thereby decrease the distinctiveness of the opposition and the chances for a strictly competitive contest between government and opposition. As a result of all these factors, both separation of powers and federalism confront an opposition with the alternative of carrying out a revolution to sweep the whole fragmented structure away, or else adopting a strategy for gaining goals by influencing the existing personnel of government rather than relying exclusively on winning elections and displacing the governing party or coalition- To increase the relative magnitude of the political resources available to the president, prime minister, or cabinet operates in the opposite direction- The more the power of the chief executive is increased relative to the power of legislators, the more an opposition will have to concentrate its efforts on the chief executive. This in turn increases the importance of winning elections in order to replace the executive. All three factors taken together generate pressures to concentrate the oppositions into a single coalition which, by its unity, has a chance to win the election; this in turn will emphasize the distinctiveness of opposition and push the system closer toward strict competition,
As to electoral arrangements, it is unnecessary to recapitulate that extensive controversy. The view that PR is a necessary condition for multi-party systems, and that single-member districts with plurality elections are a sufficient condition for a two-party system, is definitely untenable. Yet PR does seem to be a sufficient condition for more than two parties;
none of the nations using a complete PR system for national elections has a two-party system, and if the reasoning about the United States and Britain employed a moment ago is correct, then both these countries would in all likelihood move toward multiparty systems if they were to adopt PR for national elections. PR decreases the concentration and distinctiveness of opposition; reduces strict competition and increases the need for cooperation; and thereby increases the rewards to be gained from bargain- , ing strategies of various kinds. The argument of these paragraphs is summarized in Table 12.1.
Thus if constitution-makers wished to concentrate the opposition and
encourage the existence of two distinct and strictly competitive parties each "mploying a strategy of winning elections in order to form a powerful one-party executive, they should recommend a parliamentary system without marked constitutional separation of powers; an executive with ' relatively great political resources in comparison with the legislature; a unitary rather than a federal system; and election of members of parliament by a relative majority of votes in single-member districts. However, constitution-makers in some countries would be well advised, as we shall see, to hesitate about adopting such a constitution because of its explosive potentialities if political attitudes were to become highly polarized. Conversely, if constitution-makers wished to diffuse the opposition into a

variety of different sites and parties and to encourage cooperative-competitive strategies with a strong emphasis on gaining entry into parliamentary coalitions, they should recommend proportional representation, a relatively weak executive, constitutional separation of powers, and federalism. However, the price of such a constitution might be a certain paralysis in the executive except in a country with rather high consensus -which, since it would have nothing to fear from a concentrated opposition, would have no great need to employ these constitutional devices m order to diffuse and fragment the opposition.
Widely Shared Cultural Premises
That we are examining the products of historical developments should serve as a warning. Our ten countries have had different histories: even Swedes and Norwegians do not share the same past. The history of a country is in this one respect analogous to the past of an individual: behavior at any given moment is a product of interplay between the present situation and what has already been learned from responses to earlier situations- In an individual, the sum total of these learned orientations is his personality; in a nation or a country, its culture.
Both notions are diffuse, and we are barely past the threshold of scientific knowledge about personality and culture, particularly as they bear on politics. Nonetheless, few students of comparative politics doubt that certain countries do have different political cultures: that leaders and a great part of the people differ from one country to the other in their orientations toward politics. For example, as Alfred Grosser suggests, underlying attitudes about opposing the government are rather different in France from what they are in, say, Britain or the United States. A

Frenchman, typically, is more disposed to oppose, less disposed to support the government-any government.8
Despite the slender evidence, the conjecture that patterns of opposition may have something to do with the widely shared cultural premises of a country is much too important to ignore. Our data-indeed, all existing data-do not permit one to do justice to the conjecture. We do not even know what weight to assign cultural factors as compared with others. But we cannot ignore them. What follows, then, is not a summary of findings
but a sketch for a theory. Four kinds of culturally derived orientations toward politics seem to
have a bearing on patterns of opposition:
1. Oriepitations toward the political system. These orientations may be classified as allegiance, when attitudes, feelings, and evaluations are favorable to the political system; apathy or detachment, when attitudes, feelings, and evaluations are neutral rather than positive or negative; and alienation, when attitudes, feelings, and evaluations are unfavorable.7 In the United States, Great Britain, Sweden, Norway, and (in a more complex way) Holland, the political culture evidently generates widespread allegiance. In West Germany (and perhaps Austria), it seems to generate detachment. In Italy and France, and possibly to some extent in Belgium, it generates alienation 8 mixed with a large measure of apathy.
2. Orientations toward other people. Beliefs that one can have faith and confidence in others, or conversely that one should exercise distrust and suspicion toward others, seem to be culturally rooted to some extent, even if there are great individual variations around the cultural norm. Though comparative evidence is scanty, Almond and Verba found sizable variations in "faith in people" among their samples in five countries; "the Americans and British tend to be consistently most positive about the safety and responsiveness of the human environment, the Germans and Italians more negative, and the Mexicans inconsistent." 9
6. That the difference is one of degree and is subject to long-run changes is suggested by Bagehot's comment that "The natural impulse of the English people is to resist au'horky." The English Constitution, (New York, Dolphin Books, Doubleday,
n.d.), p. 306.
7. These concepts are derived from the first: systematic empirical study of political
culture on a comparative basis: Gabriel Almond and Sidney Verba, The Civic Culture, (Princeton, Princeton University Press, 1963), pp. 11-11. The concept of allegiance was in turn derived from Robert Lane, Political Ideology, Why the American Common Man Believes What He Does (New York, Free Press, 1961), pp. 170 ff.
8. For Britain, U.S., Germany, and Italy, see Almond and Verba, The Civic Culture, Ch. 14, p. 401, and passim. For the others, the classification is a purely qualitative Judgment based primarily though not exclusively on the essays in this volume.
9. Ibid., p. 168. For example, the percentages agreeing that "most people can be trusted" were: U.S., 55 per cent, U.K., 49 per cent; Germany 19 per cent; Italy 7 per cent (Table 4, p. 267). Whether the expression in quotation marks has precisely equivalent meanings in the different countries is uncertain.

3. Orientations toward cooperation and individuality. Some cultures emphasize the virtues of cooperating with others, conciliating opposing views, compromise, willingness to submerge one's own special ideas in a larger solution. Other cultures stress the value of maintaining one's individuality, distinctiveness, the integrity of one's personality and ideas preserving personal integrity by avoiding compromises. The political culture may, of course, stress these attitudes more-or less-heavily than the "general" culture. Though concrete evidence of cultural differences among our countries is difficult to find, it hardly seems open to doubt that in Sweden, Britain, and the United States, the political culture strongly emphasizes the virtues of compromise and conciliation and the possibility or compromising without threatening personal integrity; indeed, in these countries compromise is widely hailed as virtuous. In France and Italy, on the other hand, the virtue both to individuals and to groups of maintaining personal integrity and distinctiveness, even at the price of conflict,
seems to be relatively more heavily stressed both in the general culture and in political life.10
4. Orientations toward problem-solving. Sartori has emphasized the importance of looking "ac the underlying cultural patterns if we want to understand the difference between democracies of the Anglo-American type and, let us say, of the French type." u Englishmen and Americans, he argues, tend to be characterized by an empirical or (in a loose sense) pragmatic approach to problems, whereas among the French, Italians, and Germans,12 a rationalistic approach is more likely to dominate thinking about politics- Sartori sketches the two orientations as follows:
10. Direct evidence is hard to find. One indirect piece of evidence for "cooperation" versus "individuality" is membership in voluntary organizations. Almond and Verba found such memberships much more frequent among Americans and Britishers than among Italians and Mexicans, The Civic Culture, pp. 301-06. On Italy, see also Arnold Rose, "On Individualism and Social Responsibility," in European Journal of Sociology, 2, No. i (1961), 163-69. For insights on French "individualism" see Francois Bourricaud, "France" in Arnold M. Rose, ed., The Institutions of Advanced Societies (Minneapolis, University of Minnesota Press, 1958), pp. 408 ff.
n. Giovanni Sariori, Democratic Theory (Detroit, Wayne Stare University Press, 1062), p. 233.
12. While a weakness in empirico-pragmatic styles of thought is often ascribed to Germans, other observers might nor agree that their scvie is "rationalistic" in quite the sense of French and Italian "rationalism." Thus Deutsch and Edinger speak of "the two underdeveloped traditions of empiricism and equality: tv,'o muted themes in German culture . . - The mercantile traditions of empiricism, rationality, adaptability, and ease of compromise seem markedly underrepresented . . . German cul'ure thus offers its members two quite different roles for imitation: on the one hand the obedient, dependable craftsman, and on the other, the bold, romantic knight and his intellectual cousin, the daring, demonic magician." Karl W. Deutsch and Lewis J.
Edinger, Germany Rejoins the Powers (Stanford, Stanford University Press, icto), P- ^

While the empirical (empirico-pragmatic) mentality stays in medias res, close to what can be seen and touched, the rationalist mentality soars to a higher level of abstraction and hence tends to be far removed from facts. While the former is inclined to accept reality, the ra'ison tends to reject reality in order to re-make it in its own image; while empiricism tends to be anti-dogmatic and tentative, rationalism tends to be dogmatic and definitive; while the former is eager to learn from experience and to proceed by testing and re-testing, the latter goes ahead even without tests; while the empiricist is not deeply concerned with rigorous coherence and distrusts long chains of demonstration, the rationalist is intransigent about the necessity for deductive consistency-and therefore, in the summing up, while the former prefers to be reasonable rather than rational, the latter puts logical rigor above everything and thus is rational even if it means being unreasonable. While the empirical approach takes the attitude that if a program does not work in practice there must be something wrong about the theory, the rationalist will retort that what is true in theory must also be true in practice-that it is the practice, not the theory, that is wrong.13
It seems reasonable to think that opposition in these two hypothetical systems would differ in the following ways:
In System I, opposition would In System II, opposition VJcyitId stress the importance of: stress the importance of:
maintaining a stable government^ achieving goals, even if this leads adhering to the political and to instability in government constitutional rules of changing the rules if this is the game required by goals making evolutionary changes by making major structural changes,
marginal adjustments possibly by revolutionary means factual analysis rather than ideological consistency rather than ideological consistency factual analysis
Perhaps the most interesting difference in these two contrasting cultural orientations is the extent to which they would encourage consideration by political elites and activists of major structural changes in society, economy, and policy. Because System I would deal only with marginal alternatives, proposals for major structural changes would receive slight attention. Thus System I might prove to be rigid if it were confronted by problems that could not be met satisfactorily by the existing structures. System II, on the other hand, would encourage political elites and activists to consider major structural changes; but the emphasis on deducing the program from an abstract ideology, and the willingness to violate or change the constitutional and political rules to achieve political goals, might so thoroughly fracture consensus that discussion-a "dialogue"- about the alternatives would, in practice, be impossible. The structural alternatives presented by different groups would not be analyzed and discussed as much as merely proclaimed: in the place of discussion, there would be a dialogue of the deaf.
The two systems, it should be said again, represent the extreme cases. In practice, concrete political systems, including the ten examined in this book, would fall short of the extreme types. Probably the United States comes closer than any of the other nine countries to our hypothetical System I; the orientations and the opposkional behavior ascribed to System I are, in fact, usually attributed to the United States. Present-day Britain, the Scandinavian countries, and Holland also seem to display the characteristics of System I. Italy and France, on the other hand, probably approach System II somewhat more closely than any of the other countries described in the preceding chapters.
Yet it bears re-emphasizing chat while the evidence we have lends credence to these conjectures, our observations in this area are still mainly impressionistic. Much research remains to be done before we can be at
all confident about the interplay between political cultures and patterns of opposition.

