¹ Twelve-year-old Colchester girl interviewed in 1968.

² Shils, E. and Young, M., ‘The Meaning of the Coronation’, The Sociological Review, I (1953), 63–81, p. 63.CrossRefGoogle Scholar See also Birnbaum, N., ‘Monarchs and Sociologists: A Reply to Professor Shils and Mr. Young’, The Sociological Review, III (1955), 5–23.CrossRefGoogle Scholar

³ On that topic see Hardie, F., The Political Influence of the British Monarchy: 1868–1952 (New York: Harper and Row, 1970);Google ScholarJohnson, P., ‘The Political Power of the Monarchy’, in Brown, J. Murray, ed., The Monarchy and Its Future (London: George Allen and Unwin, 1969), pp. 111–22Google Scholar; and Martin, K., The Crown and the Establishment rev. ed.,(Harmondsworth: Pelican, 1965).Google ScholarOlder discussions are Macdonagh, M., The English King (London: Ernest Benn, 1929)Google Scholar and Fairer, J. A., The Monarchy in Politics (New York: Dodd Mead, 1917).Google Scholar

⁴ Presumably the following statement from Britain 1973: An Official Handbook (London: HMSO, 1973), p. 29Google Scholar, can be taken as the nearest approximation to the official view (or at least the publicized official view) of what it may be that presently hedges royalty in Britain: “The Queen is the personification of the State. In law, she is the head of the executive, an integral part of the legislature, the head of the judiciary in England and Wales, Northern Ireland and Scotland, the commander-in-chief of all the armed forces of the Crown and the temporal head of the established Church of England. In practice, as a result of a long evolutionary process during which the absolute power of the monarch has been progressively reduced, the Queen acts only on the advice of her ministers, which she cannot constitutionally ignore. She reigns, but she does not rule. The United Kingdom is governed by Her Majesty's Government in the name of the Queen’. Also see the various essays and prefatory writings in Burke's, Guide to the Royal Family (London: Burke's Peerage, 1973).Google Scholar Attempts at empirical delineation of aspects of the symbolic meaning of the monarchy include Abramson, P. R. and Inglehart, R., ‘The Development of Systemic Support in Four Western Democracies’, Comparative Political Studies, II 1970, 419–42Google Scholar, and Blumler, J. G., Brown, J. R., Ewbank, A. J. and Nossiter, T. J., ‘Attitudes to the Monarchy: Their Structure and Development During a Ceremonial Occasion’, Political Studies, XIX (1971), 149–71.CrossRefGoogle Scholar Examples of speculative discussions are Adam, T., ‘The Queen as Political Symbol in the British Commonwealth’, in Bryson, L., Finkelstein, L., Hoagland, H., Maclver, R. M., eds., Symbols and Society (New York: Cooper Square Publishers, 1964Google Scholar; originally published 1954), pp. 9–22, and Jones, E., ‘The Psychology of Constitutional Monarchy’, in his Essays in Applied Psychoanalysis (New York: International Universities Press, 1964), pp. 227–33Google Scholar (essay originally published in The New Statesman and Nation, February 1,1936). Commonwealth children's perceptions of the Queen are reported in Pammett, T. H., ‘The Development of Political Orientations in Canadian School Children’, Canadian Journal of Political Science, IV (1971), 132–41CrossRefGoogle Scholar, and Connell, R. W., The Child's Construction of Politics (Melbourne: Melbourne University Press, 1961).Google Scholar On Japanese children's perceptions of their monarch and prime minister see J. A. Massey, ‘The Missing Leader: Japanese Youth's View of Political Authority’, American Political Science Review, forthcoming.

