CHAPTER II

METHODOLOGY

In this chapter, the design for this multinational research project will be laid out. First, the underlying philosophical stance of the entire work will be briefly consider, a conceptual framework that draws from the literature but that attempts to close gaps in that literature that may pertain to the issues under study. Then the methodology to be employed in the research will be considered.

Conceptual Framework

Building on the ideas of Gellner (1983), Anderson (1991), Giddens (1985) and others, the overarching stance taken in this study is to conceive of nation-states not as entities of objective reality but as "imagined communities" (Anderson 1991). The project thus builds on social constructivist (Onuf 1989, Berger and Luckmann 1966) arguments to see national identity as constructed and never fixed; that it can be meaningful to persons; that it is constantly reworked from a repertoire of traditions, myths, and representations; and that it can be employed when it makes political sense (Breuilly 1982, 382).

Furthermore, the emphasis will be primarily on Billig's (1997, 6) conception of banal nationalism: "the ideological habits which enable established nations...to be reproduced." While Billig describes how nations are reproduced co-constituitively—i.e., both by agent (individuals) and structure (the nation)—he focused more on that part of the process that involved how nations inculcate citizen habits. The emphasis here, however, is on the inverse: the way that ordinary people in their everyday lives respond to (supra)nationalist discourses and formulate their world views based at least partly on this response. A goal will be to examine how the two entities—the nation and the EU—are thus being reproduced by their citizens.

This approach shifts the emphasis of study from objective characteristics of the national states or the EU as a supranational entity, and instead focuses on people themselves and how they conceive of their nation's existence and reality, how and to what extent that nation may be changing in the face of an encroaching European Union, and whether that is, in the view of the citizen, a good thing.

To a certain extent, the national discourse “out there” in society is about national character, what constitutes a nation. As Stephenson (1967, 93) has written, "national character is what a nation is prepared to talk to itself and others about.... It is what gives a nation a sense of its own identity." These studies thus begin with discourses about national and European institutions, and touch upon national character matters, but they end with how people perceive those discourses and whether or not they seem attached to one or more of them.

Such an approach necessitates, as Giddens has noted (1985), a strong psychological component to the analysis. Thus, subjective understandings people have of their attachments to political institutions are central to the study. An intensive analytic approach (Brown 1980, 112-114) is utilized in an attempt to gain a deeper understanding of participants' views. An approach based securely on subjective reality would help to ground the analysis in a way that previous theoretical literature on nationalist practices tends not to be (Terhune 1964, Worchel 1997). In that sense, this study joins a growing body of literature on the impact on the self of events of culture-wide significance (Breakwell and Lyons 1996, Wicklund and Oosterwegel 1995, Stryker 1980, Eller 1999).

It may be useful to point out that “subjectivity” referred to here is not employed in the sense of an opposition to scientific “objectivity” (or presumed scientific objectivity, [Rosaldo 1994]), but in the sense of individual, “self-referent notions” (Stephenson 1953, 248). This has the sense of viewing the world through one’s own mind, of individual opinion as a valid object of study. More about the methodology employed in this study and its basis in the scientific study of subjectivity is presented below in the methodology section.

Following Pentland (see previous discussion in the introductory chapter), this project made use of at least two theoretical positions—the supranationalist and the pluralist—as bases for two possible forms of attachment to the EU. In addition, following the nationalist literature, a third and essentially opposing factor—the nationalist—was posited and investigated. Basically, the project then examined whether European participants in this study would connect with these forms of affiliation, or would exhibit different ones. More on these forms of attachment will be discussed below in this chapter.

Social psychological theory, and especially social identity theory (Tajfel and Turner 1986), is naturally helpful to an analysis whose focus is on personal impressions of political discourse. Thus, following social identity theory (SIT), this analysis did not center on proximate group behavior—i.e., behavior within a group where participants can see and hear each other but on how grand cultural and political forces influence the individual, and how national/supranational issues affect individual structuring of social identity.

SIT advances the notion of identity salience, that people will form and reform their social identity from a repertoire of multiple identities, depending on contextual demands. A focus of the study was on those demands that may provide greater or lesser salience for the participant. Similarly both to SIT and some political scientific analyses, nationality here is conceived as resting in large part on an "us vs. them" basis, and attempts were made to focus on outgroups that the individual believes to be relevant.

In contrast to much of the research in SIT, this project eschewed structured experimentation in artificial settings for data collection, instead attempting to gather data in as naturalistic a setting as possible. This meant operating on site, doing fieldwork at various places around Europe and utilizing methodologies such as Q and interviewing techniques to elicit face-to-face, in-depth comments. It is hoped that a project that is focused on subjectivity will contribute to a more nuanced reading of individual differences than is evident in SIT's concept of the self (Cinnirella 1996, 255).

Finally, it may be appropriate at this point, after noting what this study attempts to do, to comment as well on what it does not include. In this study there will be no speculation on the psychological need people have for group attachments. This will be accepted as given and will not be explored in any fundamental way. Rather, emphasis will be on the relative strength of those ties and with which entity at the national or supranational levels people feel more comfortable associating.

There will also be no philosophical speculation on the ultimate nature of nationalist sentiment, whether that might be situational or primordial (Richmond 1981; Davis 1997), but rather an acceptance that nationalist feelings are very real and present. The sense people have of their national identity, it is assumed, constitutes a force in itself worthy of study.

Finally, this project will not attempt to examine "national character" as such, what characteristics of that character might be, or even whether it exists or not in fact. There will be no speculation on whether or not the participants in the study fit a modal personality (Inkeles 1997) that may be associated with a national character. Instead, as noted previously, the focus will be on how people understand their relation to their nation vis-a-vis the European Union.

Methodology

Because the focus of this research is on subjective understandings of political discourse, the methodology employed should necessarily be focused on the individual. To avoid as much as possible the pitfalls of methodological individualism and the confusions of infinite variation, however, the methodology should at the same time attempt to seek out and explain patterns of behavior across individual expression. It should thus bridge the chasm between quantitatively- and qualitatively-based research. And of course it should be rigorous and adhere to all scientific canons. In answer to all these demanding criteria, the primary research tool used in this project is Q methodology (Brown 1980, McKeown and Thomas 1988, Stephenson 1953), an approach that seeks the orderly examination of human subjectivity.