Subcultures
Almost any difference in behavior or beliefs can lead to the development of so many special patterns of thought, language, identity, and other forms of behavior that we can appropriately label these patterns as a subculture. Once it develops, a subculture often displays remarkable tenacity, for the levers of change cannot easily be manipulated by outsiders, and those inside the subculture are rewarded not for changing their way of life but for adhering to it.
Conflicts involving subcultures are likely to be especially intense, and therefore particularly difficult to manage, because they cannot be confined to single, discrete issues; to the person sharing the perspective of a subculture, conflict over a "single" issue threatens his "way of life," the whole future of the subculture. The historic conflict of anticlericals and Catholics over funds for parochial schools was surely not often perceived, on either side, as a simple, straightforward question of how to finance education: it was a matter of rival ways of life or fundamental notions about freedom. In the United States, the question of voting rights for Negroes has never meant simply whether Negroes should vote; the typical white Southerner has perceived this, or any other extension to Negroes of the rights and liberties enjoyed by whites, as a threat to "the South"-in short, to a way of life built directly upon the subjection of
the Negro.
To some extent, all of the countries examined here are societies divided into subcultures. Occupation; social status; race, language, and echnic group; religion; residence; size of community; and region-these are nodes, everywhere, around which cultural distinctiveness develops. But it is when the numbers who participate in a subculture are large, the differences sharp, and the subculture comprehensive that the effects on political conflict are greatest.
Not one of our ten countries has wholly escaped conflict exacerbated by differences among subcultures (or between a subculture and the dominant culture), and in six of the ten countries conflicts of this kind have-at least for a time-directly shaped the characteristics of opposition:
In Britain, the conflict over Ireland is now substantially ended, but while it endured it involved a subculture based on region, religion, and ethnic identity. The conflict proved to be incapable of solution within the framework of the British political system.
In Norway, the conflict, now diminishing, between "center" and "periphery," based on region, residence, occupation, language, and to some extent religion has been important in political life during much of this century.

In Holland, the conflict over religion has led to the division of Dutch
society into three rather distinct communities.
In Belghmi, the conflict between Walloons and Flemings reflects subcultures different in language, religious views in part, and region.
In Austria, the conflict between two Lager has dominated the political scene since the beginning of the Republic.
In the United States, the South has for nearly two centuries formed a distinctive regional subculture with profound effects on American political life.
Since opposition between a subculture and a government that represents a different subculture or the dominant majority culture is highly explosive, how have our countries responded to those conflicts?
In general, conflicts involving subcultures rarely seem to be handled- for long-by the normal political processes employed in other kinds of issues. For this sort of conflict is too explosive to be managed by ordinary parliamentary opposition, bargaining, campaigning, and winning elections. The chief ways in which these conflicts have been dealt with are these:
1. Violence and repression. This has been a response in Britain, in Belgium, and, most notably, in the United States, where violence has even erupted into a civil war. The possibility of violence and civil war always lurks as a special danger in countries with hostile subcultures; and this danger undoubtedly stimulates a search for alternative responses.
2. Secession or separation. This was the solution to the problem in Ireland. It was the solution sought by the South that eventuated in the American Civil War.
3. Mutual veto on government policies. In this case, each opponent can veto changes in the status quo involving his subculture. This is most clearly the system in Austria, But it is also characteristic of Holland and, to a considerable extent, Belgium and the United States.14
4. Aiitono7ny. Autonomy may be granted to a regional subculture, as in the case of the South, which after about 1874-80 received, de facto, a grant of autonomy from Congress, President, and Supreme Court. Or, as
in Holland, autonomy may be granted to subcultures that do not have a regional basis.
5. Proportional representation has been used to guarantee a subculture that it will be represented in parliament even though, as in Belgium, it does not receive much autonomy. And proportional representation in the broader sense of representing groups more or less in proportion to their
numbers can be applied in all kinds of agencies and organizations, as in Belgium.

6. Assimilation. This seems to be the evolutionary pattern in Norway. For a century and a half it was the characteristic response to ethnic group differences in the United States; but the process of assimilation has failed, so far, in the case of Negroes.
The Record of Grievance
The extent to which citizens of a country are allegiant or alienated depends in some measure on the way the government has responded to grievances in the past and is expected by citizens or subjects to respond in
the future. Reflecting on observations made during 1950-51 in the Provencal village
of "Peyrane," Lawrence Wylie noted that
there never has been a time since the beginning of Peyrane's history, when contact with organized humanity has meant anything but the exploitation and manipulation of the individual. The wandering hordes, the Romans, the feudal lords-including the neighboring papal rulers, the agents of Provencal counts and French kings, the nineteenth-century regimes set up by Paris, the twentieth-century bureaucracy centralized in Paris-all these form an unbroken past in the vague memory of the village. They all mean domination by a human power beyond the control of the individual. At best the domination has brought unsought modifications in living habits. At worst it has brought disaster. And so it has become conventional to think of human power as a plague to be classed with the plagues of nature: the odious government, the leveling mistral, the flooding Durance.16
If the citizens of the United States and Britain are much more confident than the citizens of Germany, Italy, and Mexico that they could do something to prevent the passage by the national legislature of a law they considered to be unjust or harmful, as Almond and Verba report,16 surely these differences have something to do with the historic record of how grievances have previously been handled. To be sure, these differences in attitudes may now be embedded in the political cultures of these countries; but the political cultures themselves can be explained in part by the record of the past.
The burden of grievance is not, however, invariable in human societies. The institutions of a society generate a greater burden of grievance during some periods than in others; moreover, for a variety of reasons some societies evidently generate a smaller burden of grievances than others. Thus the accumulated burden of grievances may be low and allegiance high: either because the social and economic institutions of a country tend
15. Village in the Vaucluse (rev. ed. New York, Harper Colophon Books, 1964),

to generate a relatively small burden of grievance; or because, even though
great grievances have been generated in the past, the government has responded to alleviate them.
Returning to the French village ten years later, WyJie was surprised to discover marked changes even in that brief period:
As I went out into the country and talked to the farmers I realized the importance of the changed economic situation and state of mind of the farm people in the commune- They have a new awareness of their professional status and dignity, which is reflected in the growth of their professional organizations and in their sense of solidarity with farmers elsewhere ... It was obvious that French farmers were in agreement that if they wanted action there was nothing to be gained by talking to elected officials ... At last they had come to feel that their most effective political representatives were not the traditionally elected officers- deputies and senators-but farmers themselves who had been elected to office in professional organizations . .. Once there was no more hope of
turning back, the inevitability of change was accepted, it at last became possible for people to devote their energy to seeking new solutions for their problems. People acquired a greater sense of freedom to act for their own welfare. The rejection of the old political system, the modernization of the farms, the tractor demonstrations, the development of
farmers' organizations are. to a certain extent, manifestations of this changed spirit.17
For nearly a century the most insistent internal threat to the allegiance of citizens was the burden of grievance accumulated among the urban and (particularly in Italy) rural proletariats, a burden that threatened to alienate the working classes if the grievances were not heeded-or to alienate the middle classes if they were. The problem, then, was how to gain the allegiance of the working classes to democratic institutions without alienating the other social strata. This problem has not been solved in all ten countries; France and Italy in particular do not seem to have worked out a solution, though they may be in the process of doing so. In Germany (and perhaps in Austria) commitment to democratic institutions is untested by adversity and may still be somewhat weak; but working classes, at any rate, are as allegiant as other major social strata, and in this narrow sense a solution has been reached to the problem of political alienation among the working classes. In the remaining six countries, the problem has been solved, at least in the main. One of the significant developments of this century has been, then, the integration of the working classes into a large number of Western democratic systems and the decline
in these countries of any serious threat to parliamentary democracy 17. Village in the Vavcluse, pp. 359, 362, 364.

generated by class conflict. The importance of this development for the future of oppositions can scarcely be exaggerated.
Even aside from France and Italy, however, the integration of the urban and rural workers into political life has not followed an identical pattern.18 Despite the diversity of specific patterns, it is possible to clarify some of the major aspects of the process by drawing a simplified and abstract picture of how a political system might solve the problem of acquiring the allegiance of the working classes to democratic institutions without alienating the other strata, particularly the middle and upper strata. One might conjecture that the likelihood that this problem would be solved would be relatively high if the following conditions were met:19
1. As the urban (or rural) working classes increase in numbers and in demands on the political system, there already exists an operating parliamentary system supported by a large and allegiant middle class and led by an experienced and allegiant political elite.
2. These incumbents, the middle classes and their leaders, head off severe and prolonged frustration over the operation of the political system by peacefully yielding an increasing degree of participation in political life to the working classes by
the extension of the suffrage,
allowing the development of political leaders representing the working strata,
permitting the participation of these political leaders in political decisions,
and accepting their entry into the government.
3. Entrepreneurs and managers yield enough of the benefits made possible by more efficient technology and organization to reduce the frustrations that had been generated by social'and economic conditions during early industrialism.
4. The government undertakes regulation or structural reforms in such a way as to reduce social and economic sources of frustration ito people in
working-class occupations, without, however, seriously alienating other social strata.
Six of our ten countries have, in varying degrees, satisfied these four general conditions: Britain, Belgium, Holland, Norway, Sweden, and the United States. Of these, the five European countries have followed a roughly similar development, while the United States has traced a different
18. Cf. Vfll R. Lonvin, "Working Class Politics and Economic Development: in Western Europe," American Historical Review, v. 63 (January 1058), 338-51.
19. Although the formulation here is not identical with S. M. Upset's, it has much in common with his. See his discussion of legitimacy and democracy in "Some Social Requisites of Democracy," reprinted in Nelson W. Polsby, Robert A. Dentler and
Paul A. Smith, eds.. Politic! and Social Life (Boston, Houghton Mifflin, 1963), pp. 541-68, at pp. 554 ff.