⁵ For example, 80 per cent of a sample of Campaign for Nuclear Disarmament activists who responded to Parkin's mail questionnaire ticked ‘disagree’ or ‘strongly disagree’ in response to the statement ‘The monarchy is an institution we should be justly proud of and only 12 per cent indicated ‘agree’ or ‘agree strongly’. F. Parkin, , Middle Class Radicalism: The Social Bases of the British Campaign for Nuclear Disarmament (Manchester: Manchester University Press, 1968), p. 26Google Scholar (percentage recomputed from the subdivided data in his table). A survey conducted in 1971 for The Times of men and women listed in Who's Who found many fewer (15 per cent) of this elite group of respondents describing the monarchy as ‘very influential’ in ‘shaping society in Britain today’ than Parliament (42 per cent) and the Civil Service (23 per cent). Only the Church ranked lower (2 per cent). Thomson Group Marketing Services, Britain Today (London: Times Newspapers Ltd., 1971), p. 8.Google Scholar

⁶ The general-population respondents interviewed by Blumler et al., ‘Attitudes to the Monarchy’, for example, exhibited high levels of support for the monarchy; indeed, positive attitudes toward the monarchy in this study were at about the level of the negative attitudes of Parkin's CND respondents and, if anything, these positive attitudes improved somewhat in the immediate aftermath of the investiture of Prince Charles. Many surveys reported in the press have testified to the popularity of the monarchy with the general public: for example, the following spate of polls during the period of the investiture and the Royal Family film: ‘The Queen Tops Royal Popularity Poll’, a Sunday Times Poll by Opinion Research Centre, Sunday Times, 23 March 1969, in which only 6 per cent of the sample (with no significant variation by sex or party) agreed that ‘the monarchy should be abolished’; Daily Mirror, 17 November 1969, in which Prince Philip received first choice for a hypothetical presidency of Britain if the nation became a republic ‘and we had to elect our first president’; Daily Mirror, 18 November 1969, reporting widespread approval of the level of expenditures of the Royal Family and related royal practices such as not sending the Queen's children to State schools; Whale, John, ‘Why Britain is Royalist’, STP Poll, Sunday Times, 29 06 1969Google Scholar, in which two-thirds of the respondents approved of the present ‘power to influence matters of importance in Britain’ and about three times more of the remainder wished the monarchy had more rather than less influence than at present (the influence attributed to the monarchy by more than half of the respondents included ‘keeping the Commonwealth united’, ‘making it less likely that Britain will have a violent revolution’, and setting ‘standards of morality and family behaviour'); and the Gallup Poll reported in ‘Is Britain Still Really Democratic?’, Daily Telegraph Magazine, Number 259, 26 September 1969, 23–5, with more evidence of Prince Philip's popularity. For a less fugitive source in which data are presented more systematically (with similar conclusions), see the periodic surveys reported by National Opinion Polls Limited — for example, NOP Political Bulletin, July 1969, October 1969 (a particularly full survey), June 1971, and June 1972.

⁷ Bagehot, , The English Constitution (London: Collins, 1963; first published in 1867; definitive text, 1915), the quotations in the text at pp. 82, 86, 90, 97, and 249Google Scholar, respectively. L. J. Sharpe reminds us ‘that the English Constitution, admirable as it is, was written over a hundred years ago’. His plaintive call for ‘a moratorium of, say, ten years… for quotations from it that seek to describe British politics today’ ought to be extended to a permanent injunction since Bagehot demonstrates nothing about the present and perhaps little about his own time. But Bagehot as a source of hypotheses, whether about the past or the present, is not too often drawn upon but rather too rarely read closely. Sharpe's, comment is in his ‘American Democracy Reconsidered: Part I’, British Journal of Political Science, III (1973), 1–28, p. 17.CrossRefGoogle Scholar

⁸ See fn. *. We conclude that our finding was at once recognized as accurate and as credible but not ‘already obvious’ by the considerable attention to that article in the press and other media.