Invented by William Stephenson in the 1930s (Stephenson 1935, 1953), Q methodology involves the rank-ordering of a set of statements (Q sample) by a participant in an array that reflects that person's overall perspective relative to aspects of discourse expressed in society on a particular topic. As Stephenson (1978) observed, there is a "shared communicability" in society on most topics, and one can locate a "concourse" of statements that approximates this. Then, by a process of "focalizing attention," a participant will arrange these statements to reflect his/her subjective structure relative to the topic in question (Stephenson 1978, 28).

Factor analysis of the results from all participants in a study will then bring operant factors to the surface. In this way, Q attempts to bring a structured and statistically relevant analysis to the data. The goal of a Q study is to reveal common groupings of opinion within a given population on a particular topic. These patterns are revealed through factor analysis, a powerful procedure that aids in finding statistically significant groupings in a mass of seemingly highly diverse opinion. This then allows the researcher to investigate the factors in greater depth than otherwise, to help in describing and interpreting shared perspectives.

In the process, Q inverts the traditional procedures of statistical analysis. Often referred to as “R” methodology, traditional quantitative analysis is primarily concerned with correlating opinion with traits such as age or gender across populations, while Q methodology is concerned with correlating patterns within persons (McCormack 1998, 5).1

Stephenson argued that subjectivity is an actual event that can be measured with great accuracy through subjective behavior analysis (Brown 1999) and that Q methodology could be a powerful tool in this approach. For Stephenson, a person’s opinion is an object worthy of study and forms the basis of his or her subjectivity.

Aided by Kantor’s tenets on what constitutes psychological events (1938, 1959) and interbehavioral psychology (1959), Stephenson (1953; 1983, 112-113) attempted to establish the study of subjectivity as a natural science and thus rescue it from a sort of dustbin to which it had been assigned by much of behavioral psychology primarily because of concerns over mentalism (Brown 1997). As Stephenson wrote:

Subjectivity is all about us in every branch of human knowledge–-in religion, politics, economics, literature, science, education, philosophy. Authorities speak with its tongue, and [yet] deny any theoretical and methodological advances that can put subjectivity in line with the universality of science. Objectivity, everyone believes, belongs to the “world outside.” The denial of a place for science in the “world inside,” our subjectivity, is a matter for the ultimate shame of modern science. (1983, 122)

Taking Kantor’s (1959, 16) definition of interbehavior, the “behavior of specific individuals in relation to concrete objects and events,” Brown posits further (1999, 43) a “naturalistic subjectivity” in which a researcher would be drawn to study a person’s “historical relationship to the political information provided and to its meaning when viewed subjectively.”

This does not mean that the approach can reveal the inner workings of the mind, or that a link can be discovered between thought and action. There is a lively debate in the psychological literature over the validity of research into internal causal variables, with the behaviorist view that there are none (Skinner 1953, Wolf 1978). Similarly, Stephenson (1953, 86-100) did not claim that the methodology says anything specific about internal thoughts or internal causal variables (Brown 1999, Febbraro 1995). In any event, this study will steer clear of this debate in simply claiming along with Stephenson that subjectivity is behavior and worthy of study, and that Q methodology is a valid and valuable tool in that effort. Since opinion can be revealed through Q analysis, it can be studied, measured and compared with others’ opinions. As long as this is done with appropriate operations, this can form the basis of a scientific study of subjectivity.

Q methodology has been used primarily in psychology (Stephenson 1987) but has had broad application in fields as disparate as mass communication (Stephenson 1967), phenomenology (Brown 1994; Delprato and Knapp 1994), education (Stone 1966), public administration and management (Brown, Durning and Selden 1999) and health care (Stainton Rogers 1991), among others. It has even been used to study humor (Kinsey 1993) and music (Maxwell, 1999). The literature of Q methodology is extensive (Brown 1968, 1977) and may now exceed 2500 entries (Brown 2000).

The methodology has had a valued place in political scientific research (Brown 1980; Dryzek 1990, 1996, 2002). It has been shown to be especially useful in gauging whether a particular political discourse is resonating within a population, and whether it is resonating in ways that theorists and/or general commentators suppose. Both of these characteristic strengths of Q were especially helpful for this study where the thrust was to learn more about how Europeans attach more with their nation or the European Union, and whether those affiliations might be how analysts and commentators suppose them to be. For a similar example of how Q has been useful in this way, Dryzek (1997) employed it to show that the overarching consensus on liberal democracy traditionally thought to exist in the United States is wrong. Instead, in his study he found that there are other operant patterns of subjectivity that give rise potentially to different resonant discourses on democracy in American society.

Noting Q’s power in thus revealing different discourses, deLeon and Steelman (1999) have called for its use in public policy studies in general to aid especially in fleshing out varieties of opinions on controversial policy issues.

While there have been numerous Q studies on identity (see for example Brown 1984, Davis 1997b, Goldman and Emke 1991, Marshall 1991, Wong and Sung 1998), and recent work using Q in Eastern Europe to study civic culture (Klicperová 1999), attitudes towards economic liberalism (De Holczer 1990) and democracy (Dryzek 1997, 2002), the technique has only recently (Haesly 2001) been used to examine identity change in the context of the European Union. These studies thus join with this pioneering effort in this regard. It is hoped that the procedures and results, as noted previously, will demonstrate the usefulness of Q to this important issue area.

The Concourse. For Q methodology, the concourse represents an important part of the basic phenomenon of interest for study. Taken from all that is spoken or written on a particular topic, the concourse could be a huge number of statements, perhaps on the order of several thousand. Stephenson (1978, 25) in fact posits that the number is theoretically infinite and, in what he called the “law of the concourse,” that it is growing all the time (1983, 75). The statements are reduced using Fisherian design (Stephenson 1953, 101-113) to a manageable number for a Q sample to be constructed. It is this smaller number of statements of the Q sample that is then administered to participants in the study. More is presented below on the particular Q sample for these studies following these more general comments on the properties of a concourse.