path. In the five Norch European countries, a constitutional system with an elected parliament, middle-class support, and experienced leaders drawn from the middle classes or the aristocracy was already in existence when the urban and rural workers began to develop political consciousness and demands. In all these countries viable labor or socialist parties that drew increasing support from the working classes were formed in the years from about 1885 to 1905. By the end of this period or a few years later the right to vote was gained by most workers in all these countries.20 Even the existing middle-class parties began to respond to the demands of the new strata, and they sometimes supported measures designed to eliminate, or at least to mitigate, the worst social and economic evils of unregulated capitalism: social security laws, legal support for trade unions, expansion of public education, and the like. These reforms did not, however, prevent the growth of the labor and socialist parties. During the First World War and after, socialist leaders began entering inco coalition governments. Although the labor parties and their leaders participated initially only as junior partners in coalition governments, eventually they secured a large enough vote in elections and seats in parliament so that, as the largest or second largest party in the country, they were highly influential in the conduct of government-sometimes as the major opposition to the government, sometimes as a coalition partner, sometimes as the governing party. In the extreme cases, Norway and Sweden, the labor parties have governed for three decades; in Britain, their victory in 1945 enabled them to carry out most of their immediate program; in other countries, even when the labor parties are not currently participating' in the governing coalition it is reasonable for them to expect thaC they may be able to do so within a few years,
Thus in these five countries the first two conditions mentioned a moment ago have been met, and in the process the working-class parries have acquired the sobering experience of governing their countries. Meanwhile, too, the third and fourth requirements have been met by both economic and social development and by government regulation and reform. In all five countries the labor parties have witnessed the attainment of a good many of their immediate objectives, while their ultimate goals for a socialist society have more and more become hypothetical and rhetorical.
In the United States the four conditions suggested earlier have been met-but according to a different timetable and in quite a different way. As in the five countries just mentioned, so too in the United States, well before rapid industrialization expanded the urban working class from a
20. On the development of political participation in Western Europe, see Stein Rokkan, "The Comparative Study" of Political Participation: Notes Toward a Perspective on Current Research," in Austin Ranney, ed., Essays on the Behavioral Study

tiny minority into a substantial proportion of the population, democracy was a going system of government, backed by a "middle class" (consisting predominantly of farmers) and operated by leaders skilled in the arts of managing the country's political institutions.
But the United States met the second condition by providin"' for participation of the emergent working classes in political life in a way that frequently leads European socialists to the conviction that this country has skipped "an inevitable stage of capitalist development" and that sooner or later (despite impressive and increasing evidence to the contrary) it will have to rum back and rerun its history according to the North European pattern. The essential difference is that unusual conditions in the United States permitted it to arrive at the same result-the integration of the working classes into the political system-without the mediating role of a specifically working-class party. The transition was far from peaceful. "American workers," Val Lorwin has said,21 "had to fight bloodier industrial battles than the French for the right of unions to exist and to function." Nonetheless, in the United States the working classes have always been, on the whole, strongly allegiant to the political system;22 and though many workingmen's or socialist parties have been created to appeal to urban workers, few have lasted more than a decade and none has ever attracted the continuing support of more than a small proportion of the
working classes. The reasons for this great difference between the United States and the
European democracies with which it has much in common in other respects are complex. One was the fact that workers had already acquired full political rights long before an urban proletariat of any size came into existence; thus workers were never alienated from the political system as a result tf being excluded from it. In addition, two national, "grass-roots" parties were already on the scene, ready, willing, and able to recruit workers or supporters. To call them "bourgeois" parries is to miss an important point. If they were "bourgeois"-American middle-class-in their orientations, they were by no means so in organization and recruitment. They were distinctly not elite parties of notables. Typically, they were organized all the way down to the poorest precincts of the large cities; and they had perfected techniques-in essence, rendering primitive social services-for gaining and holding the loyalty of their followers. A working-class party could offer distant collective goals; it could not compete in immediate individual palliatives for concrete grievances. Finally, in a very large country with a decentralized federal system of government and parties, many of the most pressing legislative demands of
21. Val R. Lorwin, "Reflections on the History of the French and American Labor
Movements," Journal of Economic History (March 1957), p. 37.
22. In this respect the 15 men studied in depth by Robert Lane seem representative. See Political Ideology, Ch. 10, "The Alienated and the Allegiant," pp. 161-76.

as fully integrated into, and as allegiant to, their democratic institutions as the other strata. In Germany, as Ott-o Kirchheimer indicates, class antagonism no longer serves as much of a political stimulus; and what was for generations considered the leading socialist party of Europe is now barely distinguishable in goals and strategies from the Christian Democratic Union. In Austria, coalition government and Proporz have provided the old enemies, the two Lager, with a verv great stake in maintaining the system. The commitment to democracy may not be strong in either country; but allegiance is no weaker among the working class, evidence suggests, than among other major strata. It is not so much that the battle for allegiance has been won as that it has not been lost. The evidence of Almond and Verba seems to indicate that in Germany-and the same may be true in Austria-the population is neither alienated from nor allegiant to democratic institutions but rather indifferent or detached.26
Like Germany and Austria, the political development of France and Italy did not fulfill the four conditions suggested earlier. In both countries, the last two conditions were, at least until recently, met badly. In addition, Italy scarcely even fulfilled the first condition during the 6o-year period of parliamentary government before fascism, for it failed to build up a large and allegiant middle class. Meanwhile, an extraordinary record, probably unparalleled in Western Europe, of violence and repression by the state against socialists, anarchists, trade unionists, and other workingmen's organizations must have strengthened the suspicion and hostility of the working classes toward parliamentary institutions- "What is more, Italy arrived at the second condition rather late by maintaining a highly restricted suffrage until universal manhood suffrage was introduced in 1912:
the result was an abrupt threefold swelling of the electorate. The First World War following upon the elections of 1913 and the rapid collapse of parliamentary government after the war meant that in Italy, as in Germany and Austria, the struggle for allegiance really began only two decades ago. So far, as Samuel Bames* essay makes clear, the outcome remains in doubt. Let me now draw this discussion together with three conclusions:
First, in seven of our nine European countries the politics of the working strata has converged toward a situation that has much in common with the traditional position of the American worker. An American is likely to consider desirable any change that brings European systems into conformity with his own. But the problem is not quite so simple.
For, in the second place, one cannot simply ignore the plain fact that, up to now, in Italy and France history has taken a different path. It is not yet clear, particularly in Italy, how the largest group of working-class
voters, those who have voted for the Communists, are to be integrated peacefully into a viable political system.
26. The Civic Culture, pp. 418 ff.

Some Explanations
Third, when the Left parties acquire a durable majority of votes, the problem of allegiance shifts from the working classes to the white-collar strata: lower middle classes, professional groups, businessmen. The problem would arise in Italy in its most extreme form if there were ever a coalition government in which the Communists participated. And for how long can a nation's second largest party, with the largest following of manual workers, be excluded from participating in the government? The problem is posed in much more moderate form in countries where the Left parties have become the largest parties and normally in control of the government; most notably Sweden, Norway, and the United States, where the parties representing the conservative middle classes have not been able to control the national government for the better part of the last three decades. To what extent will conservatives (whether in the Scandinavian or the radically different American sense) become politically alienated by their continuing exclusion from governing their countries? Certainly the American Right appears to be fully as hostile to the dominant tendencies in current American politics as the working classes of the United States ever were. And the strike of the Belgian doctors in 1964 shows to what length a professional group may go in opposing government policy.
Social and Economic Sources of Political Cleavage
An alternative though not necessarily contradictory way of describing and explaining these historical changes in the patterns of opposition in our ten countries is to attribute the changes in politics to long-run changes in social and economic factors, using these terms very broadly to include not only class, social status, and occupation but also such social factors as religion, ethnic group, and language. The political affiliations, loyalties, and attitudes of an individual, it might be said, are heavily dependent on his durable social and economic roles, functions, and affiliations: political cleavages are, according to this hypothesis, the expression of social and economic differences.27 If the political behavior of various strata has become more alike in recent years, then this is because these strata have also become less distinct in their social and economic characteristics.
There is, surely, a great deal of truth in this explanation. The core of truth is to be found in the twin assumptions that social and economic differences usually are associated with differences in rewards and deprivations, in relative advantages and disadvantages; and that these differences in rewards and deprivations stimulate cohesion among those who are socially similar and conflict with those who are different. The difficulty with the explanation and the axioms on which it is founded is not so much
27. A special problem implicit or explicit in thi^ kind of explanation, the relation of social and economic cleavages to "polarization," is dealt with below, at pp. 380-86.
that they are false as that they oversimplify a highly complex matter and thereby leave a great segment of political behavior unaccounted for.
To begin with, in the countries examined in this volume, the political affiliations, loyalties, and attitudes-in short, the political behavior-of individuals and groups cannot be traced to any single social or economic characteristic. Quite the contrary, political behavior is evidently influenced by a great variety of different social and economic characteristics.
The most important of these are differences with respect to economic position (in a broad sense), social position, religion, language or ethnic group, and region. Differences in economic position, ordinarily allied to differences in social position, are sources of conflict in all our countries. Religion is usually an important source of differences in political behavior in countries with sizable Catholic populations, because religious commitments generate differences in political attitudes between Catholics and Protestants or between Catholics and anriclericals, as in France, Italy, and Belgium. Differences in language or ethnic identification have been critically important in Belgium and the United States. And regional differences-associated often with some of the others-have been significant in the United States. Norway, Belgium, and, to some extent, Italy,
Second, some social differences are associated with differences in relative advantage, and hence with political behavior, in some countries but not in others. Differences in language are one of the commonest sources of cleavage in all parts of the world. In Belgium, as Lorwin points out, the bitterness between Walloons and Flemings arises not only from the language difference but also from the fact that language has been associated with a pattern of inequalities in occupations, education, social class, prestige, and opportunities to rise in Belgian national life. In Switzerland, on the other hand, although there are often abrasive relations at the cantonal level, conflict between German- and French-speaking Swiss has been largely avoided in national politics, evidently because few persons in either language group feel that: they have been unfairly treated in Swiss national life as a result of their language.
Third, the ways in which different socioeconomic factors are related to one another do not produce a single pattern of cleavage in the different countries, buc a great profusion of patterns. In a rough way it is possible to distinguish three patterns:
a. Countries highly homogeneous except for social standing and economic position (which are highly correlated). Among our countries these are Protestant nations where religious differences have ceased to be salient, and where ethnic, language, and regional differences have left only slight traces. Britain and Sweden fall most clearly into this class; Norway is a somewhat marginal member because of the declining but not insignificant influence of region, religion, and loyalty to the rural language. Because

