In 95 Theses Anne Norton picks a fight with conventional political science. She asks conventional political science to use different measures (indeed, conceptualize what we do as something other than “measuring”), use different methods, and ask different questions. Some may read this as a threat to an entire way of life. I read it as an intriguing and exciting challenge.
I am sympathetic to much of what Norton argues in this book, but I also struggle with the task of putting her theses into practice. That, I think, is the value of this book. For those of us political behavior scholars who have been trained in traditional empirical methods, reading this book gives pause to our first impulses. Indeed, it reminds us that our sense of what is appropriate political science is not a given, but is the product of history and power, as well as the accumulation of knowledge. By stimulating us to notice the nonessential nature of prevailing modes of political science, Norton has given us the opportunity to improve our understanding of what we do as scholars and the ways we go about doing it.
To capitalize on the tensions that Norton excites, this essay focuses on two challenges that I find to be particularly promising for the field of political behavior. These are the calls to focus on the particular and on the meaning that lies in the ordinary. Considering these challenges in light of our tendencies as traditional political behavior scholars generates the possibility of advancing our understanding of political attitudes and behavior in ways that respect the circumstances of the people we study and refreshes the project of theorizing in this field.
Focus on the Particular
The first challenge I would like to take up is Norton's call to empirical researchers to focus attention on the particular. This is not just an injunction to take social context into account, but to pay careful attention to the creation of context. To Norton, culture is a matrix (T1). It is all around us. It is in institutions, in language, in people, in roles, and in laws. This means that if we want to think about an American national culture, we need to notice how culture is imbedded and produced in certain histories, and in the memories of those histories (esp. T86–90). Culture is not something that simply exists. Particular communities of people in particular places at particular times create culture—and indeed, cultures produce particular communities of people in particular places at particular times.
The challenge for the field of political behavior is generalizability. One might read Norton and ask, if we were only to study particular people in particular places, how would we say something about American political behavior in general? This is both an empirical dilemma, and precisely her point. My first impulse as a political behavior scholar is to judge the utility of work based on how broadly we can generalize from the findings of the sample at hand to a broader population. But the insight that emerges from the contrast Norton presents is that this is but one understanding of generalizability. An alternative is to use samples, cases, or instances to think about processes and develop concepts or theories that themselves have broader application.
Valuing only the first version of generalizability sacrifices pursuing important questions in favor of a drive to generalize from sample to population. A pressing research question in the political behavior literature will illustrate. Much high quality research has been conducted on the question of the competence of the ordinary citizen, specifically levels of political knowledge held by members of the American public. We know from surveys that “the mean is low and the variance high”.1
Converse 1990, 372.
Neumann, Just, and Crigler 1992 chap. 6; Mondak 1995; Prior 2003.
Burns, Schlozman, and Verba 2001, ch. 13.
But there is a whole range of questions that this style of analysis does not address. If we look at particular communities of people in particular contexts, we can ask, how do people create contexts in which knowing little is acceptable or even normative?5
If we observe people trying to interpret mediated messages in the course of political interaction, we can ask, what resources do they use if they do not have large stores of knowledge?6Delli Carpini and Williams 1994, 1998; Gamson 1992.
These questions of social process are important because of the way political communication operates. People listen and understand better when information is put in terms that resonate. Being able to answer how particular people in particular places interpret information can result in a better understanding of the way political information could be packaged to reach people who are otherwise uninterested in the business of self-governance. Another benefit of asking how people make sense of politics when they do pay attention is that it reframes our conception of ordinary citizens from conceiving of them as dolts, to conceiving of ordinary citizen as people acting in a matrix that values and produces doltishness, and actually acting quite reasonably within it.
Obviously I am sympathetic to a focus on particulars. At the same time, my methodologies are rooted in a tradition that values conclusions about broad populations more than explorations of particulars. Here is the dilemma I face. Recall the recent Matrix movies (directed by Andy and Larry Wachowski),8
In Thesis 4, Norton laments the common practice of regarding popular culture as trivial. I follow her lead in making use of movies (i.e. Monty Python, note 92, p.71).
