1 For Philip Converse, government is an object of low centrality (“Attitudes and Non-Attitudes: Continuation of a Dialogue,'” revised version of a paper read at the Seventeenth International Congress of Psychology in Washington, DC, August 1963. Reprinted in Tufte, Edward R. (ed.), The Quantitative Analysis of Social Problems [Reading: Addison-Wesley, 1970], 168–69)Google Scholar. According to Robert Dahl, “in New Haven as in the United States generally one of the central facts of political life is that politics—local, state, national, international—lies for most people at the outer periphery of attention, interest, concern, and activity” (Who Governs? Democracy and Power in an American City [New Haven and London: Yale University Press, 1961], 279Google Scholar). Similarly, Butler and Stokes suggest that for the ordinary British subject, politics is very remote (Butler, David and Stokes, Donald, Political Change in Britain: Forces Shaping Electoral Choice [New York: St. Martin's Press, 1969], 25Google Scholar). The French, according to Alain Lancelot, have little “interest” in politics (L'Abstentionnisme électorate en France [Paris: Armand Colin, Cahiers de la Fondation Nationale des Sciences Politiques, 1968], 164–67Google Scholar). Ralf Dahrendorf analyzes the ordinary German who is “unpolitical because the political is deeply unimportant to him” (Society and Democracy in Germany [Garden City: Doubleday, 1969], 377Google Scholar).

2 The relevant literature is enormous. For insights into the scope of this area of inquiry, see Robinson, John P., Rusk, Jerrold G., and Head, Kendra B., Measures of Political Attitudes (Ann Arbor: Center for Political Studies, Institute for Social Research, University of Michigan, 1968)Google Scholar.

3 The term “centrality” is sometimes employed to identify a particular attitude which is pivotal to a general belief structure. For example, according to Harry C. Triandis, “centrality … implies that the change of a belief high in this characteristic also will change many other beliefs” (Attitude and Attitude Change [New York: John Wiley, 1971], 9Google Scholar). This notion is distinct from the idea of “subjective importance” and, consequently, from the use of the term “centrality” in this paper.

4 Newcomb, Theodore M., Turner, Ralph H., and Converse, Philip E., Social Psychology: The Study of Human Interaction (New York: Holt, Rinehart and Winston, 1965), 58–59Google Scholar.

5 Dahl, Who Governs?, 279.

6 Ibid.., 279–80.

7 Ibid.., 178.

8 Nie, Norman H. with Andersen, Kristi, “Mass Belief Systems Revisited: Political Change and Attitude Structure,” Journal of Politics 36 (1974), 572CrossRefGoogle Scholar.

9 Newcomb, Turner and Converse, Social Psychology, 59.

10 Stouffer, Samuel A., Communism, Conformity and Civil Liberties: A Cross-section of the Nation Speaks Its Mind (Gloucester: Peter Smith, 1963Google Scholar; originally published in 1955).

11 Robert Dubin has carried out extensive empirical research on a phenomenon related to centrality. He has been studying the “central life interests” of workers, that is, their “expressed preference for a given locale or situation in carrying out an activity” (Dubin, Robert, “Industrial Workers' Worlds: A Study of the ‘Central Life Interests’ of Industrial Workers,” Social Problems 3 [1956], 134CrossRefGoogle Scholar). In this research, there are no data on the centrality of politics. See also: Dubin, Robert and Champoux, Joseph E., “Workers' Central Life Interests and Personality Characteristics,” Journal of Vocational Behavior 6 (1975), 165–74CrossRefGoogle Scholar; Dubin, Robert, Champoux, Joseph E., and Porter, Lyman, “Central Life Interests and Organizational Commitment of Blue-Collar and Clerical Workers,” Administrative Science Quarterly 20 (1965), 411–21CrossRefGoogle Scholar; and Dubin, Robert and Goldman, Daniel R., “Central Life Interests of American Middle Managers and Specialists,” Journal of Vocational Behavior 2 (1972), 133–41CrossRefGoogle Scholar.

12 Stouffer, Communism, Conformity, and Civil Liberties, 15–17.

13 Ibid., 59.

14 Ibid., 67–68.

15 Ibid., 70. Emphasis is mine.

16 Ibid., 70.

17 Ibid., 75–76.

18 These limitations stem from the attempt to employ his findings to illuminate the problem of governmental centrality, a topic toward which Stouffer's research was not specifically directed.

19 This research logic is based on arguments presented by Karl Popper and Harry Eckstein. See Popper, Karl R., The Logic of Scientific Discovery (New York: Harper Torchbooks, 1965)Google Scholar; and Eckstein, Harry, “Case Study and Theory in Political Science,” in Greenstein, Fred I. and Polsby, Nelson W. (eds.). Handbook of Political Science Vol. 7 (Reading: Addison-Wesley, 1975), 79–137Google Scholar.

20 “Centrality” and “subjective importance” are, by their very nature, relative concepts. Things are not important or unimportant, central or noncentral, to people; rather they are more or less important, or central, than other things.