these differences are at a minimum, the most salient differences are those in occupations, incomes, and social standing-socioeconomic status. As a consequence, in countries like Britain and Sweden the political parties reflect more clearly than in other countries the differences in the class composition of their followings; the labor-socialist party draws the bulk of its strength from blue-collar workers, while the middle-class panics (or party) receive considerably more votes in the middle or upper strata than does the labor-socialist party. In this sense there is a higher degree of "status polarization" in politics than in other countries. But, as we shall see, this polarization is probably statistical rather than psychological.
b. Countries in 'which several kinds of socioeconomic differences coincide and thus reinforce one another. One of the surprising results of our inquiry is that this pattern does not exist in anything approaching a pure form in any of our ten countries. This may well be because a pattern of this kind would lead to such severe conflicts that a parliamentary system would founder. In each one of our countries where several different kinds of socioeconomic differences stimulate political conflicts, the planes of cleavage in one conflict do not coincide exactly with those in other conflicts-fortunately for the survival of the system. However, some countries
approach this pattern more closely than others. Belgium falls most clearly-even if imperfectly-into this category, for
there, as we have seen, two of the three main planes of socioeconomic cleavage do coincide to some extent. The explosive nature of the question of the Negro in the United States also stems in considerable part from the fact that the most pronounced line of cleavage, that between North and South, goes hand in hand with differences in ideologies, economic issues,
social systems, and regions.
c. Countries in which several kinds of socioeconomic differences crosscut one another. It is in this case that the effects of "overlapping memberships," "conflicting identifications," and "cross-pressures" so much discussed in the literature of American political science may occur.28 Thus a Dutch Catholic worker may have a strong sense of his identity both as a Catholic and as a member of the working class. Potentially, then, he is in conflict with non-Catholics on religious matters and with nonworkers on economic issues. Both his class and his religion are durable features of his life. As a worker living in a working-class ambience he finds common ground with other workers, be they Catholic or non-Catholic; as a Catholic living among Catholics he feels solidarity with other Catholics,
28. For a summary and critique as of 1958 see Robert E. Lane, Political Life (Glencoe, Free Press, 1959), pp. 187-203. The overlapping membership model par-riculariy as it applies to the United States is examined by William C. Mitchell in "Interest Group Theory and 'Overlapping Membership,'" a paper prepared for the 59th Annual Meeting of the American Political Science Association, New York City,
September 1963.

whether bourgeois or workers. He therefore has an incentive, particularly if his identification with each group is strong, to seek (or to accept: at the behest- of his leaders) ways of reducing conflict: within each group. His desire for group solidarity may stem from psychological needs, from a fear of the collective consequences of internal divisions, from a strategy of influence in politics and economic life, or from all of these and other reasons. Whatever the reasons, though responses other than compromise are possible, he is likely to be receptive to compromises on questions involving religion or economic matters, for conflict threatens to divide a group whose solidarity he wants to maintain.
Although socioeconomic differences help to account for patterns of opposition, they leave a good deal unexplained. Even in a country with such sharp social and economic differences as Italy, political conflict is by no means purely a matter of conflict amorig different social and economic groups. Industrial workers evidently divide their votes among Communists, Socialists, and Christian Democrats. Even in a homogeneous country like Britain, where socioeconomic status has greater influence because of the weakness of other kinds of social distinctions, a third of the workers vote Conservative. Or again, compare the fate of the labor parties in Britain and Sweden. In Britain, the most urbanized and industrialized of our countries, the Labor Parry has spent most of its history in opposition;
while the middle classes are rather solidly united in a Conservative Party which, thanks to the support of its working-class voters, managed until the election of 1964 to remain in office for all except 6 years out of the previous 30. In Sweden, less urban and industrial than Britain, the Socialist Party not only wins the support of two-thirds or more of the working class, as the Labor Party does in Britain, but also gains a quarter of its votes from nonworking-class groups.29 Thanks to middle-class support it
has been in office for 30 years, and it is the bourgeois parties that have formed the opposition.
Why do socioeconomic factors account for only a part of the variation in patterns of opposition? Principally, it seems, because the causal chain from one's socioeconomic position to one's overt political action is long and tenuous; and each link in the chain may be weak enough to be broken by the pull of other forces. A pure social determinist might postulate a causal sequence in which one's overt political acts are completely determined by one's socioeconomic position.
But we know that each of these links can be so weak that it cannot bear the weight of the others. Whenever objective differences in socioeconomic

positions are blurred and ambiguous, as they are, for example, amon^ clerical workers, then there may be only a weak correlation between socioeconomic position and subjective identification; hence while a majority of white-collar workers might identify themselves as middle class a sizable minority might see themselves as working class. Moreover, for a variety of reasons subjective identifications vary in strength, and if one's identification with one's occupational group is weak, they may not have much to do with one's political opinion-as was probably the case with the wirer in a California radio factory who said: "Well, I work for a living so I guess I'm in the working class." 30 Finally, political opinions may be weakly related to overt political acts, particularly among people who are uneducated, who are unable to conceptualize abstract political ideas, or who are not interested in politics,
The link at c, between political opinions and overt political aces, is usually stronger among educated persons, intellectuals, and political activists than among the general population; conversely, however, among these very groups, the links at a and b connecting objective position, identifications, and political opinions are somewhat weaker than among the general population. Thus with the general population the breakdown in the hypothetical causal chain of the social decerminist is likely to occur closer to the terminal end, at c or b; but among the political elites, the break is more likely near the beginning, at a.
The determinist's chain of causation is thus rather easily broken by the intrusion of factors that he assumes are irrelevant or extraneous. The British middle classes may not agree in their ideas more than the Swedish middle classes; but in Sweden PR encourages the middle classes to distribute their votes among several parties, whereas in Britain the single-member district and plurality elections make it simple for them to concentrate their votes on the Conservative Party. Other factors also have an opportunity to express themselves, as we shall see in a moment.
Specific Patterns of Attitudes and Opinions
I have already referred to attitudes and opinions in order to explain patterns of opposition. Widely shared cultural premises consist of attitudes, feelings, and evaluations held, presumably, by a substantial proportion of leaders and the general population. Subcultures are relatively distinctive sets of attitudes, opinions, and values that persist for relatively long periods of time in the life of a country and give individuals in a particular subculture a sense of identity that distinguishes them from individuals in
other subcultures.
In this section, however, I am not concerned with the widely shared
30. V. 0. Key, Public Opinion and American Democracy, (New YoA, Knopf, 1961), p. 141,0.5.

in Figure 12.i.31 Suppose, further, that the distribution were unimodal and similar to the familiar bell-shaped pattern (Figure 12.1); that political leaders were more or less familiar with the distribution of opinion; and that they were anxious to win elections by adopting a position on this
attitudes that go into a general political culture nor with the content of political attitudes found in specific subcultures or other distinctive groups-I am concerned only with the -patterns of cleavage and consensus formed by the ways in which political attitudes are distributed over the population of a country. What follows is an effort to draw together a tentative theory to explain how four factors may increase or decrease the incentives of political leaders to pursue conflicting goals and strategies (or, conversely, conciliatory goals and strategies).
These four factors are:
The distrib^ition of opinions on political questions-specifically whether
the distribution is single-peaked or bimodal.31
The coincidence of opinions among different individuals, that is, the
extent to which individuals who agree on one question agree on others.
The salience or intensity of opinions on different questions. The institutional means for aggregating opinions, and specifically the
oparty system.
Let us begin with the last factor, for a belief widely expressed in the literature of political science is that a two-party system has a moderating influence on the selection of goals.32 In a two-party system, it is often said, the government party and the opposition tend to converge Coward common ground because both parties compete for the great mass of voters whose opinions on political questions differ only very little: the center. Yet this hypothesis assumes that there is a great mass of voters responsive to "centrist" ideas and proposals. But this might not be the case. If not, our hypothesis is that a two-party system would not necessarily lead to conciliation and compromise but might actually intensify political conflict. Suppose that opinion on some critical political question33 were distributed along a continuum from the extreme left to the extreme right, as
31. Obviously there are many other distributions; but to add others not only increases the complexity of exposition but takes theory well beyond the limits of the
data at hand.
32. Assertions about the moderating effects of a two-party system are surely among the oldest and most widespread in the modern study of political parties, and a number of writers on parties have provided explanations derived implicitly from the assumptions and argument in these paragraphs. See in particular A. Lawrence Lowell, The Government of England (z vols. New York, Macmillan, 1908) and Public Opinion and Popular Government (New York, Longmans, Green, 1913)-, and E. E. Schattschneider's discussion in Political Parties (New York, Farrar and Rineharc, 1941) on "the moderating effect of the attempt to create a majority" in a two-party system (pp. 85 ff.). Lowell recognized the possibility that the distribution of opinion necessary for a two-party system to have a moderating effect might not exist: in this case, however, it was not a "true public opinion" and it would be impossible to conduct the government "by a true public opinion or by consent" (Public Opinion and
Popular Government, p. n).
-.-. T ncp rhp word "nnesrion" deliberately in this discussion to include issues, candi-

