Relating the notion of political community to those of discrimination and legitimacy — and, consequently, setting up a typology which will contrast communities by the criteria they use to distinguish members from non-members and by the location they give to legitimacy, placed either in a ruler or in a process — will be made easier if we consider first of all certain striking similarities which exist between the most private of social groups, the family, and the most public, the state, similarities which probably facilitate the transfer of the ideological constructs formed in infancy and childhood into the political expectancies and assumptions of adulthood. Let us distinguish communities defined through the brothers from communities defined through the father: communities centered around a leader from those centered on themselves. These contrasts will suggest to us a natural/ artificial continuum along which political communities can be ranked, those most resembling the family in their ideas about legitimate authority and legitimate membership being closer to the ‘natural’ end of the continuum. I will explain later the reasons for this distinction between ‘natural’ and ‘artificial’, but I must at the outset make it absolutely clear that the use of these terms does not in any way imply an evaluative preference on my part.

Administrative centralization has been a fact of life in France since Louis XIV. Political analysts have alternately praised the system as ‘the envy of all Europe’ and condemned it as ‘a glittering example of “blockage” in our society’. Very few of them have, however, made any serious effort to study the impact of centralized structures on local governments. The literature is remarkable for the absence of empirical research on the exercise of power by elected local officials in France.

Leaders and activists in most Western political parties, appreciating Schattschneider's adage1 that parties are controlled by those who control nominations to public offices, jealously guard their right to choose parliamentary candidates. They realize that candidates play a crucial role in shaping voters’ images of the parties as well as in formulating the party policies. Not surprisingly, candidates are usually recruited through rather oligarchical procedures. Extra-parliamentary party leaders normally pick the nominees in closed private meetings and then submit their choices to rank-and-file party members and/or delegates who nearly always grant their approval. Such procedures reduce the role of party members to one of docile acquiescence, and completely exclude party voters.

In the study of power structures, whether of the local community or the national society, three main approaches have been proposed: the reputational, the sociology of leadership and the issue approaches. Of these the least popular in practice is the last-named, which would attempt to discover ‘who rules?’ through an analysis of actual decision-making in a series of issues. This is particularly true of studies of national power, where evidence on how high-level political decisions were made is cited only to illustrate or exemplify an argument.² Although some studies of local community power have relied primarily on the issue method, studies of national power have almost invariably utilized the sociology of leadership method,³ and no systematic comparative analysis of a sample of national issues has been made. Since even those writers like Bachrach and Baratz who have raised objections to ‘the assumption that power is totally embodied and fully

This article is an account of a tentative attempt to solve a problem which arose in the course of an investigation of the career mobility of CPSU obkom first secretaries. At the outset we were only vaguely aware that a problem meriting investigation existed and, even now, all we would claim to be doing is presenting a report of the progress made so far towards its solution. Given, however, that research into the obkom tier is being carried on by an increasing number of scholars, we present such information as we have so far; and also recount the false starts and errors we made in the hope that this will be an aid (if only in a prophylactic sense!) to others working in the same area.

In the October 1973 issue of this Journal, John Orbell and Brent Rutherford published an article which sought to test the central thesis of Hobbes’ Leviathan. Although I shall try to show that as a piece of research it was fatally flawed, the authors deserve acclaim for their boldness, and for showing the way back to serious comparative government research in the Aristotelian manner. To test the basic principles of constitutions against the claims made for them is so obviously worthwhile that it is inexplicable and sad that their article stands virtually alone. In part its rarity reflects the unspoken assumption in modern political science that, after the behavioural revolution, anyone taking constitutions seriously was bound to be returning to the discredited legalism of the past. This overreaction has had its worst excess in political theory, for we seem all to have forgotten that political theory used to be an attempt to devise constitutional frameworks within which human nature would be led to produce the good life. As a theorist I hope we will return to that, and as an empirical researcher I hope we take up Orbell and Rutherford's lead and start to test such general constitutional claims. There are plenty of examples to go on with, starting perhaps with the claim that the United States Constitution is apt to ‘establish justice, insure domestic tranquillity, provide for the common defence, promote the general welfare, and secure the blessings of liberty’.

In a recent article in this Journal several conclusions concerning career patterns in the Irish political elite were drawn A.S. Cohan, ‘Career Patterns in the Irish Political Elite’, II (1972), 213–28. First, the revolutionary generation was shown to have effectively passed from the political scene. This was bound to happen with the passage of so many years since the war for independence. Second, the local orientation of the members of the Dail was shown to be very important even if TDs were members of the political elite, i.e., the Government. Third, with regard to a point discussed elsewhere but relevant to the above, the family connections of members of the elite were shown to be particularly important to their careers. The results of the parliamentary election of 28 February 1973 and of the presidential election of 31 May 1973 have in no way altered these conclusions.

Various writers have developed models dealing with the problems facing a political party about to choose its election platform. 1 This Note shows, by means of a simple model, the importance that parties ought, in making this choice, to assign to a particular class of voters. This class consists of a subgroup of the usual ‘floating voters’ (i.e. those who change their vote from election to election). The subgroup consists of those among the floating voters whose votes may be influenced by their satisfaction or dissatisfaction with the programs the parties adopt. Such voters may be designated ‘swing voters’. Whether or not such voters exist it is not the purpose of this Note to argue. Certainly political parties - in the importance they attach to their election manifestos and to the promises they contain - behave as though they believe they exist.

Ian Budge 'A Comment on Self-fulfilling and Self-defeating Predictions’, in this Journal, III (1973), 249–50 attempted to show that, despite objections to the contrary, enduring generalizations about political behaviour can be made even while acknowledging the existence of what Sir Karl Popper has called the ‘Oedipus effect’. Interestingly, Budge proceeds, not by questioning whether change in the level of information leads to change in patterns of behaviour, but by showing that under present conditions the amount of information possessed by the vast majority of the population is sufficiently low to preclude marked alterations of behaviour. A critic, however, might point out that Budge's line of argument leaves open the possibility that future changes in the level and accuracy of the information held by the population will lead them to alter their behaviour and that, consequently, the objection has not really been met. The imaginary critic's response highlights the difficulty of dealing with the so-called Oedipus effect when we proceed from the prevailing approach to the study of politics. I wish to indicate why this may be so, and to suggest a possible solution.