears1966; Reference Sears1994). In the present context, the adversarial approach would pit the findings of liberal and conservative researchers studying a common question against one another. Each side would argue its position to the best of its ability, marshalling evidence in support of its position and accepting or disputing the conclusions of its scholarly adversaries. In the legal world, the adversarial model is practiced by the American justice system (i.e., with a prosecution and a defense) and follows the contours of a political debate. In the inquisitorial model, by contrast, ostensibly neutral researchers attempt to gather and present evidence fairly from both sides of an issue. That also has a long legal tradition, of course, and is widely practiced in court systems around the world (e.g., China, Russia, Germany, and Scotland).
When viewed through an interdisciplinary lens, social psychology has many potential adversaries in more conservative disciplines such as economics and political science. Social psychologists can pit, say, a theory of symbolic politics against theories of self-interest or prospect theory against the rational-actor model typical in economics. In these cases, social psychologists can usefully adopt the adversarial model by presenting their best case while those with other philosophies, ideologies, and viewpoints do the same. What typically emerges is a richer, more nuanced picture of the phenomenon under study. Boundary conditions and limitations of each approach can be identified and potentially reconciled.
Nevertheless, when viewing the internal dynamics of a decidedly liberal field, many of the shortcomings identified by Duarte et al. are important. When a field develops a political consensus with no one to argue for other views, an adversarial model is no longer viable and an inquisitorial model is likely to be hampered by bias. With respect to the SPSSI tradition, Duarte and colleagues' point to a hostile climate toward conservatives and illustrate how problems of embedded bias in theory and methods can hinder scientific inference and discourse. It is ironic that a field that so aptly documented the dangers of insular groupthink and ideological homogeneity is so susceptible to them. We agree with Duarte and colleagues that such homogeneity of ideology adversely impacts the field's intellectual richness and creativity.
While a diversity of political viewpoints would benefit social psychological science, we also believe that political diversity is not necessarily the most important type of diversity for the field as a whole to pursue. Exposure to politically diverse viewpoints – and challenges to the insights of the field – is always possible from other social science disciplines and from society as a whole. Social psychologists may therefore do well for the science by increasing their commitment to interdisciplinary collaboration and public dialog. However, diversity in traditionally disadvantaged ethnic groups, racial groups, and economic classes is still sorely lacking, not just in social psychology, but across most of social science as a whole (see Medin & Lee Reference Medin and Lee2012). Given the historical substantive foci of social psychology within the social sciences – often with explicit goals to seek justice for disadvantaged people – we are not sure that political diversity should be the leading priority.
In the target article, Duarte et al. argue that social psychology has become more politically lopsided over the years, with liberals and liberalism all but dominating a field that was once much more politically diverse. They go on to suggest a number of ways that the science of social psychology would benefit from increased representation of conservatives and conservative ideology in our departments, leading journals, and academic discourse. Our goals in this brief commentary are threefold.
First, we question the authors' conclusions about increasing liberal homogeneity in social psychology, and we come to a somewhat different conclusion from our own historical analysis of the field: Social psychology has been (and continues to be) a politically liberal social scientific discipline since at least the 1930s. Second, we consider how liberalism within the field can be both a weakness and a potential strength. And finally, we suggest that increasing political diversity, while important, may be less important for the health of the field than increasing other types of diversity, especially racial and ethnic diversity.
The starting point for our historical analysis of the field is the foundation of the Society for the Psychological Study of Social Issues (SPSSI) in 1936. From the outset, SPSSI members focused on liberal concerns such as racial prejudice, class conflict, and war. Prominent psychologists in this camp included Otto Klineberg, Gordon Allport, David Krech, Ralph White, Dan Katz, Nevitt Sanford, Daniel Levinson, Else Frenkel-Brunswik, Theodore Newcomb, Brewster Smith, and, later, Tom Pettigrew, Herb Kelman, Irving Janis, and Bob Abelson. Most contemporary researchers studying stereotypes, prejudice, and intergroup relations more broadly and, increasingly, environmental and health psychologists, are the intellectual heirs of the SPSSI researchers from the 1930–1960s eras.