A concourse has several empirical properties (Stephenson 1983, 75-76). Since statements are taken from general public discourse, they should all be understood by the participants. But they may be understood differently by each of the participants, even apprehended differently by the same participant at different times of life, an important and telling property. The statements reflect something of the general culture and as such should reflect understandings that people have about an aspect of that culture.

In Q methodology, the basic phenomenon worthy of study is the "whole response" (Brown 1980, 173), or the viewpoint of a person with respect to the topic in question. The concourse is therefore critical in helping to elicit a person’s response to a topic. The goal in collecting the statements from the concourse for the Q sample is representativeness, to present a fair rendering of a particular discourse on the topic in question.

In these studies, the concourse reflects expressions of national and transnational sentiment. The intent was to reveal salient factors within the concourse on national and European Union identity and the degrees of diversity within a community regarding attachment to the nation and/or to a supranational entity. An important distinction thus needs to be drawn on the one hand between what can be described or what has been said by commentators about a nation or a supranational entity, and on the other hand what people think about those descriptions or statements, and whether and to what extent there is agreement with them.

An individual's picking and choosing a way through a sample of statements, finding those that s/he agrees or disagrees with, could be conceived as that individual's attempt to establish an identity on the particular issue in question. It is a way to work out a positive distinctiveness with regards to a topic, or a viewpoint that would distinguish that person from another, which is a critical part of establishing a person's social identity (Cinnirella 1996, 253-254).

The concourse for this project was first collected in 1998 for dissertation work in France by the editor of this volume. Subsequent fieldwork was carried out in France using this concourse. Following the successful completion and defense of the dissertation (Robyn 2000), a cohort of researchers was gathered to then translate this concourse into the several languages and different cultures for this seven-nation project. More about this translation process and problems encountered will be addressed below.

The concourse for the original French study was collected from a variety of sources: books, journal articles, newspapers, and statements by interviewees. Some statements were located in the theoretical literature on nationalism and the EU, some from mainstream discourse and the popular press, some drawn out of and modified from the social psychological literature concerning aspects of motivation, and some from comments elicited in administering Q sorts to European participants. All statements were thus either derived from the literature or otherwise indigenous to the population, and related to concepts of national and supranational identity formation. The statements for this French study are listed in Appendix 1; in this section of the study, only certain ones are excerpted for illustrative purposes. The Q statements for each of the country studies are appended at the end of the appropriate chapter.

It is pertinent to note here that, other than the forms of attachment described below, which themselves were advanced to articulate intellectual positions relative to national and integration theory, there is a dearth of explicit description in the literature on forms of European attachment to national versus supranational identity. This points again to a genuine need in the literature for studies of this kind.

Reiterating the EU discussion in the previous chapter, Pentland (1973) described two theoretical positions on European integration as predominant: the supranationalist and the pluralist. Supranationalists, believing that modern technology and social problems on an international scale compel transnational cooperation and perhaps even possible political unity, are skeptical of the ultimate capacity of the nation-state to deal with the complexities of modern life. They may have been profoundly affected by the past performance of the nation-state, particularly its implication in atrocities in European history. Individuals who hold these views would likely agree with the following statements from the concourse:

The sovereign nations of the past can no longer solve the problems of the present; they cannot ensure their own progress or control their own future. (From the autobiography of Jean Monnet, an architect of the early forms of the European Union)

European disunity is the result of a deliberately-fostered fiction of full national sovereignty and of the absolute effectiveness of national policies. (From a speech by Jacques Delors, former President of the European Commission)

National divisions will disappear as a new generation of Europeans grows up. (From an interview)

The nation-state is too weak to secure us equality and too strong to allow us liberty.... It is the focus of man's irrational, dysfunctional and often destructive emotions. (From an article by David Mitrany, an early proponent of supranational government)

I think of myself as a European first, then my own nationality next. I feel I am a citizen of Europe more than of France. (Derived from the identity literature)

I am very attached to Europe as a whole. (Derived from the identity literature)

Greater European integration can give national minorities more a sense of their own identities. (Elizabeth Meehan, Citizenship and the European Community)

In contrast to supranationalists, pluralists might have a greater appreciation for the diversity of national cultures, and the inevitable need for a national state to protect and defend that culture. They might see European integration as a much more deliberate process of interaction, debate and discussion among nations whose sovereignties might be surrendered only under the most critical circumstances. According to pluralists, any emergent supranational government would have to be run at the behest of the member states. Along with nationalists, to be described below, pluralists see a natural limiting factor to political integration: "a significant but persistent core of national identity in each state, in whose service governments may feel obliged to act in ways unproductive for integration" (Pentland 1973, 47). Pluralists therefore may have a mix of national and supranational identity tendencies but allied strongly to neither. In the sample of statements, the following are representative of the pluralist viewpoint:

I want a Europe of nation states that are as politically and culturally different as they are geographically diverse. (From a speech on the future of Europe by former British Prime Minister Margaret Thatcher)

Europe will be stronger precisely because it has France as France, Spain as Spain, Britain as Britain, each with its own customs, traditions and identity. It would be folly to try to fit them into some sort of identikit European personality. (From the same speech by Margaret Thatcher)

Unity in Europe does not create a new kind of great power; it is a method for introducing change in Europe ... it is not a blueprint, it is not a theory, it is a process that has already begun of bringing peoples and nations together to adapt themselves jointly to changing circumstances. (Jean Monnet, from his autobiography, and in his later period of more pluralist leanings)

European countries are diverse, but it is possible for a European Union to work. (From an interview)

Our national leaders are capable of managing our institutions in a global environment. (From an interview)

Both the supranationalist and the pluralist positions above share an interest in and a certain commitment to political integration in one form or another. The significant differences between the two relate to the speed and inevitability of the process, and the character of the resulting governance structure.

What is not investigated in the analysis so far is a view diametrically opposing both. That view would be the nationalist. More than simply skeptical of supranational governance, the nationalist would be opposed to it. Like Bauer (1924, 135), the nationalist would have a belief in a nationally shared "community of fate," a common history of glory and defeat that would tie together a people within a certain bounded territory into a "community of character." Like Montesquieu and Herder, the nationalist would believe explicitly in the idea of a national culture, and accept that it differs from any other culture. This fundamental incommensurability would inevitably thwart true political union, not simply slow it down or change in some way its character, as a pluralist might argue.