⁹ On children's partisanship and their evaluations of the Prime Minister see Stradling, Robert and Zureik, Elia, ‘Political and Non-political Ideals of English Primary and Secondary School Children’, The Sociological Review, XIX (1971), 203–27.CrossRefGoogle Scholar

¹⁰ For an important critique of the widespread use of ‘deference’ as a concept for explaining English political socialization see Kavanagh, Dennis, ‘The Deferential English: A Comparative Critique’, Government and Opposition, VI (1971), 333–60.CrossRefGoogle Scholar For an example of deference theorizing which nevertheless remarks on the high degree of non-deference in English industrial relations see Goldthorpe, J. H., ‘Social Inequality and Social Integration in Modern Britain’, Advancement of Science, XXVI (1969), 190–202.Google Scholar This article deliberately skirts the theoretical thicket of clarifying the notion of deference and the various alternative ways of explaining the behaviour attributed to that putative orientation. Two basic writings that need to be consulted by trail blazers are Parkin, Frank, ‘Working-class Conservatives: A Theory of Political Defiance’, British Journal of Sociology, XVIII (1967), 278–90CrossRefGoogle Scholar; and Goffman, Erving, ‘The Nature of Deference and Demeanor’, American Anthropologist, LVIII (1966), 473–502.Google Scholar

¹¹ Needless to say, it remains to be demonstrated that there is a ‘cooler, more detached view’ of the British Prime Minister (or prime ministers in constitutional monarchies in general, or even prime ministers in general) than of leaders who combine the functions that are shared in Britain by the Queen and the Prime Minister. One immediately sees the complexities that would be involved in operationalizing the key terms and devising appropriate measures, as well as the impossibility of administering those measures retroactively and the considerable problems in distinguishing between role and incumbent in studying popular orientations to a political office. One convenient source of insight is the survey item regularly asked over the years by the Gallup affiliates in many nations on how good a job the incumbent chief executive is doing. Using statistics provided us by British Gallup and the various publications of the American Institute of Public Opinion, we have made a variety of comparisons of the run of presidential and primeministerial popularity statistics since 1945. Taking 1945–1969 as a period of comparison, for example, we note little difference in the overall average of citizens in each nation approving of the way the leader carried out his duties: 52 per cent in Britain and 57 per cent in the United States. The low end of the range is the same in both countries — an average of 28 per cent approving Truman in 1951 and an average of 26 per cent approving Wilson in 1969. However, the US range appears capable of going much higher: 81 percent for Truman's two readings in 1945, and 74 per cent for Kennedy in 1961 and Johnson in 1964, to 68 per cent for Macmillan in 1960. Another measure of whether much feeling is invested in the role might be the capacity for sharp oscillation in support in relatively short periods — for example, the 40 per cent variation in Johnson's popularity within the year 1966 and Nixon's drop of about 40 percentage points during the Watergate-disclosure-filled year of 1973.Whether the Americans and the English convey quite the same level of feeling by ‘approve’ and ‘disapprove’ is itself an empirical question of more than a little interest, calling for greater research ingenuity than has been common in cross-cultural research.

¹² It should therefore be remembered that we were administering questionnaires during a period when there was a Labour Prime Minister who was faring poorly in public esteem and that the research was not conducted in an area of Labour strength.

¹³ A very few of the children who said that the Queen is more important were not innocent believers in a fairy-tale conception of British politics, however, but rather Tory partisans expressing negative views of Harold Wilson. This precocious Toryism strengthens Zureik's findings of greater middle-class political realism since partisan, as opposed to unrealistic, references to the Queen's importance are largely by middle-class children, and if these were eliminated the class differences in political realism would be sharper. In addition, some children say the Queen is more important out of positive attraction to royal display and symbolism rather than out of political realism. Our open-ended data discussed below and tabulated in Table I distinguish among the different meanings behind common fixed-choice responses.

¹⁴ This is not to imply that screening devices such as the eleven-plus examination identify ‘intrinsic merit’ or that they are free of class bias, but only to suggest that the mixture of children so selected are likely to display more political realism than children in the sorts of public schools described in the text.