21 This study of governmental centrality was part of a larger inquiry into the nature of behaviour toward authority by pupils in secondary schools. In particular, I have been concerned with student-teacher interactions. A study of this phenomenon was conducted in 1967–1968. The 1974 study sought primarily to investigate changes which have occurred since the May crisis. Research included not only pupil questionnaires but also classroom observation and teacher interviews. For further details on the structure of French education and the results of the first study, see: Schonfeld, William R., Youth and Authority in France: A Study of Secondary Schools (Beverly Hills: Sage Professional Papers in Comparative Politics, 1971)Google Scholar; and especially Obedience and Revolt: French Behavior Toward Authority (Beverly Hills: Sage Publications, 1976), 18–68Google Scholar, 205–27.

22 This item was placed at the very beginning so that responses to it would not be polluted by the other items included in the questionnaire, which focussed attention on student-teacher authority relations.

23 The most frequently-mentioned items combined under the general rubric of family are: “the family,” “family relationships,” and “parents.” “My friends,” “buddies,” “comrades,” “relations with boys and girls,” and “friendship” are the most often cited items in the peer group category. The rubric sex typically captures such references as “love,” “relations between boys and girls,” and “relations with girls” (stated by a male). The most commonly cited items categorized as referring to the school are “homework,” “school,” “the school environment,” “my friends in class,” “my education,” and specific subject matters. Individual recreational activities include, in particular, “listening to music,” “films,” and a wide range of sports and hobbies which do not require other people—such as boating, golf, painting, photography and hiking. Collective recreational activities basically include sports that cannot be played alone. Items judged as referring to politics include naming a politician (usually Giscard or Mitterrand), “the right to vote,” “elections,” “political events,” “political life,” “being an activist,” and, most common of all, simply the word “politics.” The category societal issues identifies problems or situations which, from the perspective of the scholar, may depend on governmental action. However, the respondent does not present the issue in a manner which suggests an awareness of governmental involvement. The most commonly cited items coded under this rubric are: “ecology,” “to be able to express oneself freely,” “freedom to do what one wants with one's own life,” and “social justice.”

24 Space was also provided for respondents to identify two additional relationships and state the amount of personal importance and thinking they gave to each. Although a minority of students did identify additional relationships, the heterogeneity of their responses militates against meaningful analysis.

25 For example, David Easton defines political science “as the study of the authoritative allocation of values for a society” (The Political System [New York: Alfred A. Knopf, 1953], 129Google Scholar). Similarly, Lasswell and Kaplan conceive of it as the study of “influence and power as instruments of … the integration of values realized by and embodied in interpersonal relations” (Power and Society: A Framework for Political Inquiry [New Haven: Yale University Press, 1950], xiiGoogle Scholar). Robert Dahl suggests that a “political system is any persistent pattern of human relationships that involves, to a significant extent, power, rule, or authority” (Modern Political Analysis [Englewood Cliffs: Prentice-Hall, 1963], 6Google Scholar). To cite a final example, Harry Eckstein equates political science with the study of authority patterns (“Authority Patterns: A Structural Basis for Political Inquiry,” American Political Science Review 67 [1973], 1142–61CrossRefGoogle Scholar).

26 The proportion of the total sample attributing intense subjective importance to each of the other attitude domains is as follows: individual recreational activities, 29.5 per cent(N = 142); school, 20.8 percent(N = 100); collective recreational activities, 14.6 per cent (N = 70); peer group, 11.2 per cent (N = 54); family, 4.6 percent (N = 22); societal issues, 4.4 per cent (N = 21); and sex, 3.5 per cent (N = 17).

27 Gammas rather than Pearson correlations were used for two reasons: (1) responses to the importance and thinking about questions do not follow a normal distribution; and (2) it is not obvious to me that the scalar values are genuinely interval—for example, is the distance between “never” and “rarely” equal to the distance between “rarely” and “from time to time?”

28 The gammas relating the importance and thinking-about questions for each predefined relationship are: parents, .55; members of the opposite sex, .87; friends, .85; brothers/sisters, .70; teachers, .70; government, .76; neighbours, .82; shopkeepers, .75; and police, .77.

29 These categories were developed by combining the responses to the importance questions (“fundamental importance” was coded as I; “a great deal of importance” as 2; “a little importance” as 3; and “no importance” as 4) with the responses to the thinking-about questions (“almost all the time” was coded as 1; “very often” as 2; “from time to time” as 3; “rarely” as 4; and “never” as 5). The category labels correspond to the following cumulative scores: great centrality, 2 or 3; some centrality, 4 or 5; little centrality, 6 or 7; and no centrality, 8 or 9.

30 This image of relative centrality, obtained by using the combined measure, is the same as the results derived simply from either the importance or the thinking-about question.

31 The importance attributed by the French to privacy, and its separation from the public domain, are discussed in the literature on national character and in community studies. See, for example, Metraux, Rhoda and Mead, Margaret, Themes in French Culture: Preface to a Study of French Community (Stanford: Stanford University Press, Hoover Institute Studies, 1954)Google Scholar; and Wylie, Laurence, Village in the Vaucluse (Cambridge: Harvard University Press, 1957)Google Scholar.