Drawing on the nationalist literature from Montesquieu and Herder to Bauer and Kohn, Smith (1983, 21) derived seven propositions from the "core nationalist doctrine":

1. Humanity is naturally divided into nations.

2. Each nation has its peculiar character.

3. The source of all political power is the nation, the whole collectivity.

4. For freedom and self-realization, men must identify with a nation.

5. Nations can only be fulfilled in their own states.

6. Loyalty to the nation-state overrides other loyalties.

7. The primary condition of global freedom and harmony is the strengthening of the nation-state.

Along with several of these core propositions, other statements of nationalist character were collected. Thus, the following statements in the concourse are representative of the nationalist position:

Humanity is naturally divided into nations with their own peculiar characters. (From Smith, above)

Globalization means the death of my national culture. (From an interview)

I do not hate foreign persons in my country, but I feel they are polluting our national identity. (From a speech by Jean-Marie Le Pen, leader of the right-wing Front National in France)

Loyalty to the nation-state overrides other loyalties. (Smith)

The baptism of Clovis is the baptism of France. (Slogan of a right-wing political group in France on national origins)

The European Union represents an unattractive homogenization of European cultures. (Constructed by the author from political discourse on European Union matters)

European countries are too diverse for a European Union to work. (Interview)

The Nation is essentially the source of all sovereignty; nor can any individual, or any body of men, be entitled to any authority which is not expressly derived from it. (Declaration of the Rights of Man, Article III)

France is not really herself except in the first rank of the countries of the world. (From the autobiography of Charles de Gaulle)

Our identity as a nation is migrating to the European Union and we are left empty as a result. (Constructed by author from political discourse on European Union matters)

The nationalist/pluralist/supranationalist sentiment is thus displayed schematically in the Q sample structure in Table 1:

TABLE 1. Q Sample Structure for (Supra)National Identity Study

Main effect Levels N

Identity Nationalist 12

Pluralist 12

Supranationalist 12

__________________________________________________________

The Q sample structure is incorporated in the full listing of the Q statements found in the appendix to the French chapter.

Following the successful administration of these statements in the French study and once the decision was made to attempt a cross-cultural multinational study, the task was to construct Q samples for each of the countries under study. This was done by the authors of each of the chapters in this book.

In most cases, simply translating the statements directly posed no special linguistic problem. But for several of the statements—especially numbers 5, 15, 24 and 27-the meanings were so directly connected to the French experience that alternative statements needed to be found that would convey the same meaning to persons of that country. Using their intimate knowledge of the cultures in question (for most of them their native cultures), the researchers then found cultural equivalents for these statements.

Some may object to the cross-cultural nature of this project, arguing that Q is so finely tuned to the particular culture that it cannot be safely carried out of that context. In their recent study on societies in transition to democracy, for example, Dryzek and Homes (2002) argue as much, insisting in their multinational studies that their researchers in the field in effect construct culture-specific Q samples.

However, we feel on safe ground in arguing that, with the few statements especially reconstructed as mentioned above, the rest of our statements can cross the cultures represented in our studies. Our approach can accommodate this because of the subject matter. The intent of the entire European integration project, after all, is to establish some commonalities among Europeans. These commonalities may involve basic principles of federalism or the dynamics of European integration, which can be discussed with the same statements across cultures because a European-wide discourse is taking place on these issues. Or the commonalities may simply be to establish and react to the EU as one common source of governance. At the other end of the spectrum in question, the national, the notion of national identity may have particularities for each nation but has as well universal principles, as Smith has pointed out.

In any event, the Q samples in these studies provide less a theoretical construct than a departure point for further exploration of issues of subjective impressions of national and supranational identity. The factors that would eventually emerge from these investigations, as described in each of the country chapters that follow, may or may not reflect the theoretical constructs described above, and might reveal different positions relative to the question under study.

Since this did in fact occur, and four or more factors were revealed for most of the countries, rather than the three posited in the literature, the task was then to examine those various factors that did in fact emerge, to gauge whether they had significant strength, to describe them, and finally to speculate about whether people have actually changed with respect to the theoretical issues Pentland laid out.

It could well be asked at this point how it is that any other forms of identity could emerge if only the statements from these three forms were utilized in the concourse. Perhaps there is a position that would not be represented by these statements. In response to this question, it should first be remembered that only these three possible identities have been posited in the literature, and thus the study to some extent is limited by the limited discourse that has been theorized on this topic. People are similarly confined in the choices they can make about any topic from the discourses available to them in society.

Secondly, the question misunderstands the potential variety of opinion that Q allows participants. Q principles and procedures permit great variability in individual attitudes. The participant is not making unconnected judgments about a series of statements, but is asked to weigh the response to a statement in conjunction with other statements. In this way the sorted Q sample is a picture of identity in which statements are woven into a larger mosaic. Each person was thus asked implicitly to compare each of the N=36 statements with the other, resulting in (½)(N)(N-1)=630 different possible judgment evaluations.

In addition, in simply separating out three piles of statements (agree, disagree or neutral), the way that a participant normally begins a Q sort, the variety potential is on the order of 3.384731763 × 10¹⁵. Putting this into perspective, with 6 billion people in the world, there are 564,122 times as many ways to separate the 36 statements into only three piles as there are people in the world.

This would seem to indicate in itself plenty of opportunity for individual variability in the sorting process. But participants do not stop at sorting into only three piles, and instead continue to place the statements into an array (described below). Here, even more variability is displayed, for there are 36! = 3.719933267 × 10⁴¹ different ways that the statements could be rank-ordered from “agree” to “disagree.” It should be noted here that a Q-array is a modified form of rank ordering in which each statement in a column is not ranked in relation to other statements in that column, but instead in relation to the statements in each of the other columns. This has the effect of reducing the potential for variability noted above, although it will still be quite large.

Allowing such latitude in subjective interpretation contributes to the credibility of the factors that emerge in the study. When people have this much opportunity to express individuality but nonetheless ultimately group into distinctive factors, it is strong evidence that that form of discourse resonates in the population.