¹⁵ The procedures, rationale, and findings of this study are more fully described in Greenstein, F. I. and Tarrow, S., ‘Political Orientations of Children: The Use of a Semi-Projective Technique in Three Nations’, Sage Professional Papers in Comparative Politics, I, Series No. 01–009 (1970), 479–560Google Scholar, and F. I. Greenstein, ‘Children's Images of Political Leaders in Three Democracies: The Benevolent Leader Revisited’, paper delivered at the 1973 Annual Meeting of the American Political Science Association, forthcoming ('The Benevolent Leader Revisited: Children's Images of Political Leaders in Three Democracies') in the American Political Science Review. The open-ended technique was used to escape at least some of the difficulties pointed to by various of the critics of Dennis, J., Lindberg, L., and Mccrone, D., ‘Support for Nation and Government among English Children’, British Journal of Political Science, I (1971), 25–48.Google Scholar Rejoinders to this paper include Marsh, D., ‘Political Socialization: The Implicit Assumptions Questioned’, British Journal of Political Science, I (1971), 453–66CrossRefGoogle Scholar, and Birch, A. H., ‘Children's Attitudes and British Politics’, British Journal of Political Science, I (1971), 519–20CrossRefGoogle Scholar; Budge, I., ‘Support for Nation and Government among English Children: A Comment’, British Journal of Political Science, I (1971), 389–92CrossRefGoogle Scholar; Marsh, D., ‘Beliefs about Democracy among English Adolescents: What Significance have They?’, British Journal of Political Science, II (1972), 255–9.CrossRefGoogle Scholar

¹⁶ Greenstein, F. I., ‘The Benevolent Leader: Children's Images of Political Authority’, American Political Science Review, LIV (1960), 934–43CrossRefGoogle Scholar, later expanded on in Greenstein, F. I., Children and Politics (New Haven: Yale University Press, 1965), Chap. 3Google Scholar; Hess, R. D. and Easton, D., ‘The Child's Changing Image of the President’, Public Opinion Quarterly, XXIV (1960), 632–44.CrossRefGoogle Scholar For useful anthologies of writings from the political socialization literature, see Adler, N. and Harrington, C., The Learning of Political Behavior (Glenview, III.: Scott, Foresman, 1970)Google Scholar; Sigel, R. S., Learning about Politics: A Reader, in Political Socialization (New York: Random House, 1970)Google Scholar; and Dennis, J., Socialization to Politics (New York: Wiley, 1973).Google Scholar The first and third of these anthologies have excellent bibliographies appended.

¹⁷ Connell, , Child's Construction of Politics, p. 28.Google Scholar

¹⁸ On the absence of civic instruction in standard British curricula and the current proposals to change this state of affairs, see, for example, Heater, D. B., ‘Contemporary History Justified’, Teaching History, I (1969), pp. 2–7Google Scholar, and the various issues of the periodical Teaching Politics, which was founded in 1972. We are indebted to Kavanagh, Dennis for the point about children's difficulty in uncoding the official language of British government. On class and the learning of linguistic codes, see the work of Basil Bernstein — for example, his Class, Codes and Control (London: Routledge and Kegan Paul, 1971).Google Scholar

¹⁹ ‘Rich’ relative to our comparison groups. The following image scores for the French President of the Republic and Premier and the American President were computed in the same fashion as the scores for the Queen and the Prime Minister in Table 2: France: President of the Republic, 2.7; Premier, 1.5; United States, white subsample: President, 3.1; black subsample: President, 2.7. As can be seen both of the prime-ministerial roles are more dimly perceived than any of the head-of-state roles, but the British Prime Minister occupies more pre-adult cognitive space than the French Premier. (The latter is largely viewed simply as a stand-in for the President of the Republic in the event of that individual's absence from the country and as a presidential assistant.)

²⁰ Lindzey, G. and Borgatta, E. F., ‘Sociometric Measurement’, in Lindzey, G., ed., The Handbook of Social Psychology, I (Cambridge, Mass.: Addison-Wesley, 1954), 405–48.Google Scholar Another set of social psychological boxes relevant to our data is Robert Bales’ well-known distinction between social-emotional leadership and task leadership. This is very much the way many respondents seem to perceive the differences between the Queen and the Prime Minister. See Slater, P. E., ‘Role Differentiation in Small Groups’, American Sociological Review, XX (1955), 300–10CrossRefGoogle Scholar, and the sources there cited.

²¹ See fn. 16.