issue that would gain them maximum popular support. Suppose, now, that the ideologues of Party A were to advocate an extreme left-wing policy represented by the position Ai. The leaders of Party A would see that this extreme position would very likely cost them nearly all of the votes of people whose opinions were to the right of Ai, provided only that Party B were astute enough to take a position slightly to the right of Ai; hence if Party A were to take a position Ai, it might retain its extreme left wing bur would probably lose votes catastrophically to Party B- In precisely the same way, if the ideologues of Party B were to advocate Bi, the party leaders would see how heavy a price they might ultimately pay in order to satisfy their extremists. Thus both parties would have strong incentives driving them toward positions close to the middle that have the support of the preponderant majority of citizens, say As and Bs. If the parties were to adopt As and B,, most voters would feel that the policy adopted by the government was an acceptable one no matter which party were to win the election. And compromises between the two parties would not be par-
14. Spatial models, like those used in the exposition that follows, derive indirectly from Harold Hotelling, "Stability in Competition," Economic Journal, 39 (1929), 4'-57i but became an explicit basis for a theory of party competition in Anthony Downs, An Economic Theory of Democracy (New York, Harper, 1957). They were also employed in my A Preface to Democratic Theory (Chicago, University of Chicago Press, 1956), pp. 93 ff., which set out some of the argument that follows. The unrealistic character of some of the assumptions made in using spatial models has been severely criticized by Donald E. Stokes in "Spatial Models of Party Competition," American Political Science Review, ^7 (June 1963), 368-77. Scokes' criticisms are, I

ricularly difficult, since positions As and B, are not, after all, very far
apart.
The effect of party competition in this situation would be to reinforce moderate or central opinion. By responding to the moderates and proposing moderate policies, the parties would help to strengthen moderate opinions and moderate leaders, which in turn would reinforce the tendencies of the parties to adopt moderate positions.35 The moderating effects of party competition are likely to be particularly strong in a system under which legislators are elected in single-member districts by winning more votes than any single opponent. For in such a system, an extremist party advocating, say, Ai might gain the votes of left extremists only to be swamped in the elections by the votes of the center. Moreover, an extreme left-wing party at A i might well cause enough defections to prevent a moderate left party at At from winning the election; hence to vote for the extreme left party could produce policies even more intolerable to the left extremists than the policy advocated by the moderate left party. Consequently voters at the extreme left are confronted by the prospect that they will not only throw away their votes by supporting a party that advocates their views but they may well help to bring about a victory of a party on the right. Similar reasoning applies, of course, to the right.30
Even under a system of proportional representation with, let us say, four parties, dominant coalitions would tend to move toward moderation. For the moderate center parties would stand the best chance of winning the most votes; by winning the most votes they would acquire the most seats; and thus they would have the greatest influence in a governing coalition. Indeed the two moderate parties might well combine to form a governing coalition. Buc even a moderate party that considered the prospect of forming a coalition with the party on its extreme flank would hardly find it worthwhile to bargain away its own policies for chose of its lesser partner simply to gain the support of the smaller party; for to do so would cause the moderate party to lose the bulk of ks supporters to its moderate rival. Consequently the moderate parties would bargain from positions of strength, the extreme parties from positions of weakness.
The preceding analysis is probably correct, at least roughly. Yet it rests entirely on the assumption that opinions do in fact have a single mode. And there is no reason at all to rule out the possibility of radically different
35. In this discussion "moderate" means no more than being near the center of some distribution of opinions. In this sense, "moderation" has no intrinsic or necessary virtues and implies no psychological qualities in the "moderate" person or the supporters of a "moderate" party.
36. All these arguments were used, in less formal language, by moderate opponents of Senator Goldwacer in their unsuccessful attempt to block his nomination by the Republican Party in 1964. The election demonstrated, I think, that their arguments were correct.

distributions of opinions. For example, some great question might divide the opinions of citizens into two camps as in Figure 12.2. Now a center scarcely exists; and the ideologues of moderation labor under the same handicaps as the ideologues of extremism under the single-peaked dis-

tribution. To the contention of moderates in Party A that they should hold to position At, the left will argue that the great bulk of Party A^s support lies toward the extreme; and if Party A takes a moderate position it might well see its support sapped by another party of the left. And so, too, on the right. Once again, the single-member district system with plurality elections reinforces the cleavage in opinion. If, for example, a moderate left party were to adhere to ics traditional position of centrist moderation, it would run the risk that a new party would arise and advocate a position to the left of As. Such a party would be likely to swamp the old moderate left party (as Labor in Britain did with the Liberals.) 3T Thus when opinion is bimodal a two-party system based on single-member districts and plurality elections is likely to intensify rather than to mitigate conflict either by inducing the existing parties to shift to the extremes or by generating a new extreme party (and thus ceasing, at least temporarily, to be a two-party system). For a center party will, in rime, be reduced to a corporal's guard by the effects of the election system. Indeed, it is altogether likely that with a bimodal distribution of political opinions a two-party system may intensify conflict even more than a
multiparty system and proportional representation might do.
To be sure, a multiparty system cannot create moderation and conciliation where they do not exist. With proportional representation, the for-
37. Cf., for example, John Bonham, The Middle Class Vote (London, Fabcr and Faber, 1954), pp. 149 ff., and Ivor Jcnnings, Party Politics, The Growth of Parties (Cambridge, Cambridge University Press, 1961), pp. 250 ff.

marion of extreme parties would be inevitable; and by making it difficult for the center parties to form a viable coalition the growth of support for extreme parties would increase the chances for coalitions influenced by the extremes. This was the problem of Weimar Germany, and it is to some degree the situation in Italy. Yet where polarization is less than complete, proportional representation and multiple parties may help a center to survive as long as any significant number of left-center, center, and right-center opinions persist. Thus the center parties may manage to retain some bargaining power in the coalitions of government and opposition and thereby to weaken, at least partly, the strength of incentives to antagonism and conflict.
One must therefore qualify the hypothesis that two-pany systems necessarily foster moderation and compromise while multiparty systems encourage antagonism and severe conflict. We conclude instead that u'Aen voters* opinions are (and are thought to be) tinimodal, both a two-party system and a multiparty system are likely to lead to moderation and compromise among the leading parties. When^ on the other hand, opinion is strongly polarized in a bimodal pattern, two parties, each striving to retain the sztpport of the extremists on its flank, 'will only exacerbate a conflict;
and in multiparty systems the center parties 'will decline in votes and influence.
So far in this discussion I have implicitly assumed that it is possible to characterize opinions on all important questions or conflicts by means of a single, summary distribution. Thus if Smith is to the "left" and Green to the "right" of Jones on one question. Smith will be to the left and Green to the right of Jones on other questions. But of course this is not necessarily the case.38 The extent to which the opinions of individuals or groups coincide from one question to another may vary; and these variations are likely to encourage different patterns of opposition.
Thus the relative strength of incentives for conciliation and antagonism does not depend only on whether opinions are unimodal or bimodal, but also on the extent to which opinions coincide from one political question to another, that is, on the extent to which individuals or groups who agree (or are close to one another) on one question agree on other questions.
If there is a high degree of coincidence, so that individuals who agree on one political matter agree on others, then the effect is evidently quite straightforward: high coincidence reinforces the effects of unimodal or bmzodal distributions of opinion.
If people who are moderate on one question are moderate on others, and if most people are moderate on every question, then the incentives of party leaders to engage in a. search for moderate policies are, of course, very much increased. If, on the contrary, people who are at opposite

extremes on one question are likely to be at opposite extremes on all questions, and if most people are extremists, then moderate, conciliatory parties or coalitions will be even more likely to yield to intransigent political groups implacably at odds with one another.
If the effects of high coincidence seem clear, the effects of low coincidence are complex- To begin with, as compared with high coincidence, low coincidence of opinions increases the number of different clusters of persons with divergent opinions. Each of these "opinion-clusters" is potentially a separate political following-potentially, therefore, the basis of a separate political movement, whether pressure group, faction, or party. The point may be clarified by using Belgium as an example.
Belgians, as Lorwin has shown, have been at odds over three sets of questions. If we consider them as pro-con questions for ease of exposition, then in Belgium an individual might be characterized as:
either pro-labor or pro-property;
either anticlerical or pro-Catholic; and
either pro-Walloon or pro-Flemish.
Because three dichotomous attributes can be combined in eight ways, in principle every Belgian might be located within one of eight ppssible combinations; for example, individual A, B, C, D, etc. might be:
A: pro-labor/anticlerical/Walloon
B: pro-labor/Catholic/Walloon
C: pro-property/anticlerical/Walloon
D: pro-labor/Catholic/Flemish
etc.
At one extreme, coincidence among the opinions of Belgians might be maximally high. Individuals who agreed on one question would agree on the others; those who disagreed on one would disagree on the others. Thus by knowing how a Belgian stood on one of these questions one could predict perfectly how he would stand on the other two. Under the particular conditions of our example, this would mean that every individual would fall into one of two opinion-clusters. (Technically, one cluster might be empty, but we ignore that possibility.) For example, all Walloons might be pro-labor and anticlerical; and all Flemings Catholic and pro-property. In this case, the society, and presumably political life, would be polarized.39 At the other extreme, coincidence among opinions might be at a minimum, so that if two persons agreed on one issue (if, for example, they were both Walloons) nothing could be predicted as to their opinions on the other issues. Hence all eight opinion-clusters might be
in. Thit result follows from the particular example. Conditions could be specified