In its remarkable capacity for allowing this kind of individual expression, Q is of “almost subatomic nature” (Stephenson 1983, 78) in its focus. It thus responds well to Kantor’s (1978) specificity principle in scientific analysis, which alerts researchers to the huge potential for individual variability in natural phenomena. As Kantor wrote on the specificity principle as applied to psychology, “no two individuals are alike nor perform any type of adjustmental action the same way in recurrent situations” (1978, 123). Research in psychology must account for such potential variability. In its attention to specificity, Q is particularly strong in this regard.

Yet there is another side to Q as well, its attention to communability; that is, what an individual’s response has in common with responses from others. Even with all this potential for variability, in this study (as in all Q studies) participants grouped themselves together onto a relatively small number of factors. These factors will be described in the chapters that follow.

This communability of responses has two key advantages for a Q study: it permits the researcher to more easily describe factors that actually exist in the community and it allows a smaller number of participants to make up a valid population set. If, despite such great potential for individual variability, people actually do fall into relatively small numbers of groups of opinion, then a proportionately smaller number of participants may be used in the population sample. This is another important distinction between R and Q methodologies. Adding more participants will likely not add more factors but simply more persons on each factor, displaying similar patterns of identity. The respondent design helps insure diversity in the population sample, and thus a larger “net” to catch the necessary “fish.” A number of studies attests to the fact that repetitious patterns arise as larger numbers of participants are included in the sample (see Coke and Brown 1976; Dryzek, Clark and McKenzie 1989). More about the population sample for this study is presented below.

What makes people reduce to such groupings? Why are we not as idiosyncratic in our views as some would have us, or as we ourselves might wish to be at times? Speculating in this way on basic human nature would be beyond the scope of this study. It may be enough to say simply that factor analysis proves that it is so. But Stephenson did theorize on this (1983, 93-94) and posited a “law of affectability” in which he argued that groupings in the concourse form by way of “confluences of feeling.” Feeling is

a common element running through a factor from one end of it to the other, in such a way that the statements of the Q sample arrange themselves in a perfect order, each statement in its appropriate place, like pieces of a jig-saw puzzle fitted neatly into place. (1983, 93)

With such a powerful tool to reduce variability to manageable proportions for study, it is no wonder that one scholar (Duijker 1979, 18) lauded Q in this way:

Human situations are to some extent like snowflakes: They are innumerable, they exhibit a multitude of forms, and above all they are highly perishable. Q-methodology, Stephenson’s great contribution to psychology, was designed to deal with this “infinite variety,” to make it accessible to scientific explanation without distorting it.

Procedures. In these studies, participants were asked to sort 36 statements that make up the Q sample, 12 each from the three theoretical positions described in Table 1, from agree (+4) to disagree (-4), following this instruction in the language of the country: "These are statements made by people concerning national identity and the EU. Sort these statements from those you agree with to those you disagree with." The point of the exercise was to elicit subjective impressions and as such, there was of course no "correct" manner of sorting the statements. The data gathered were then factor analyzed using PQMethod (Schmolck and Atkinson 1998), a computer software program made available specifically for Q-methodological studies, in order to gauge the emergent factors.

Participants were first asked to read over the statements to get a broad impression, then to sort them into three piles—agree, disagree and more or less neutral—before more carefully examining the statements to decide finally how they would fit precisely into the array. Thus, Q is a "forced choice" methodology (see discussion of the forced choice aspect of Q methodology in Brown [1980, 201-203, 288-289] and Stephenson [1953, 59-61]) in which participants are required to choose how to characterize their attitudes about a statement and set it into a general descriptive picture about nationality and supranationality. For example, there are only three spots in the array for those statements that the participant would feel were most characteristic of his/her feelings about the topic. This requires the person to make distinctions that otherwise might not be made in the course of thinking about or discussing a topic.

The remainder of the statements are then set into the array, according to a continuum represented in Appendix 2. When the sorting is finally done, it should be structured as in the continuum, with statements "most disagree with" to the left and those "most agree with" to the right, with the statements more or less "neutral" in the middle.

The neutral area in the middle part of the Q sort (0, +1, -1) is reserved for those statements that create conflict for the participant, or about which the person feels less certainty. Far from being an area of less interest to the observer, the more neutral statements can be equally as critical to revealing how the participant feels. These statements can be particularly useful in eliciting discussion, clarifying the meanings of statements, or probing the feelings of the participants.

When done, this array is a structured representation of the subjectivity of the participant on the topic of nationality and supranationality. Subjectivity emerges from virtually the beginning of the sorting process as a participant decides how to place the statements, and as the participant and the observer discuss how the statements are placed.

The underlying dynamic is that of Stephenson's "psychological significance" (cited in Brown 1972, 81-84), that people react to these statements in terms of the degree of significance they feel the statements have for them. Especially statements at either end of the array are of greatest salience for the individual, whether in a positive or negative sense. It would be in this part of the structure that individuals would be expressing those sentiments that they would be most attracted to or repelled by, in effect creating visibly their social identity. Those statements in the middle part are of lesser significance, although as stated previously, they can be of great interest in establishing more nuance for the person's social identity.

The P-Set. Because of the intensive nature of Q Method and because, as we have seen, factors within discourses are relatively few in number, the methodology can be administered to a comparatively (i.e., compared to quantitative procedures) small group of persons. In Q Method parlance, the group of participants in a study is called a population set, or P-set for short. The P-set in the original French study, described in more detail in the appropriate chapter, was comprised of 37 French participants. The other country studies ranged from that to a few more than 40 each. In total, then, more than 300 Europeans participated in the entire project.

In a study with the objective to examine how nationalist/supranationalist discourse is resonating in the population, participants needed to be selected on a purposive basis and come from a wide variety of backgrounds and from varied settings. The P-set for the project is represented in Table 2.

The search was for representativeness in the sample. This is not to be confused with random sampling; breadth and diversity are more important in this case, based on how the participants are likely to exhibit "theoretical saturation" (Glaser and Strauss 1967). Theoretical saturation refers to the point beyond which “no additional data are being found whereby [the researcher] can develop properties of the category” (Glaser and Strauss 1967, 61). Beyond this point, research is simply adding more participants to the study and nothing to the theory.