However, the field of social psychology during much of this era was comprised of more than just the SPSSI social-justice tradition. A second tradition developed in the 1950s, which centered more on rigorous scientific experimentation and distinguished itself from the “softer” side of social psychology that focused on social issues. Prominent in this group were Leon Festinger, Carl Hovland, Hal Kelley, Don Campbell, Stanley Schachter, John Thibaut, and Bill McGuire. Probably most of these now-deceased researchers also at least leaned toward political liberalism, even though their research was not as clearly connected to politics. Duarte et al. seem less concerned about liberalism in these less politically relevant branches of social psychology.
Duarte and colleagues rely on a survey by McClintock et al. (Reference McClintock, Spaulding and Turner1965) to suggest that psychology was more politically heterogeneous a half century ago than it is today. However, it is difficult to make inferences about the politics of social psychologists from those data. The difficulty arises for two main reasons: (1) The survey lumped the then-small field of social psychology in with the larger other specialties in academic psychology, such as learning, sensation and perception, and physiological psychology, whose subject matters were far from social issues; and (2) as Duarte et al. note, party identification was not as highly correlated with political ideology in the American public then as now: r=.00 in 1956; r=.62 in 2004 (Lewis-Beck et al. Reference Lewis-Beck, Jacoby, Norpoth and Weisberg2008). The “sorting” of political parties into distinct ideological camps is a well-documented but fairly recent phenomenon (LevenduskyReference Levendusky2009), and there is no reason to think that academic psychologists were immune to such broader societal trends. As such, no doubt numerous psychologists who identified as Republicans in the McClintock survey were liberal and some who identified as Democrats were conservative. As noted, we think it is a good bet that most social psychologists in those days were at least somewhat politically liberal, regardless of their party identification.
One of us has long argued that scientists with differing theoretical or ideological priors have two competing models available to them in presenting their findings: the adversarial model and the inquisitorial model (Sears Reference Sears1966; Reference Sears1994). In the present context, the adversarial approach would pit the findings of liberal and conservative researchers studying a common question against one another. Each side would argue its position to the best of its ability, marshalling evidence in support of its position and accepting or disputing the conclusions of its scholarly adversaries. In the legal world, the adversarial model is practiced by the American justice system (i.e., with a prosecution and a defense) and follows the contours of a political debate. In the inquisitorial model, by contrast, ostensibly neutral researchers attempt to gather and present evidence fairly from both sides of an issue. That also has a long legal tradition, of course, and is widely practiced in court systems around the world (e.g., China, Russia, Germany, and Scotland).
When viewed through an interdisciplinary lens, social psychology has many potential adversaries in more conservative disciplines such as economics and political science. Social psychologists can pit, say, a theory of symbolic politics against theories of self-interest or prospect theory against the rational-actor model typical in economics. In these cases, social psychologists can usefully adopt the adversarial model by presenting their best case while those with other philosophies, ideologies, and viewpoints do the same. What typically emerges is a richer, more nuanced picture of the phenomenon under study. Boundary conditions and limitations of each approach can be identified and potentially reconciled.
Nevertheless, when viewing the internal dynamics of a decidedly liberal field, many of the shortcomings identified by Duarte et al. are important. When a field develops a political consensus with no one to argue for other views, an adversarial model is no longer viable and an inquisitorial model is likely to be hampered by bias. With respect to the SPSSI tradition, Duarte and colleagues' point to a hostile climate toward conservatives and illustrate how problems of embedded bias in theory and methods can hinder scientific inference and discourse. It is ironic that a field that so aptly documented the dangers of insular groupthink and ideological homogeneity is so susceptible to them. We agree with Duarte and colleagues that such homogeneity of ideology adversely impacts the field's intellectual richness and creativity.
While a diversity of political viewpoints would benefit social psychological science, we also believe that political diversity is not necessarily the most important type of diversity for the field as a whole to pursue. Exposure to politically diverse viewpoints – and challenges to the insights of the field – is always possible from other social science disciplines and from society as a whole. Social psychologists may therefore do well for the science by increasing their commitment to interdisciplinary collaboration and public dialog. However, diversity in traditionally disadvantaged ethnic groups, racial groups, and economic classes is still sorely lacking, not just in social psychology, but across most of social science as a whole (see Medin & Lee Reference Medin and Lee2012). Given the historical substantive foci of social psychology within the social sciences – often with explicit goals to seek justice for disadvantaged people – we are not sure that political diversity should be the leading priority.