represented in the population. Each would be a potential foundation for a separate political movement antagonistic to all the others-40
Yet low coincidence does not necessarily produce this kind of fragmented antagonism among opinion-clusters. In some circumstances, as we shall see, low coincidence may strengthen incentives for conciliation. The extent to which incentives for conciliation (or antagonism) are strengthened (reduced) appears to depend on two additional factors: the salience
of a political question or intensity of opinions, and the available instim-tional means for aggregating opinions.
Suppose that in Belgium there were not only a low coincidence of opinions but that all three questions evoked extremely intense views on the part of everyone. In this case each opinion-cluster would tend to become a separate political movement, and it would be extremely troublesome for leaders to build political coalitions combining two clusters. For example, Cathplics would split down the middle over labor and property and Walloon Catholics would split with Flemish Catholics. There would be four distinct Catholic parties, or four distinct factions within one Catholic party. If everyone held to all three of his opinions strongly, compromise would be impossible. Each question would involve a straight fight between a rigid majority and a rigid minority. And a majority that held together on one question would collapse completely on the next.41
It is quite possible, however, that one of the three questions in Belgium might be more salient than the others. In the simplest case, the same question would be salient to everyone, and therefore the other two would be less important to everyone. Suppose, for example, that everyone in Belgium believed that language was the overarching problem. In this case, the leaders of the Walloon community might profitably search for support among both socialists and pro-property Walloons, among both anticlerical and Catholic Walloons; and so, too, with the Flemish leaders. Hence while conflict over language would doubtless grow more intense, the possibilities of conciliacion on the other questions would increase,
since everyone would have some incentive Co reduce conflict on these questions.
But there is sdil another possibility: although everyone might believe that one of the three issues was the most important, different individuals might rank the importance of che issues in different ways. To one citizen,
40. This is equivalent to "the superposition of dualisms" of Maurice Duverger, Lei Partis folttsques (Paris, Colin, 1954), pp. 260-65.
41. Discussions of the effects of cross-cutting cleavages in mitigating conflict usually ignore this possibility. The usual assumption, at lease in recent American political science, seems to be that cross-cutting cleavages inevitably have a unifying rather than a disintegrating effect. But the analysis here indicates that unifying effects cannot occur if all the cleavages are felt with equal intensity- Conciliation is encouraged by
cross-cutting cleavages only H some cleavages arc less significant than others (as is shown in the paragraphs that follow),

the language question would be dominant; to another, Catholicism; to a third, socialism. When opinions are held in this fashion, the possibility of building coalitions of different opinion-clusters is evidently considerably increased. For example, it would now be easier to unite strongly pro-labor Walloons to whom anticlericalism is of secondary importance with fervent Catholic-Flemish workers who might otherwise support the Christian Social Party. Thus there would be a tendency for each opinion-cluster to yield somewhat on secondary issues in order to gain support on primary questions. Hence heterogeneous coalitions or parries might develop; and party leaders would exert great efforts in trying to conciliate their diverse followings.
The extent to which a nation with strong cleavages m^iy be held together in part by its cleavages is perhaps best illustrated by the Netherlands. Abstractly considered, the Netherlands would surely be counted a prime candidate for political disintegration, since the country must bear not only some class antagonism but an even more profound cleavage over religion. Yet the very way in which the one conflict is superimposed on the other probably helps to reduce the potential for conflict. A Catholic party, as Hans Daalder shows, unties both working- and middle-class Catholics. The Protestant parties also cross class lines for their support. Although the Socialist and Liberal parties have had only modest success in gaining support among Catholics and orthodox Protestants, they do nonetheless persist in striving for votes among these religious groups and the liberal Protestants; consequently, less because of their actual support than because of the support they hope to acquire, Socialists and Liberals are a moderating influence on religious cleavages. Taken together, then, the efforts of the parties help to knit the community together.
As the Dutch example suggests, the effects of low coincidence cannot be explained simply by the salience of the questions among the various opinion-clusters, but also by the way in which political institutions encourage people in the different opinion-clusters to unite in coalitions. The most important of these institutions is, no doubt, the political party. If we assume that the number of important parties is not solely a function of the various patterns of opinion we have been describing but also of election systems, historical events, tradition, and the weight of institutions, then the effects of low coincidence of opinions (and hence the existence of a number of different opinion-clusters) will depend in part on the number of parties.
The matter, it must be admitted, is somewhat unclear. However, there is some reason for thinking that where political opinions do not coincide and are unequally salient, a two-party system strengthens incentives for conciliation and reduces incentives for conflict. The reasoning is often applied to the United States, which seems to meet the requirements for

our hypothesis: low coincidence, unequal salience, and two major parties. Given low coincidence, the two parries are bound to be heterogeneous- as indeed they are in the United States. There, as we have seen, the coincidence of opinions is too low to permit: homogeneous parties. But these very circumstances create strong incentives for conciliation and compromise. For party leaders have nothing to gain from yielding entirely to one opinion-cluster: They cannot possibly gain majorities in Congress or win presidential elections by that strategy. Hence they must devote unremitting effort to building coalitions of different opinion-clusters not only in order to win elections but also to pass legislation. A major preoccupation of the American politician is therefore to persuade any given opinion-cluster to give way on its secondary demands in order to gain- if only temporarily-the support of other opinion-clusters for its primary demands. Given the unequal salience of opinions on different questions among- different opinion-clusters, politicians stand a fair chance of succeeding in their efforts at conciliation.
How different are the effects of a multiparry system? Given the same conditions-low coincidence and unequal salience-the same process could, presumably, lead to similar results. Moreover, if a government is to receive the support of a majority in parliament, presumably the process must and will take place. Yet it does seem rather likely that the incentives for conciliation as compared with conflict are stronger if there are only two parties. For one thing, with two parties the process of conciliation is necessary not only in the legislature (as in a multiparcy system), but also in nominations and elections. For another, in a two-party system a good deal of the process of negotiation takes place among politicians of the same party rather than among politicians of different parcies; in this case the
incentive to arrive at a viable compromise is strengthened, not weakened, by party loyalties.42
This somewhat abstract discussion of the effects of different factors on incentives for conciliation or conflict is summarized in Table 12.2.
Polarization
It is a common assumption, one by no means confined to Marxists, that urbanization, industrialization, the growth of national economies, and Other features associated with social and economic modernization tend to result in a high degree of political polarization. A contradictory and more
recent hypothesis is that, beyond some range, in Western societies modernization leads to a decrease in political polarization.
42. This is a specific application of a familiar proposition. Drawing on George Simmd, Lewis Coser has formulated a number of propositions about conflict that are relevant to the discussion in this section- Thus "Conflict with another group leads to the mobilization of the energies of group members and hence to increased cohesion of the group." The Functions of Conflict (Glencoe, Free Press, 1956), p. 95.

Factors that strengthen incentives for conciliating conflicting opinions
1. All uni-modal distributions of opinions.
2. High coincidence of uni-modal distributions,
3. Low coincidence, if of unequal salience.
4. Two-party system:
a. If opinions are uni-modal. b. If opinions have low coincidence but unequal salience.
5. Multiparty system, if opinions are distributed bimodally with high coincidence.
To test either hypothesis is, unfortunately, difficult. To begin with, the term "polarization" is used in various ways. Typically, it is intended to refer to the extent to which a population is divided politically into two
antagonistic camps distinguished by differences in socioeconomic characteristics.
Unfortunately, the concept is bristling with difficulties, for it combines at least three different notions that need to be distinguished.
1. It contains the idea of a dimension of bipolarity, dualism, a U-shaped distribution, the extent to which a population is divided into two categories. A measure of dualism might be the proportion of a population contained in the two largest categories.
2. It also contains the idea of a dimension of distance between the two largest categories. The greater the distance the more polarized the society.
3. Finally, it contains the idea of different characteristics with respect
to which bipolarity or dualism might exist. The most important of these are:
a. Social and economic characteristics. Socioeconomic dualism, then, would be a measure of the extent to which a population is divided into two categories by class, status, language, religion, income, wealth, etc, A country where everyone considered himself a member either of the working class or the middle class would be completely duaUstic with respect to these class categories.
b. Psychological, emotional, or affective characteristics. Psychological or affective dualism would require a measure of the extent to which a population is divided into two antagonistic or hostile groups.
c. Political characteristics. Political dualism would be measured by the extent to which a population is divided into groups distinguished by differences in political behavior-ideology, political goals and demands, party memberships or identifications, and voting. In the simplest case, a country

owhere everyone voted either for Party A or Party B would be (by definition) completely dualistic with respect to voting.
Since the three main dimensions are logically independent, and the three types of characteristics are also logically independent of one another, managing the concept of polarization is inordinately difficult.43 A moment's reflection will reveal the vast variety of different patterns that might be regarded as examples of polarization. Indeed, the concept is so complex that it has seemed advisable to avoid the term "polarization" wherever possible in the discussion that follows and to refer instead to some specific patterns. Among a very much larger number of theoretical possibilities, it is useful to imagine four different general situations in which a high degree-'of political dualism might exist.
1. Political dualism unrelated to social or affective dualism. To offer a hypothetical example, this would exist in the United States if everyone identified strongly with either the Democratic or Republican parties, or if everyone adhered either to a liberal or a conservative ideology, or to an isolationist or interventionist foreign policy; if, however, there were little antagonism between the two groups; and if there were little or no
correlation between party loyalties or ideology and occupation, income, religion, etc.
2. Political dualism closely related to a social dualism, but not to affective dualism. This would exist in Britain if all manual workers were socialists and people in nonmanual occupations were conservative; and if there were little antagonism between the two groups.
3. Political duality closely related to affective duality but not to any social dualities. This would exist in France if everyone "were strongly pro-or anti-De Gaulle; if there were very great antagonism between Gauliists
43. See, for example, The American Voter, where the concept of polarization is discussed at some lengch and used fruitfully in exploring American survey data wkh-out, however, escaping some ambiguity (pp. 338 ff.). The authors begin by defining polarization in psychological terms, i.e., in terms of antagonism; "The condition of active disc-ord between social strata [is] status polarization" (p. 338, their italics). "Status polarization, then, refers to the degree to which upper and lower status groups in a society have taken up mutually antagonistic value positions" (p. 339). They then go on to suggest that status polarization (or, in terms of the distinction above, psychological dualism) is a function of a certain kind of social dualism, namely with respect to social class: "We can say in a rough way that variation in the status polarization of a society reflects variation in the intensity and extent of class identification among its members. When polarization is high most of the citizenry must have perceived a conflict: of interests between strata and have taken on class identifications with fair intensity. When polarization is low, either few people are identifying or extant identifications are weak, or both" (p. 339, their kalics). The measures they use do not, however, bear directly on the question of antagonism; for they measure "status polarization" by using indices of occupation in some cases (Figure 13-2, p. 347, and Figure 13-6, p. 358), and "subjective status" or class identification in others (Figures 13-3, p. 352-, 13-4, p. 354; 13-7, p. 364; 13-8, p. 367). Thus while they define "scatus polarization" in psychological rcrms, they do not in fact measure antagonism. A.
Canipbcll, P. Converse, W. Miller, and D. Stokes, The America?! Voter (New York, Wiley, 1960).