As a reminder, in Q methodology, representativeness is as much or more in the Q-sample statements as in the P-set. In effect, what is being sampled are statements from the discourse, and not the persons themselves. Thus the concern in constructing the P-set is simply to cast an effective “fishing net” in an attempt to understand perceptions among a group of people. In Q, one is not searching for concreteness in generalizations to the population in its entirety, or findings based specifically on personal characteristics of the participants.

The P-set has then been constructed to be as broad as possible, and yet divided along levels in which one might expect to find significant national identity differences. In these studies, the levels were determined to be: age, gender, socioeconomic class and setting.

Older Europeans, who are likely to have experienced World War II and/or its aftermath, might have internalized a deeper attachment to national symbols than younger persons. As we have seen, the European Union has often been characterized as an elite construction based largely on considerations of global economic matters. One might then expect differences in attitudes based on class and professional working status.

Because global economic matters tend to impact differentially in urban compared to rural settings, it might be expected that greater support for supranationality might be exhibited among city dwellers. At least one analyst (O'Brien 1993) has found just such a noticeable urban/rural dichotomy over questions of national identity.

Interviews with a variety of French participants during the lead-up to the original study suggested differences based on gender as well: French males, it is suggested, many of whom take part in military service and also participate more (at least as spectators) than women in sporting events that often take on nationalist character in Europe, might be expected to exhibit more nationalist leanings.

In Table 2, it will be noted that there are 16 combinations within each sample. Participants for the studies were selected based on a combination of the four effects listed. Thus, one participant was a young working-class male from an urban setting (aceg), another a young working-class male from a rural setting (aceh), and so forth for the remainder of the 16 combinations. In general, we aimed for two participants of each type, which would yield a P-set size of n = 32.

TABLE 2. P-Set Structure for (Supra)National Identity Study

Main effects Levels N

_____________________________________________________

A. Gender (a) male (b) female 2

B. Age (c) young (d) older 2

C. Class (e) working (f) middle 2

D. Setting (g) urban (h) rural 2

_____________________________________________________

ABCD = (2) (2) (2) (2) = 16 combinations

Setting. As the following chapters will show, each country offered its own interesting “living laboratory” for analysis of national/supranational identity issues. A particular country’s history and contemporary relationship with the European Union movement would offer hypotheses for prediction on whether its citizens would have generally favorable or unfavorable attitudes towards a European ideal. These hypotheses would then be tested by the research from the field.

To fulfill setting requirements for the P-set, we administered roughly half the Q sorts in urban and half in rural settings. The quality of the data gathered was only enhanced by the fact that nearly all of us live on site in the countries under study and so know the conditions well. The briefest time spent in country was for the original French study, and that was a lengthy three months.

Interviews. In addition to Q, we relied on in-depth interviews to validate and cross-check observations, and to reduce systematic bias in data collection. This can be critical when studying an issue of such complexity as national identity. It helps to avoid the limitations of observation inherent in qualitative approaches (Patton 1980, 244-245). These limitations include how the observer can affect the study situation, the selective perception of the observer, and the possibility of the participants' observations being subject to bias, anger, emotional feelings of that day, to recall error, to reactivity to observer, or to self-serving responses.

Q sorts were set up with various participants who best fit the P-set criteria. During the Q sorts we asked often detailed questions prompted by how the participants were constructing the sort. Finally, in some cases we could return to do more in-depth interviews with certain participants who seemed best to exemplify factors that emerged in the study, to understand better the factors highlighted from the Q sorts.

Second-Order Analysis

The preceding has described the procedures used in the administration of the Q sorts in each of the countries of these studies. From the data that had been gathered and analyzed, arrays were constructed for each factor from each country. These groupings are then described in the chapters that follow.

Once this data had been thoroughly examined in its particulars by each author, then another analysis was performed on all of that data together by the editor of this volume. This second-order analysis takes all of the arrays from all of the various factors from each of the countries, re-enters them into the PQMethod program to factor analyze them again, this time in comparison with one another. The result of this second-order analysis is a rigorous recheck of all factor arrays, and a valuable overview of the entire project, providing useful comparisons of the relative strength, intensity and position of the factors cross-culturally. The results of the second-order analysis provided the grounding for the narration of overall findings in the introductory chapter.

Appendix 1.

Q Sample and Factor Arrays:

European/French National Identity

Factor Arrays

Statements A B C D

1. Our identity as a nation is migrating to the European Union ‑4 2 ‑2 -1

and we are left empty as a result. (N)[1]

2. Rapid technological innovations and the global economy has 0 0 2 ‑1

contributed to my optimistic mood for France. (SN)

3. Humanity is naturally divided into nations with their own ‑1 4 4 4

peculiar characters. (N)

4. Globalization means the death of French national culture. ‑3 3 -3 3

(N)

5. I do not hate foreign persons in my country, but I feel ‑4 0 -3 ‑2

they are polluting our national identity. (N)

6. The sovereign nations of the past can no longer solve the 4 -2 1 0

problems of the present; they cannot ensure their own

progress or control their own future. (SN)

7. Loyalty to the nation‑state overrides other loyalties.(N) ‑2 2 0 ‑1

8. Adopting a single European currency will be difficult but 3 -1 4 1

necessary for France. (PL)

9. My parents would be unhappy if I were not married to ‑3 0 ‑3 ‑4

someone who was also French. (N)

10. The European Union is a means to ensure peace and 3 -4 4 0

stability for France in the future. (PL)

11. Our national leaders are capable of managing our 0 -1 1 ‑1

institutions in a global environment. (PL)

12. I think of myself as a European first, then my own 4 -2 ‑4 ‑4

nationality next. I feel I am a citizen of Europe more

than of France. (SN)

13. I want a Europe of nation‑states that are as politically ‑1 4 ‑1 2

and culturally different as they are geographically

diverse. (PL)

14. I am very attached to Europe as a whole. (SN) 2 0 3 ‑4

15. The baptism of Clovis is the baptism of France. (N) ‑2 3 ‑2 -2

16. Unity in Europe does not create a new kind of great power; 2 1 ‑1 0

it is a method for introducing change in Europe. It is not

a blueprint, it is not a theory, it is a process that has

already begun of bringing peoples and nations together to

adapt themselves jointly to changing circumstances. (PL)