²² See fn. 17.

²³ Table 3 also presents as a source for readers’ speculations the class differences in our subdivided positive affect categories, but here the small size of the categories combined with the smallness of the subsamples discourage us from attempting interpretations.

²⁴ On the French sample and the two US samples, see the second reference in fn. 15. The French respondents were interviewed in the late de Gaulle and early Pompidou periods and are from the south of France and the Paris area. The Americans are from the eastern United States and were interviewed in 1969–1970 during Mr. Nixon's ‘low profile’ period of his first term in office and well before the Watergate episode early in his second term. In each case the populations of respondents interviewed were selected on a quota basis with even divisions of boys and girls, and roughly even division by class, except among the predominantly low socio-economic-status American blacks. A variety of controls have been introduced to insure against reporting ‘national’ differences that are merely artefacts of sample composition, but it should be evident that such small-sample findings can be suggestive, but not conclusive.

²⁵ On the reasons for using different types of measurement, not merely different measures, see the following classic paper: Campbell, D. T. and Fiske, D. W., ‘Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix’, Psychological Bulletin, LVI (1959), 81–105.CrossRefGoogle Scholar

²⁶ See Kavanagh's excellent review on this point and on the problems of measurement of ‘deference’ (fn. 10.)

²⁷ In seeking to satisfy ourselves as to whether we ‘really’ are measuring deference, we have listed for ourselves every response coded as deferential and devoted a good bit of thought to whether we had appropriately put various entries on the list. In the final analysis we cannot prove ‘deference’, but we can share with others the specific content of what the children we interviewed said. In addition to the quotations already given, here are further examples: The policeman recognized the Queen and apologized very much for being rude. The Queen shouldn't be driving anyway; she should have a chauffeur to drive for her. So, really,

²⁸ For a general discussion see Greenstein, , Children and Politics, 75–84.Google Scholar This formulation about why early political learning may be consequential has been criticized by Searing, Donald D., Schwartz, Joel J., and Lind, Alden E. in ‘The Structuring Principle: Political Socialization and Belief Systems’, American Political Science Review, LXVII (1973), 415–32.CrossRefGoogle Scholar For a critique of their assumptions, and an effort to discuss circumstances under which early learning is likely to be consequential, see Greenstein's, Letter to the Editor American Political Science Review, LXVIII (1974), 720–2.CrossRefGoogle Scholar

²⁹ The Sunday Times of 15 July 1973 reported a poll in which a national sample of Britons was given a list of twelve institutions and asked to indicate the two that ‘… have the greatest effect on how the country is run’. Trade unions received the most references (44 per cent) and the Prime Minister the second most (33 per cent), with the Queen garnering a mere 5 per cent. Dennis Kavanagh informs us of unpublished research showing that adult attitudes toward the monarchy have very little independent predictive power for explaining other attitudes held by the same individuals. If Britons are asked how important the Queen is to Britain rather than in running Britain, only a few (13 per cent) deprecate her importance, at least circa the 1964 study of Butler, David and Stokes, Donald, Political Change in Britain (New York: St Martin's, 1964), p. 484.Google Scholar

³⁰ ‘Prince of Wales’, The Economist, 5 July 1969, 12–14.

³¹ Barry, Brian M., Sociologists, Economists and Democracy (London: Collier-Macmillan, 1970), pp. 73–4Google Scholar, appropriately criticizes the view that monarchy causes political stability. But it seems premature to conclude as Barry does that the causal sequence is wholly in the opposite direction: stability makes the Crown possible. At least for Britain, one suspects that there are more complex feedback influences. Barry, like The Economist authors, is prepared to reason cross-nationally on the basis of the logic of variable-by-variable correlations. Compare the attack on this ubiquitous assumption by the anthropologist Clifford Geertz. Cross-cultural variation is not ‘… experimental variation,’ Geertz, writes, ‘because the context in which it occurs varies with it, and it is not possible… to isolate the y's from the x's to write a proper function.’ The lnterpretation of Cultures (New York: Basic Books, 1973), p. 23.Google Scholar