Some Explanations
and anti-Gaullists; and if these differences were not correlated with sociil status, income, occupation, region, etc.
4. Political dualism closely related to both social and psychological dualisms. This is full-scale political polarization. It would exist in France if, in the example just given, the last condition did not hold and instc;ul the middle classes and farmers were uniformly Gaullist while the workin"-classes were uniformly anti-Gaullist. ;
It is obvious that each of these patterns of political dualism is significantly different from the others. What has been called "status polnri-/^-tion" sometimes refers to the second, sometimes to the last. Political polarization might refer to the third or fourth, but it might also refer to the first. Yet the consequences for oppositions, as well as for government would surely be very different in each case.44 Keeping these various possibilities in mind, what conclusions can we draw from the experiences of the countries examined in this volume? Although the data arc hanllv sufficient for solidly based generalizations, some tentative conclusions arc suggested by the evidence at hand.
i. Objective status differences seem to be on the decline throuc:hout Western Europe. Rising incomes, the redistributive effects of tax and welfare measures, increased consumption, standardization of consumer goods, expansion of white-collar occupations, and increased cduc;iE:i(ni (among other things) all contribute to a blurring of status differences and
44. The notion of political polarization is troubled by still other problems rli^r we cannot pursue here. In particular, there are formidable problems of constructing s;"iii-factory indices and measures. Since each characteristic-social, psychological, pnlicic.il -is a broad category, one problem is to choose a satisfactory atfribmc or sec (if attributes within that category. In discussions of social dualism, the attribiiic must frequently referred to is class or status, for which occupation, income, ciliic.uion, class identification, etc. are used as indicators. But status polarization is only one type of social dualism, though an important one, for a society may also divide aloii[; other lines-religion, ethnic group, region, for example. Indices of psychological pol.irny arc difficult to construct. Gabriel Almond and Sidney Verba in The Civic Culture offer some highly interesting possibilities; how one political group views another-positive, negative, and neutral qualities (Tables 1-6, pp. 125 fr.); and psychological disi;mcc between parties as indicated by willingness to have one's son or daughrcr nurry across party lines (Tables 7-11, pp. 151 ff.). Political duality is also broad; indicators might include party identifications or orientation to candidates, though these arc all coiii[)li-caced by simple differences in the number of alternatives presented to voters in two-party and rnuttiparty countries; by differences in policy preferences, ideology, and
the enlargement of middle strata -who shade off into the skilled working class on the one side and the wealthy upper strata on the other,45
2. Political duality as reflected in political attitudes, aspirations, and loyalties varies a good deal among our countries; but once sharp differences in political attitudes also seem to be declining. Nearly all the essays in the volume emphasize the decline of socialism as a distinctive ideology;
as Otto Kirchheimer has indicated in his essay, the change may have been greatest-it is, at any rate, the most remarkable-in Germany, the home of the classic European Social Democratic Party. Strong loyalties to different political parties do, of course, persist; but as the experience of the United States demonstrates, party loyalty may be strong and yet have little to do with differences in ideology or policy. "While the United States is no doubt the extreme case, the essays in this book suggest that European democracies are moving toward a somewhat similar situation.
3. It would be hasty to conclude, however, that the same pattern is emerging in the United States and the European democracies: that is, the first pattern suggested a moment ago, where political duality, even where it exists, is unrelated to social differences or psychological antagonism. There are two difficulties with this hypothesis.
First, the relation between status or occupation and party preference does not seem to have declined in Europe. Indeed, Lipset concludes:
A comparative look at the pattern of working-class voting in contemporary Europe reveals that with the exception of Holland and Germany. the leftist parties secure about two-thirds or more of the working-class vote, a much higher percentage than during the depression of the ip30's . . . The leftist working-class parties have increased their strength in most of Europe during the i96o's. It is clear, therefore, that the easy assumption made by many, concerning American as well as European politics, that greater national affluence would mean a weakening of electoral support for the left is simply wrong.48
Second, the English-speaking and European democracies exhibit no single pattern- Sizable differences occur from one country to another in the extent to which status or occupation is related to party preference. Alford, who applied an Index of Class Voting to voting surveys taken
from 1952 to 1962 in the four major English-speaking democracies found that:
Class voting is almost always above zero; only one Canadian survey falls below that mark. Great Britain is consistently higher than Australia
4<T. See particularly the essay by Orto Kircheimer, above. And the essays by R. Bendix, R. Dahrendorf, M. Crozier, D. Locfcwood, and Raymond Aron in European Journal of Sociology, i. No. 2 (ip6o), 181-182.
46. S. M. Lipset, "Class Conflict and Contemporary European Society, Daedalus" in
a TI /"' -. i. ' ' -.-. " o- --

in the 1952-62 period . . . Australia is consistently higher than the United States . . . The United States is consistently higher than Canada except for one 1948 Canadian survey . . . Canada always has the lowest level of class voting with the single exception mentioned.47
Moreover, there is no consistent pattern of change. In Britain, as Alford shows, the "electorate remains sharply divided along class lines. No decline of the differences between classes as such or of the political predispositions connected with occupational status has occurred." On the other hand in Norway, as Rokkan has demonstrated, an interesting change has occurred since the i89os, for the Labor Party has gradually increased its support among the lower manual strata at the expense of all other parties, and in this sense political behavior has become more closely related to social cleavage. But the Labor Party also receives a good deal of support from the middle and upper strata. In tills respect, the pattern of incomplete polarization is the reverse of that in Britain.48
4. Yet if status polarization is not decreasing, most of our essays indicate that psychological dualism in political life-i.e., antagonism and hostility -has sharply declined since the ilnterwar period. In this case, too, differences remain. The presence of large Communist parties in France and Italy produces (and doubtless is produced by) a higher level of antagonism between a large proportion of the population (and of political elites) than is true in the other countries.48 And in the United States the level of hostility over racial issues obviously has remained high, although the antagonism has historically been sharpest between the Southern and Northern wings of the same party.
5. None of the countries examined in this book closely approaches die pattern described a moment ago as full-scale political polarization, where sharp political, socioecouomic, and psychological dualisms all coincide. In Italy, political antagonisms are probably higher than in any of the other countries; indeed, it is difficult to see how its parliamentary system could withstand much higher psychological voltages. But even in Italy, as Barncs points out, socioeconomic status is quite imperfectly correlated with political cleavages, and the political groups around the center have been able, so far, to prevent political life from turning into a straight conflict between Communists on the Left and anti-Communists on the Right. In
47. "The index of class voting was computed by subtracting the percentage of non-manual workers voting for 'Left' parties from the percentage of manual workers voting for 'Left' parties. For Great Britain, the Labor party was used; for Australia, the Australian Labor party; for the United Stares, the Democratic party; for Canada, the CCF (or NDP) and Liberal parties" (p. 102, Table 5-2).
48. Above, pp. 89 ff. See also Stein Rokkan, "Geography. Religion and Social Class: Cross-Cutting Cleavages in Norwegian Politics," S. M. Lipset and S. Rokkan, eds., Party Systems and Voter Aitgmnents (New York, Free Press, 1965).
49. Almond and Verba report that "Italian respondents show a far sharper (psychological) polarization between right and left: than do the Americans, British and

Britain, S\veden, and Norway, where socialist and nonsocialist parties reflect differences in occupations, incomes, and status a good deal more strongly than the two major parties in the United States, the effects of socioeconomic differences on political life are, as we have seen, greatly attenuated because one of the largest parties in each of these countries has a socially heterogeneous following: thus in Norway the socialists draw as much as 40 per cent of the votes of the nonmanual workers, while in Britain a third or more of the manual workers support the Conservatives, and a quarter or more of the voters in nonmanual occupations have supported Labor. In these countries, moreover, psychological antagonisms appear to be exceedingly low.
How can we account for the absence of full-scale social-psychological polarization in these countries?
First there are the various factors, frequently mentioned in these essays, that are associated with economic "modernity," the growth of national communications, the blurring of class lines, etc.
Second, in some''countries the effects of cross-cutting social characteristics have probably reduced both social and psychological dualities in political life.
Third, and paradoxically, it is in these countries, where political dualism along status lines has been most evident, that a long-run reduction in antagonisms and hostilities has been most notable. Political differences are most likely to reflect status differences, as we have seen, in highly homogeneous countries like Britain and Sweden where other differences- religious, ethnic, geographical-are comparatively slight. Yet this very homogeneity facilitates peaceful evolutionary reforms that blunt the edge of class hostilities.
Finally, it must always be kept in mind that our Western democracies, like other nation-states, are the products of an historical process during which extreme polarization has worked its drastic cures. In this book we have studied the survivors; we might have studied the casualties; we might, conceivably, be examining future casualties. Full-scale polarization is obviously an unstable condition for any polity, and particularly for a representative one. Extreme polarization of political life will be reduced somehow. Perhaps the strongest polarizing forces in modem times have been those of religion, language, and ethnic identity. If these forces do not polarize political life in most Western democracies today, this is only because they did so earlier. Typically they worked their drastic cure by destroying the political system; either by a separation into different countries, so that the cleavages became international rather than national, a5 with the separation of Ireland from Britain or Norway from Sweden; or by the creation of a one-party authoritarian or totalitarian states, as in the case of Fascist Italy, Nazi Germany, and Franco Spain.