17. France must maintain an independence in European affairs. ‑1 2 3 2

(N)

18. I want France to represent an alternative to American 0 0 2 3

influence, another voice in international affairs. (PL)

19. Europe will be stronger precisely because it has France 0 4 3 4

as France, Spain as Spain, Britain as Britain, each with

its own customs, traditions and identity. It would be

folly to try to fit them into some sort of identikit

European personality. (PL)

20. The European Union represents an unattractive homogenization ‑3 1 ‑1 1

of European cultures. (N)

21. The pace of European integration is too slow. (SN) 1 -1 ‑2 ‑2

22. European countries are too diverse for a European Union to ‑1 1 ‑1 2

work. (PL)

23. National divisions will disappear as a new generation of 0 -4 0 ‑1

Europeans grows up. (SN)

24. France is not really herself except in the first rank of ‑2 3 ‑2 ‑3

the countries of the world. (N)

25. European disunity is the result of a deliberately‑fostered 2 -3 0 0

fiction of full national sovereignty and of the absolute

effectiveness of national policies. (SN)

26. I think European integration has gone about as far as it ‑1 -1 ‑1 ‑3

can go. (PL)

27. French wine, cheese and perfume are the best in the world. 0 1 2 2

(PL)

28. If someone said something bad about being French, I would ‑2 2 0 ‑3

feel as if they had said something bad about me. (N)

29. I may obey laws, but I don't feel especially patriotic. 1 -3 ‑4 4

(SN)

30. Our society is united by a delusion about our ancestry and 1 -2 ‑4 1

by a common hatred of our neighbors. (SN)

31. It is possible to separate political identity from cultural 1 0 1 3

identity. (PL)

32. In general my country has definitely profited from its 3 -4 2 ‑2

involvement in the EU. (SN)

33. The nation‑state is too weak to secure us equality and too 1 -2 0 1

strong to allow us liberty.... It is the focus of man's

irrational, dysfunctional and often destructive emotions.

(SN)

34. Federalism with subsidiarity is a way to reconcile what 2 -3 1 1

appears to be irreconcilable: encouraging the emergence

of a united Europe while keeping loyal to one's homeland

at the same time. (PL)

35. The Nation is essentially the source of all sovereignty; ‑4 1 1 0

nor can any individual, or any body of men, be entitled

to any authority which is not expressly derived from it.

(N)

36. Greater European integration can give national minorities 4 -1 0 0

more a sense of their own identities. (SN)

Appendix 2.

Q Array

(with statements indicated for composite

Factor 1 of French Study)

Most Disagree Most Agree

‑4

‑3

‑2

‑1

REFERNCES

Anderson, Benedict. 1990. Imagined Communities: Reflections on the Origins and Spread of Nationalism, 2nd ed. London: Verso.

Bauer, Otto. 1924. Die Nationalitatenfrage und die Sozialdemokratie, 2nd ed. Vienna: Brand.

Berger, Peter L. and Thomas Luckmann. 1966. The Social Construction of Reality. New York: Doubleday.

Billig, Michael. 1997. Banal Nationalism. London: Sage.

Breakwell, Glynis M., and Lyons, Evanthia, eds. 1996. Changing European Identities. Oxford: Butterworth-Heinemann.

Breuilly, John. 1985. "Reflections on Nationalism." Philosophy and Social Science, 15:55-78.

Brown, Steven. 1968. “Bibliography on Q Technique and Its Methodology.” Perceptual and Motor Skills, 26:587-613.

_____________. 1972. "A Fundamental Incommensurability Between Objectivity and Subjectivity." In Science, Psychology and Communication, eds. Steven R. Brown and Donald J. Brenner. New York: Teachers College Press.

_____________. 1980. Political Subjectivity. New Haven: Yale University Press.

_____________. 1984. "Values, Development and Character: Appraising Korean Experience." Korea Fulbright Forum, 1:33-66.

_____________. 1994. “Q Methodology and Interbehavioral Phenomenology.” Interbehaviorist 22(3): 24-26.

_____________. 1999. “Subjective Behavior Analysis.” Presented at the annual meeting of the Association for Behavior Analysis, Chicago.

_____________. 2000. Personal email correspondence. January 21.

Brown, Steven, Dan Durning and Sally Selden, eds. 1999. “Q Methodology.” In Handbook of Research Methods in Public Administration, eds. Gerald Miller and Marcia Whicker. New York: Marcel Dekker.

Cinnirella, Marco. 1996. "A Social Identity Perspective on European Integration." In Changing European Identities, eds. Glynis M. Breakwell and Evanthia Lyons. Oxford: Butterworth-Heinemann.

Coke, James and Steven Brown. 1976. “Public Attitudes About Land Use Policy and Their Impact on State Policy-Makers.” Publius 6:97-134.

Davis, Thomas C. 1997a. "The Subjectivity of National Identity: A Q-Methodological Study of Individual Attachment to the Basque Nation." Ph.D. diss. Kent State University.

Davis, Thomas C. 1997b. "Patterns of Identity: Basques and the Basque Nation." Nationalism & Ethnic Politics, 3:61-88.

De Holczer, Paul. 1989. "Economic Nationalism and Liberalism: A Q Methodological Study of Hungarian Economics Students." Masters thesis. University of South Carolina.

DeLeon, Peter and Toddi A. Steelman. 1999. “The Once and Future Public Policy Program.” Policy Currents, 9(2): 1-9.

Delprato, D.J. and J.R. Knapp. 1994. “Q Methodology and Interbehaviroal Description.” Interbehaviorist, 22(3): 14-23.

Dryzek, John. 1990. Discursive Democracy. Cambridge: Cambridge University Press.

____________. 1996. Democracy in Capitalist Times. New York: Oxford University Press.

____________. 1997. Personal correspondence, October 12.

Dryzek, John and Leslie Holmes. 2002. Post-Communist Democratization. Cambridge: Cambridge University Press.

Dryzek, John, Margaret Clark and Garry McKenzie. 1989. “Subject and System in International Action.” International Organization 43:476-503.

Duijker, Hubert C. 1979. “Mind and Meaning.” Operant Subjectivity, 3:15-31.

