In recent years the discipline of political science has focused much of its attention on political parties. In 1967 Professor L. D. Epstein noted ‘Writing about Political Parties in Western democracies is not a novel enterprise. It is several decades since political scientists, once preoccupied with constitutional forms, ceased to neglect parties’. In Britain, however, there remains much uncharted territory which requires detailed exploration by the political scientist. Much of the research carried out in Britain has concentrated on one of two levels of analysis, either national or local. Basic information about the intermediate branch of party organization, the regional and area structures, is lacking. As Professor J. Blondel has noted: ‘The eleven regions of the Labour Party and the twelve areas of the Conservative Party are rarely examined’. There is, therefore, a gap in our knowledge of political parties in Britain, a gap which Professor R. T. McKenzie readily acknowledged in his own study of British parties.

Power has been so extensively analysed and so variously defined that a raw student (or even a seasoned practitioner) who looks at a dozen or so definitions of power might well wonder about the nature of the enterprise. The situation seems to be that there is some agreement on what power is — the word might otherwise be expected to have fallen into disuse — and some very sharp disagreement. But one is at something of a loss to know, firstly where the agreement is, secondly what is at issue between rival definitions, and thirdly which disagreements are basic and which mere refinements. A new analyst might well be puzzled about where to begin; and I rather think that all those who analyse power must be quite uncertain about what it would be to succeed in their task.

In a recent article in this journal1 we reported the results of an experiment designed to investigate, amongst other things, the effect of canvassing and polling-day activity upon the results of local government elections. On the basis of our data we argued that thorough canvassing before polling day and ‘knocking up’ on the day produced an appreciable increase in turnout. We also showed that, in a situation in which only one party organized the election thoroughly, this party would benefit disproportionately from the higher turnout; that is, it would receive an increased share of the vote.

I want to clarify the scientific status of contemporary political science. To do this, it is necessary to go into the question of what science means. I take it for granted that the samiliar charcterization of science as explanation is cssentially correct. But what does explanation mean?

Are local election campaigns just periodic acts of renewal, seasonal rites, or do they gain extra votes? Whatever may be true at general elections, low turnout is a reason for thinking that the scope for organization at local elections may be considerable.

The concept of political deference as an explanation for English voting behaviour has been both a popular one and one that has recently come under attack. Writers who have found it useful in explanation include McKenzie and Silver and Nordlinger. Writers who have attacked it include Parkin and Kavanagh. A somewhat intermediate position is taken by Jessop.

Mr Ake is not the first mechanical determinist to argue for the possibility of a political science analogous to natural science. He is aware that there are objections to this position, but he scarcely meets them.

A certain amount of evidence has accumulated in the past few years to show that local election turnout is high in marginal wards. A number of different sources make the point, but the most complete evidence is presented in a recent issue of Political Studies by Fletcher who summarizes the prevailing view when he writes that ‘there was a very strong inverse correlation between the size of the majority in a contest and turnout. The smaller the margin between the victorious candidate and his closest opponent, the higher the proportion of electors voting was likely to have been.’ In an earlier report it was suggested that the closeness of party conflict is the overwhelming influence on turnout. Only three writers fail to discover a high poll in marginal wards.

In recent years a considerable amount of work has been done on spatial models of party competition.1 The usual approach has been to represent the position of a voter on a given issue or issues as a point in a space of one or more dimensions, the position of a party or candidate similarly, and then to try to build models based on the distribution of voter positions and the relative positions of voters and candidates, using various criteria for party goals and strategy – e.g. maximizing votes, or maximizing plurality.

National stereotypes have an understandable attraction, but are often a snare and a delusion. There has been no lack of commentators, in recent years, to remind the French that their structures of local administration were laid down at the Revolution, and have been little changed since. Fragmented into nearly 38,000 urban and rural communes, which vary enormously in population and wealth, the country has more local authorities than the other five states of the EEC and Britain put together. Most people, on both sides of the Channel, accept the idea of reform, but we find something of a contrast when we look at what has happened in practice. President Pompidou said at Lyon on 31 October 1970:

It seems odd that in 1972, political scientists still find it necessary to take the scientific pulse of their discipline, to exhort, cajole, and recommend, in the manner of a reproving family physician, their colleagues to ‘pay more serious attention to what the scientific study of a phenomenon entails’. Such an injunction is doubly regrettable when accompanied by a naive and confusing analysis of concept formation and confirmation in science, an account sure to encourage a dogmatic epistemology among social scientists who, in other circumstances (specifically, the opportunity to devote scarce time to careful study of the voluminous literature in the philosophy of science), would likely identify themselves with a more catholic epistemology and methodology. Mr Ake, in his account of the logic of scientific inquiry, glosses major issues in the philosophy of science, and, as a result, tacitly represents the community of scholars in the latter discipline as univocal with regard to fundamental issues of language, logic, and epistemology. I shall briefly try to state more of the case than Mr Ake has done, in a manner so that I might not too be accused of glossing (or, at least, of glossing with a minimum of distortion) the major issues he has raised in his note.

'Party identification’ has been variously described: as an anchor; as constituting ‘psychological membership’ of, or ‘psychological identification’ with the party; and so on. The compilers of a recent collection of political attitude measures, published under the auspices of the University of Michigan's Survey Research Center (where most of the election analyses since 1952 have pivoted about the concept), even describe party identification as ‘probably the most important attitude variable in the field of political behavior’.² But the term itself and its connotations are misleading. Even where they do not suggest a borrowing from the technical vocabulary of psychoanalysisγ they tend to define away a set of questions that need to be answered: a preference for one party compared to what?; an intense preference, in what sense and compared to whom?4

In 1955 Professor H. L. A. Hart put forward the claim that ‘if there are any moral rights at all, it follows that there is at least one natural right, the equal right of all men to be free’, and this thesis and the arguments he adduces in its support have been thought sufficiently important for the article to be reprinted in a recent book of readings on political philosophy for students and general readers. The truth of Hart's thesis as stated is clearly meant to be independent of the moral stance of the asserter. Yet to my mind it is untenable as it stands, and could be modified only at the expense of taking up a particular moral position.

In the last decade, a large volume of research and speculation concerned with the beliefs and attitudes of the mass and the elite about democracy, has been published.1 Parallel with this trend has been a growing interest in political socialization. It is not surprising, therefore, that some work has concentrated on the attitudes of adolescents towards democracy and the patterns of development of such attitudes.2 The genesis of this interest can be found in the belief that a knowledge of what people in a democracy think about democracy can help us understand its operation and predict its future development. My aim in this Note is to look briefly at two pieces of work which are concerned with English adolescents in order to discuss the intellectual limitations of frameworks devoted to exploring this interest.

Local government politics, particularly in urban areas, is now dominated by the major national parties though the process and causes of the take-over are not well documented. These notes, based on research in Doncaster, relate the process of the take-over of the anti-Labour position by the Conservative party and draw some possible conclusions.

In a recent examination of size of place as a factor affecting electoral behaviour, Bealey and Dyer came to the ‘general conclusion’ that ‘the Labour vote in Britain is proportionately stronger in those constituencies which are part of large places’. Is such a relationship between size of place and Labour strength displayed in local elections?

In 1955 Professor H. L. A. Hart put forward the claim that ‘if there are any moral rights at all, it follows that there is at least one natural right, the equal right of all men to be free’, and this thesis and the arguments he adduces in its support have been thought sufficiently important for the article to be reprinted in a recent book of readings on political philosophy for students and general readers. The truth of Hart's thesis as stated is clearly meant to be independent of the moral stance of the asserter. Yet to my mind it is untenable as it stands, and could be modified only at the expense of taking up a particular moral position.

England has long been fertile ground for scholars (usually American) concerned to locate the antecedents of stable and democratic government. More often than not they have stressed a particular configuration of attitudes as a basic support for such government. Evaluation of the consequences for the political system of these attitudes has frequently proceeded along very inferential and impressionistic lines, and has resulted in a benign view of the British political system.

In his brief note on turnout and marginality in a recent issue of this journal Newton [K. Newton, ‘Turnout and Marginality in Local Elections’, II (1972), 251–5], raises some interesting questions. In our view, however, the boldness of some of his conclusions is not warranted by the evidence he presents or by his supporting argument. In particular, we think three points deserve comment.