32 The relationship between centrality and precision of partisan preference remains if respondents who did not identify any political preference are included in the calculations: 49.4 percent of those with high political centrality, 66.1 per cent of those with above-average centrality, and 79.7 per cent of those with no centrality, fail to identify a precise partisan preference.

33 The general relationship between religion and political attitudes in France is analyzed by Michelat, Guy and Simon, Michel in Classe, religion et comportement politique (Paris: Presses de la Fondation Nationale des Sciences Politiques and Editions Sociales, 1977)Google Scholar. For a possible explanation, in cross-cultural terms, of the relationship between religion and radical politics, see Glock, Charles Y. and Stark, Rodney, Religion and Society in Tension (Chicago: Rand-McNally, 1965)Google Scholar, chap. 11.

34 Some readers may bemoan the absence of an attempt to use the findings presented in this article to cast light on French politics. However, to do so would dilute the thrust of the argument which is centred not on a particular polity but on a specific concept and its implications.

35 Stouffer, Communism, Conformity and Civil Liberties, 70.

36 The Pattern of Human Concerns (New Brunswick: Rutgers University Press, 1965), 401–08Google Scholar.

37 I have always been struck by that paradoxical label of “behavioural political science,” a mode of inquiry which focusses on the study of attitudes.

38 The time differential may be associated not only with the extent of activity, but also the type of activity. I think this is a reasonable supposition, and one, I have argued elsewhere, which could resolve the debate over the meaning of democratic participation. See the author's “The Meaning of Democratic Participation,” World Politics 28 (1975), 134–58CrossRefGoogle Scholar.

39 Rosenberg, Milton J. and Hovland, Carl I., “Cognitive, Affective, and Behavioral Components of Attitudes,” in Rosenberg, Milton J. et al. (eds.), Attitude Organization and Change: An Analysis of Consistency Among Attitude Components (New Haven: Yale University Press, 1960), 1Google Scholar.

40 Converse, “Attitudes and Non-Attitudes,” 175.

41 See, for example: Bennett, W. Lance, The Political Mind and the Political Environment (Lexington: Lexington Books, D. C. Heath, 1975)Google Scholar, chap. 1; Converse, Philip E., “Comment: The Status of Non-attitudes,” American Political Science Review 68 (1974), 650–60Google Scholar; Pierce, John C. and Rose, Douglas D., “Nonattitudes and American Public Opinion: The Examination of a Thesis,” American Political Science Review 68 (1974), 626–49CrossRefGoogle Scholar; and Vaillancourt, Pauline Marie, “Stability of Children's Survey Responses,” Public Opinion Quarterly 37 (1973), 373–87CrossRefGoogle Scholar.

42 The classic reference on validity is Campbell, Donald T. and Fiske, Donald W., “Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix,” Psychological Bulletin 56 (1959), 81–105CrossRefGoogle ScholarPubMed.

43 In fact, Pauline Vaillancourt, one of the participants in the nonattitudes debate, uses the two terms “interchangeably” (“Stability of Children's Survey Responses,” 373).

44 Deutscher, Irwin, What We Say/What We Do: Sentiments and Acts (Glenview: Scott, Foresman, 1973), 106Google Scholar.

45 Etzioni, Amatai, “Communication: Mathematics for Sociologists?” American Sociological Review 30 (1965), 943Google Scholar.

46 For an overview on the subject of attitude change, see Newcomb, Turner, and Converse, Social Psychology, 80ff.

47 There have been some limited attempts to compare self-reports of voting with actual voting as determined by official records. For example, in the United States, Miller found that among the 113 respondents interviewed within three weeks before the election who claimed they were “certain to vote,” 86 (74%) actually did; and among the 132 respondents who during an interview within four weeks after the election claimed to have voted, 110 (83%) actually did (Miller, Mungo, “The Waukegan Study of Voter Tumout Prediction,” Public Opinion Quarterly 16 [1952]CrossRefGoogle Scholar, reprinted in Katz, Daniel et al. [eds.], Public Opinion and Propaganda [New York: Holt, Rinehart and Winston, 1954], 751–60)Google Scholar. The findings I have cited were calculated from Table 1, 752. Unfortunately, studies of this sort are extremely rare. However, in France, numerous scholars—sceptical of the accuracy of self-reported voting behaviour—have carefully examined official records in a variety of electoral districts to analyze the constancy of voting behaviour and the characteristics of those who abstain. This research is synthesized in Lancelot, L'Abstenlionnisme électoral en France, chap. 5.

48 “Attitudes and Non-Attitudes.” In spite of the significant attention researchers have directed at the notion of “non-attitudes,” no one has carefully examined the implications of centrality for the expression of political attitudes, even though this idea is explored in Converse's seminal article. In anotherpaper, I have examined the implications of governmental centrality for the study of political socialization. See this author's “The Focus of Political Socialization Research: An Evaluation,” World Politics 23 (1971), 569–78Google Scholar.

49 Newcomb, Turner and Converse, Social Psychology, 61.