Febbraro, Angela R. 1995. “On the Epistemology, Metatheory, and Ideology of Q Methodology: A Critical Analysis.” In Trends and Issues in Theoretical Psychology, eds. Ian Lubek, Rene van Hezewijk, Gail Pheterson and Charles W. Tolman. New York: Springer.

Gellner, Ernest. 1983. Nations and Nationalism. London: Cornell University Press.

Giddens, Anthony. 1985. The Nation-State and Violence. Cambridge: Polity Press.

Glaser, Barney G. and Anselm L. Strauss. 1967. The Discovery of Grounded Theory. Chicago: Aldine.

Goldman, Irvin, and Ivan Emke. 1991. "Communication and Canadian Identity: A Q-Methodological Study." Canadian Journal of Communication 16:129-138.

Haesly, Richard. 2001. “Euroskeptics, Europhiles and Instrumental Europeans: European Attachment in Scotland and Wales.” European Union Politics 2(1): 81-102.

Inkeles, Alex. 1997. National Character. New Brunswick, NJ: Transaction.

Kantor, Jacob Robert. 1938. “The Nature of Psychology as a Natural Science.” Acta Psycholgica 4:1-67.

Kantor, Jacob Robert. 1959. Interbehavioral Psychology, 2^nd edition. Granville, OH: Principia.

Kantor, Jacob Robert. 1978. “The Principle of Specificity in Psychology and Science in General.” Revista Mexicana de Análisis de la Conducta 4:117‑132.

Kirton, D. 1995. “Race, Identity and the Politics of Adoption.” Working Paper 2, Centre for Adoption and Identity Studies, University of East London.

Klicperová, Martina. 1999. Ready for Democracy? Civic Culture and Civility with a Focus on Czech Society. Prague: Institute of Psychology, Academy of Sciences of the Czech Republic.

Marshall, Harriette. 1991. "The Social Identities of Women Lawyers." Operant Subjectivity 14:106-121.

McCormack, Una. 1998. “Playing at Politics: First-Time Voting in the 1997 UK General Election”. Politeia (Social Science Research Series No. 1). Cambridge, UK: Fieldfare.

McKeown, Bruce and Dan Thomas. 1988. Q Methodology. Newbury Park, CA: Sage.

Maxwell, Jennifer P. 1999. “The Violence of the Night: The Perception of Trauma in a Song,” Operant Subjectivity, 22: 12-30.

O'Brien, Oonaugh. 1993. "Good to be French? Conflicts of Identity in North Catalonia." In Inside European Identities, ed. Sharon Macdonald. Oxford: Berg.

Onuf, Nicholas G. 1989. World of Our Making. Columbia: University of South Carolina Press.

Patton, Michael Q. 1990. Qualitative Evaluation and Research Methods. London: Sage.

Pentland, Charles. 1973. International Theory and European Integration. London: Faber and Faber, Ltd.

Richmond, Anthony H. 1987. "Ethnic Nationalism: Social Science Paradigm." International Social Science Journal 39:3-18.

Robyn, Richard. 2000. Forms of Attachment to the European Union: A Study of French Conceptions of Identity in the Context of a Changing Europe. Dissertation. Kent State University.

Rosaldo, Renato. 1994. “Subjectivity in Social Analysis.” In The Postmodern Turn: New Perspectives on Social Theory, ed. S. Seidman. Cambridge: Cambridge University Press.

Skinner, B.F. 1953. Science and Human Behavior. New York: Macmillan.

Smith, Anthony D. 1983. Theories of Nationalism, 2nd ed. New York: Holmes & Meier.

Stainton Rogers, Wendy. 1991. Explaining Health and Illness: An Exploration of Diversity. London: Harvester Wheatsheaf.

Stephenson, William. 1935. “Technique of Factor Analysis,” Nature 136:297.

Stephenson, William. 1953. The Study of Behavior: Q-technique and Its Methodology. Chicago: University of Chicago Press.

Stephenson, William. 1967. The Play Theory of Mass Communication. Chicago: University of Chicago Press.

Stephenson, William. 1978. "Concourse Theory of Communication." Communication 3:21-40.

Stephenson, William. 1983. “Against Interpretation.” Operant Subjectivity 6: 73-103.

Stephenson, William. 1987. “Q Methodology: Interbehavioral and Quantum Theoretical Connections in Clinical Psychology.” In New Ideas in Therapy: Introduction to an Interdisciplinary Approach, ed. Douglas H. Ruben and Dennis J. Delprato. Westport, CT: Greenwood Press.

Stone, James C. 1969. Teachers for the Disadvantaged. San Francisco: Jossey-Bass.

Stryker, Sheldon. 1980. Symbolic Interactionism: A Social Structural Version. Palo Alto, CA: B. Cummings.

Tajfel, Henri, and John C. Turner. 1971. "An Integrative Theory of Intergroup Conflict." In The Social Psychology of Intergroup Relations, eds. William G. Austen and Stephen Worchel. Pacific Grove, CA: Brooks/Cole.

Terhune, Kenneth. 1964. “Nationalism Among Foreign and American Students: An Exploratory Study.” Journal of Conflict Resolution 8:256-270.

Wicklund, Robert A. and Annerieke Oosterwegel. 1995. The Self in European and North American Culture: Development and Processes. Amsterdam: Kluwer Academic Publishing.

Wolf, M.M. 1978. “Social Validity: The Case for Subjective Measurement, or How Applied Behavior Analysis is Finding its Heart.” Journal of Applied Behavior Analysis 11:203-214.

Wong, Timothy and Milan Sung. 1998. “Dissolution and Reconstruction of National Identity: The Experience of Subjectivity in Taiwan.” Nations and Nationalism 4:247-272.

Worchel, Stephen. 1997. "The Tangled Web of Loyalty: Nationalism, Patriotism, and Ethnocentrism." In In the Lives of Individuals and Nations, eds. Daniel Bar-Tal and Ervin Staub. Chicago: Nelson-Hall.

[1]

The parenthetical notations [(N), (SN) and (PL)] after each statement incorporate the Q-sample structure discussed in this chapter; i.e., Nationalist, Supranationalist and Pluralist.