My response is to go and talk to people, to meet with them in their everyday contexts and watch as they relate to others, and relate to me. Meanwhile, my crewmates back on the mother ship want a report of what I have learned. I relay my narratives and as much information as possible to demonstrate the validity of my conclusions. But I am constrained by the reality that they know how to cull certain types of information from the numbers and that they will be more likely to pay attention to my messages when conveyed in a numerical way. Moreover, most of my crewmates are so used to interpreting summaries of large populations that they want something other than information about one group or location. They want information that sums up relationships across all zip codes, all cities, and all states.
I do not think that Norton had the Matrix movies in mind when she wrote, “culture is a matrix.” I simply offer this analogy to convey the obstacles facing someone like me, who seeks to contribute to the profession of political science, and who is also sympathetic to the charges Norton makes. My analogy is imperfect, but the lack of fit underscores the value of Norton's book. For in the practice of political behavior scholarship, the numbers are not the same as individuals' interpretations of their selves and politics. The gap between what we can gather from surveys or experiments and what we can learn from observing people in their own contexts is sizeable enough that the methods are not interchangeable. Our field would be substantially enriched by a greater acceptance of intensive methods such as participant observation, alongside our appreciation of survey methodology and other large-N approaches.
Meaning Resides in the Ordinary
Related to Norton's call to study particulars is an argument that meaning resides in the ordinary. I see at least three ways that the field of political behavior could be enriched by heeding this call. First, political behavior happens at “ordinary” times, not just during elections. Most of our reliable and longitudinal survey data has been collected during and immediately following campaigns, and thus many of our analyses are only possible on electoral behavior. Moreover, for most people, their sole act of political participation, according to conventional definitions, is casting votes, so attention to elections is justified. But if we only look for political behavior in the places we as researchers have defined that it occurs, we miss a better understanding of how people do or do not connect themselves to government.
For example, when election time rolls around, people do not suddenly morph into citizens. Yes, politics defined as the business of government and elected officials becomes more salient, but people do not become fundamentally different beings during election campaigns. Who they are throughout the election cycle informs who they are when they cast votes. That is why knowing more about the behavior of people during ordinary times is not only useful for understanding political behavior in general, but for understanding behavior during extraordinary times like elections as well.
A second implication of Norton's claim that meaning resides in the ordinary is that “ordinary” people create meaning. As I noted above, ordinary citizens typically do not measure up to our discipline's standards of political knowledge or good participation. 95 Theses suggests that rather than dwell on the extent and covariates of this poor state of affairs, we might ask instead, What standards of political knowledge do ordinary folks create for themselves? How do they see their behavior as political? How do they construct their relationships with the government?9
Investigating each of these questions could refresh our theories of civic life and avoid the elitism inherent in asking whether opinions match what they ought to look like under conditions of fuller information.A third way in which we could expand our field by heeding the claim that “meaning resides in the ordinary” is by noting that the politics that are important are not just national and international, but a bit more ordinary. Important politics also happen at the level of local, county, regional, and state governments, perhaps even at the neighborhood level.10
These contexts may carry different meaning for people than spheres of politics with which they can have only impersonal relationships. Ignoring political behavior in such settings perhaps predisposes us to conceive of the ordinary citizen as inept.I do not take Norton's call for attention to the ordinary and attention to particulars as a call to return to Columbia-style voting studies.11
Hers is more than a call to do surveys set within particular cities, or take into account whom people talk to about politics. It is an urging for us to learn how people, together, create meaning by drawing on each other, their physical environment, their particular local histories and notice how they connect these things up with the wider world. Studying local politics is not necessarily the same as studying how people in a particular locality go about the practice of politics.12Joe Soss, personal communication.
I interpret Norton's call to mean a call for the disrespected, time-intensive practice of ethnography and other qualitative forms of data gathering. Ethnography and participant observation can fill many of the gaps left by survey data. In a survey, if you ask political attitude questions, and then combine the results with data on whom people talk to and data on the demographic makeup of a person's neighborhood and city, you get some sense of the relationship between individuals and their context.
But imagine that you have survey data on people in a given community, and then you drive to their city. As you exit the freeway into a particular neighborhood, you notice that the billboards advertise Mercedes—or malt liquor. You notice the way the housing stock, the quality of the lawns and cars, and the age of the schools change as you drive. You get out of your car and walk a bit and notice the smell of a tannery, a paper mill, a brewery, freshly spread manure, or perhaps no smell at all. You shop in the local supermarkets, and notice the presence (or absence) of imported Mexican foods. All of this is valuable information, and it isn't available through conventional survey data archives.
To Norton, artifacts like billboards and smells are not just incidentals. “In culture, ideas become real in the world. They take on material form and, in doing so, are altered and advanced.” (34) Artifacts are both signifiers of culture and resources that people—ordinary citizens, elected officials, journalists—use when interacting. Therefore, to ignore these things is to hamper both positivist and interpretivist research. It hampers positivist research because these things are often factors driving behavior. It hampers interpretivist research because these artifacts are in part tools that people use to understand their world.
Treating Culture as a Variable
If the artifacts and symbols of culture are important for political behavior, how should we take culture into account? When we acknowledge that there are cultural variations across members of the American public, how do political behavior scholars typically acknowledge this in our methods? In analyses of survey data, my first impulse is to use indicators of race, gender, and objective indicators of class such as income, education, and occasionally occupation. But Norton challenges such tendencies. She reminds us that these are categories that are themselves a part of our culture—our culture as social scientists, and for many of us our culture as residents in the United States (T2).
Allow me to illustrate with reference to my current research. I am interested in the ways people in cities around the United States are using facilitator-led dialogue among diverse groups of volunteers to improve race and ethnic relations. For a time, I tried to investigate the ways the outcomes of participating in these groups vary for people of different racial backgrounds. I conceptualized this as the relationship between outcomes and cultures. But by operationalizing “cultures” as self-identified racial categories, I assumed where cultural boundaries begin and end and assumed that these categories are internally homogenous. In my statistical analyses, I included variables for different racial categories. When the coefficients on these demographic indicators were sizeable and significant, my temptation was to conclude that culture matters. But this is mistaken. All I learned is that on the margins the central tendency of attitude change among people of different racial backgrounds varies. My grounding in traditional methods tells me that this is meaningful. The sensibility that 95 Theses ignites suggests it is only a hint that something deeper is going on.
A typical response, and one I use myself, is to then analyze the data separately for different groups of people, trying to be more sophisticated about the possibility that the relationships do not just vary in degree across people of different racial categories, but differ in kind.
But this still does not live up to Norton's challenge. If I want to know how people create meaning out of interacting with one another, I need to know how particular people, set within their own web of identities, connect themselves to others and how these processes in turn result in a group context. Race will still be important here, but it will not be a dichotomous variable indicating that people do or do not have race; and it will be more than an analysis in which race works differently for different people. Norton reminds us that it will instead be a resource that people create as they use it.13
E.g., Cohen 1999; Walsh 2004.
How does one capture that in a survey? My provisional response is that we should use survey data as one window. We can use interaction terms, split samples, and open-ended questions to honor the particularities of individuals' classification into categories. My long-term hope is that Norton's insight will legitimize methods that involve listening to and watching people in their own terms, on their own turf.
Perhaps the greatest challenge to my tendencies as a political behavior researcher is Norton's claim that “culture is not a variable” (T2). She admonishes those who think they can adequately characterize a culture—a dynamic entity more than an object—or presume that a group of people shares a culture. I am sympathetic to this and wary of the arrogance involved in assuming that we can describe culture without acknowledging our own. However, straying from the language of variables is a test of character.
As I conceptualize my own projects, and as I teach research to graduate and undergraduate students, I tend toward the terminology of variables. When I try to help a student plan a feasible research project, we work on the question of “what is it that you seek to explain or understand?” I assume that the answer to that question is a dependent variable. But it is not automatically so. If by “variable,” we mean an object that can be dissected and measured and quantified, then when I seek to investigate the process of political understanding, I am not studying a dependent variable. The tension that Norton generates is a reminder that culture is continuously changing and is something that people give meaning to as they act with others. To presume that we can measure this, or operationalize it into a scale or a series of scales, is to ignore the challenge of understanding how people create culture.
Conflict and Difference
To demonstrate, I would like to probe some of the ideas that Norton proposes about conflict and difference. In my research on dialogue programs communities are using to improve race relations, I am dealing with the broad topic of the relationship between ethnic diversity and civic engagement. One way to approach this relationship is to conceptualize it as “levels of civic engagement, dependent variable; extent of ethnic diversity, independent variable” and ask, does ethnic diversity foster or inhibit civic engagement? But Norton reminds us that we can also ask, how does a society create a context in which engagement across ethnic lines is possible? And, how do people actively construct their notions of civic life in contexts of diversity? In these conceptualizations, what we would seek to understand are processes, not dependent variables.
These questions about the construction of civic life are questions about the ways ordinary citizens balance unity and diversity.14
The inspiration that 95 Theses offers me on these points is to encourage me to not look for the balance between unity and diversity, but how the tension between these two forces in civic life generates new ways of being together. In Norton's conception of culture, people do not reconcile the tension between opposing forces. Instead, it is the existence of this tension that generates meaning in their lives.There are reasons to be wary of the ability of people to create meaning out of both unity and difference. For one, difference democrats (as John Dryzek refers to them) argue that when people engage in deliberation in contexts of difference, the communication tends toward unity, at the expense of difference.15
Dryzek 2000; e.g., Young 1996, 2001; Sanders 1997.
Schildkraut 2002. By “incorporationism,” Schildkraut does not mean a melting pot conception of national identity, but one that celebrates diversity.
One might read these arguments as evidence that unity and diversity are not simultaneously possible. But Norton's understanding of the world thrives on the tension between the two. Thesis 25 states, “Identities are performed … [which] enables—or obliges—one to recognize them as made from, and containing, differences, as multiple and manifold in their singularity.”17
Norton 2004, 49.
Ibid., 54.
Ibid., 60.
Ibid., 61.
Take, for example, the concept of identity. A common understanding of social identity within political science is as psychological attachments to categories of people that are clarified and defined through social interaction.21
This conception is somewhat at odds with Norton's claim that identities contain difference. Indeed, a concern in studies of civic engagement is that shared identities all too often downplay and overshadow difference. When Norton defends the thesis that identities contain difference, she defends by example, by noting the many different people we recognize as all being American. “The historical models for American identity furnished by Abraham Lincoln, Malek al Shabazz, J.P. Morgan, Mother Jones, Billy Graham, Marilyn Monroe, and Allen Ginsberg differ radically, yet all are not only recognizably, but also vividly, American”.22Norton 2004, 49.
Norton is challenging our understanding of consciousness as the constellation of considerations that are accessible to people as they communicate. The consciousness that 95 Theses speaks of is not a mystical ability to be simultaneously aware of oneself and oneness with the universe. It is not even necessarily an awareness. Instead, it is a recognition that the way people understand themselves is informed by their experience in opposing situations and as members of opposing categories. In other words, how people understand themselves as individuals is a function of their experience as community members and vice versa. When acting as members of a community, they may not be aware of themselves as individuals in the sense of being able to articulate that their actions are the product of experience in opposing identities. However, the tools they have for perceiving their current situation are tools that they have created in their attempts to make sense of their world at previous points in time.
This expanded notion of consciousness contributes to the study of political behavior, for example, by encouraging us to move from asking whether unity or difference predominates in contexts of diversity, to asking how do people negotiate the compatibility between community and difference? How do political leaders provide recipes for reconciling the two? And it encourages us to wonder whether and how the political institutions that we create encourage or discourage a balance between these opposing forces.
In particular, it encourages a questioning of our conception of the role of government in helping people negotiate unity and diversity. A limited way of conceptualizing binaries is to see the components of the binary as polar opposites. But 95 Theses reminds us that the part in which most of the production of political life occurs is the area of the continuum between the endpoints. Just as it is the tension between unity and diversity that generates much of the performance of civic life, it is the tension between “governments” and “opposition” that could use attention, too (T23).
There is a tendency to conceptualize civic life as necessarily distinct from government. In the global context, “civil society” typically means a space in polities that is distinct from the government. The presumption is that if it were not, then activity within it would not be able to serve as a check on government.23
In this conception, government intervention in civil society is domination, not democracy.However, in communities throughout the United States, and indeed around the world, there are examples of co-production, of the act of governments and citizens cooperating together to improve civic life.24
Fung 2004; Macedo et al. 2005, ch. 3.
These questions are testament, once again, to Norton's insight. She has provoked us, and called attention to the possibilities in conflict and difference. The potential outcomes are more varied and important than winning or losing, balancing or not. As people go about the business of creating their culture and creating their lives within the culture that is imposed upon them, they take on the task of making sense of difference. We have much to learn from how they do so.