To one who believes in the essential worth of a democratic polity, how much opposition is desirable, and what kinds? What is the best balance between consensus and dissent? Even among democrats there is not much
agreement on the answers to these questions.
It is easy to see "why. These questions seem to demand nothing less than a complicated assessment of democracy itself. Or to put the matter more precisely, one can Judge the desirability of different patterns of political opposition only by employing a number of different criteria that would be used if one were appraising the extent to which a political system as a whole achieves what are usually considered democratic goals or values.
Eight of these standards seem directly relevant in fudging different patterns of opposition. In comparison with other possible arrangements, one might ask, to what extent does a particular pattern maximize:
1. Liberty of thought and expression, including opportunities for dissenting minorities to make their views known to other citizens and
policy-makers?
2. Opportunities for citizens to participate in political life?
3. When political conflicts occur, control over the decisions of government by majorities (rather than minorities) of citizens, voters, and
elected officials?
4. Rationality in political discussion and decision-making, in the sense of increasing understanding by citizens and leaders of the goals involved and the appropriate means? l
5. Consensus in political discussion and decision-making, in the sense that solutions are sought that will minimize the size, resentments, and coercion of defeated minorities, and will maximixe the numbers of citizens who conclude that their goals have been adequately met by the solution adopted?
i. Cf. Bagehoc, who refers to "one of the mental conditions of Parliamentary Government, by which I do not mean reasoning power, but rather the power of hearing the reasons of others, of comparing them quietly with one's own reasons, and then being guided by the result." The English Constitution, p. 44. See also p. 180.
6. The peaceful management of conflicts and the minimization of political violence?
7. Resolution of urgent policy questions, in the sense that the government directs its attention to any question regarded as urgent and important by a substantial proportion of citizens or leaders, and adopts solutions satisfactory to the largest number of citizens?
8. Widespread confidence in and loyalty to a constitutional and democratic polity?
A number of other criteria might be advanced, but these are enough to give an idea of the magnitude of the problem of evaluation. What is most obvious and most important about these criteria is that, like most standards of performance for complex achievements, they conflict with one another;
if a political system were to maximize one of these ends it would probably do so only at considerable cost to some of the others. Moreover, because different individuals disagree about the relative importance of different goals, they disagree as to what is the best solution in general, and even for a specific situation. How then can one prescribe an optimal balance among competing goals, when the goals are nonquandtative and imprecise, and when one man's optimum may be another man's prison? Nor are these the only sources of disagreement:. The eight criteria conflict with one another; there is a certain tension among them; we cannot maximize one goal beyond some range without sacrificing another goal.
In spite of all these obstacles to finding an "optimal solution," it is possible to clarify some of the costs and gains of different solutions, actual or proposed. Let me start by examining the tension created by wanting- as most good democrats do-freedom, dissent, and consensus.

The first criterion listed above emphasizes opportunities for dissent; and it is no doubt their concern for this goal that explains, in the main, why liberals and radicals have usually been keenly sensitive to problems of political opposition. For to look at any political system from the point of view of an opposition inclines one to stress the virtues of dissent, of opposing. Yet the last criterion in our list emphasizes the virtues of stability; and the penultimate criterion, the importance of resolution and dispatch, avoiding deadlock, paralysis, impotence in government. Sensitivity to these criteria leads one to be concerned with the high costs of unlimited dissent and to stress the importance of consensus, particularly if governments willing to protect dissent are to survive.
There are, we all know, many varieties of freedom. One variety of freedom exists to the extent chat every citizen has opportunities to engage in political activities without severe social and governmental constraints. In all political systems this freedom-let me call it Freedom of Political

Action-is, like other freedoms, limited by government and society vet it is the differences in these limits that distinguish libertarian from authoritarian systems. In libertarian systems (like the ten described in this book) the right to dissent from the views of government-to oppose the government-Is a vital form of Freedom of Political Action. And political oppositions are a crucial expression of this Freedom.
Yet the very existence of dissent and political opposition is a sure sio-n that someone is constrained by government to do or to forbear from doing something that he would like to do and very likely feels he has a moral right, or even an obligation,! to do. To feel politically free because one obeys laws one believes in, to obey a government one approves of. to obey governmental policies one wants or agrees with-here is a second variety of freedom. Since this variety, like the other, bears no accepted name, let me call it Freedom in Political Obligations.
Now if the existence of political opposition is evidence of Freedom of Political Action, it is also a symbol of the Unfreedom in Political Obligations 2 of those opposed to the government:. I expect that some readers will now move a well-known objection. Even citizens who are opposed to the laws enforced by their government may nonetheless yield their implicit consent, provided these laws are adopted by procedures they regard as legitimate; in this sense, their Freedom in Political Obligations is not diminished by their need to obey specific laws to which they object. Let me recognize the force of this familiar argument and put it to one side as irrelevant here. I do so in order to distinguish (i) a polity in which a large and permanent minority accepts the constitutional procedures and arrangements, yet detests the policies of government, which seems to it tyrannical in what it does if not in the way it aces; from (2) a polity in which agreement is so extensive that minorities are microscopic and evanescent, and no one ever feels much injured by the laws he is obliged to obey. In the first case, members of the outvoted minority might accept the obligation to obey the laws because these were adopted according to legitimate constitutional processes, and yet feel constrained to obey laws they hold wrong. In the second case, they would feel no such constraint.
If you will allow me this distinction, it follows, I think, that in a democratic system where Freedom of Political Action is widely enjoyed, the less the dissent, the greater the Freedom in Political Obligations- In fact the only system in which every citizen would be completely free in his Political Obligations would be one in which political consensus was perfect; for no citizen would then feel constrained by government to do something he believed he should not do. The more extreme the dissent permitted, the greater the range of Freedom in Political Action; yet the
2. The notion of "unfreedom" is defined in Felix E. Oppenheim, Dimensions of Freedom (New York, St. Marrin's Press, 1961), Ch. 4, "Unfreedom."

more numerous the extreme dissenters, the greater the number who are (at least temporarily and perhaps indefinitely) Unfrec in their Political Obligations.
Let me try to make these abstractions more concrete by comparing a high-consensus system like Sweden with a low-consensus system like Italy. In a high-consensus system most citizens are only moderately opposed, if at all, to the character and conduct of government; by comparison, in a low-consensus system a great many more people are strongly opposed to the conduct and even the form of the government. The proportion of citizens who feel themselves coerced or constrained by government, and thus Unfree in Political Obligations is, naturally, much larger in the low-consensus systems than in the high-consensus ones.
Yet an extreme dissenter may enjoy more freedom to express his dissent in a low-consensus system like Italy than in a system with considerably more consensus like the United States. For (aside from any other reasons) the very magnitude of extreme dissent in Italy and France limits the extent to which dissent is coerced by social and governmental actions; in the United States, however, where extreme dissent is so small that it can be coerced at less cost, social and governmental constraints are rather powerful. Thus in the United States opportunities for discussing one's views with others, attending meetings, reading newspapers sympathetic to one's cause, Joining in a like-minded party, and voting for like-minded candidates are extensive for most citizens-but not, often, for the extreme dissenter.
Thus a low-consensus country like Italy may actually provide more Freedom of Political Action (to Communists, Monarchists, and Fascists, for example) than a country like the United States where there is considerably higher consensus. Is low consensus a better guarantee of political freedom, then, than high consensus? Hardly, for a low-consensus system greatly increases the amount of Unfreedom in Political Obligations among its citizens. What is more, widespread Unfreedom in Political Obligations is inescapable as long as consensus remains low; for even if the Outs were to displace the Ins, their positions would only be reversed. The Freedom in Political Obligations of the one-time Outs would now be greater; but so would the Unfreedom in Political Obligations of the one-time Ins. Then, too, a low-consensus system is much more likely to impose other costs such as deadlock, political violence, constitutional instability, and destruction of democracy itself.
If, then, the most desirable long-run solution for a low-consensus country would be to increase consensus, surely tlie most desirable long-run solution for a high-consensus country would not be deliberately to foster extreme dissent! An obvious alternative solution would be to reduce the legal obstacles that limit the Freedom of Political Action among dissenters

until they are legally on a par with all other citizens. This is, in fact, the solution adopted in a number of high-consensus countries. In this respect, the United States is a somewhat deviant case: most other stable democracies have not imposed as severe a set of legal and social obstacles to political dissent as exist in the United States,
If freedom of dissent is thought (by most libertarians and democrats) to be a desirable freedom in itself, advocates of libertarian democracy have usually contended, as John Stuart Mill did, that an opportunity for the expression of dissenting opinions is also a necessary (though definitely noc a sufficient) condition for "rational" political action- The citizens of any country, in this view, need dissenters and oppositions in order to act wisely, to explore alternatives, to understand the advantages and disadvantages of different alternatives, to know what they want and how to go about getting it. Yet there is a certain conflict, one not always recognized, between the conditions required for a relatively rational consideration of alternatives, and the existence of extensive dissent or extensive consensus. Where dissent is slight, the alternatives presented by political leaders for consideration among themselves and by the voters are likely to represent relatively small marginal changes. For in a society where nearly everyone is already rather satisfied with the conduct of government, alternatives profoundly opposed to existing government policy are not likely to be generated, proposed, or considered. Changes are likely to come about by paying attention to a relatively small number of marginally different alternatives to existing policies, examining a limited set of possible consequences, comparing the results of whatever changes are made, and making whatever further modifications are suggested by subsequent experience:
in short, by incremental action,
Although incrementalism evidently seems to a great many people a less rational process than comprehensive and deductive approaches, in fact it offers great advantages as a process for relatively rational change.3 The characteristics and effects of existing policies and institutions are more easily, more accurately, and more confidently known than for hypothetical policies and institutions. The effects of small changes are usually much easier to predict than the effects of large changes. Current processes generate information about effects, and since this information can be fed back to policy-makers, changes can be reversed, accelerated, or altered. In practice, moreover, peaceful change is usually highly incremental.
3. Cf. die discussion in R. A. Dahl and C. E. Lindblom, Politics, Economics, and Welfare (New York, Harper, 1953), pp. 81 ff., and the much more highly developed theory in C. E. Lindblom and D. Braybrooke, A Strategy of Decision (Glencoe, Free Press, 1963), Chs. 5 and 6, and C. E. Lindblom, The Intelligence of Democracy
(Glencoe.Free Press, 1965).

Partly for this reason, no doubt, incremental change is the characteristic method of democracies: liquidation of the Kulaks and the Great Leap Forward would not have been carried out by parliamentary governments.
Yet if high-consensus societies can profit from the advantages of incremental change, they run an opposite danger. Where there is little dissent both political leaders and citizens escape the compulsion to wei?h the relative advantages offered by a comprehensive, large-scale change, even when a large-scale change might prove less costly in the long-run than either the status quo or a series of incremental changes. The history of politics is writ large with the results of costly timidities that have produced too little, too late.
'Reflecting that incremental responses have frequently failed to match the magnitude of a challenge, one is tempted to conclude that sharp political conflict, clashing ideologies, and even low consensus are needed for a rational examination of alternatives. Yet the historical record seems to offer little support for this view. For intense conflicts create their own irrationalities, particularly when conflict is fortified by ideology. It is a reasonable hypothesis that the greater the discrepancy between the "on!s of the parties to a conflict, the more that problem-solving and persuasion are likely to give way to bargaining and coercion.4 The true believer does not judiciously appraise the arguments of the infidel. Has the clusl-i "f ideologies in France and Italy provided a more "rational" examination of alternatives than the low-tension conflicts and unideological analyses among Britons and Swedes?
In sum, high-consensus polities are able to give relatively rational consideration to small changes but they are prone to ignore the possible advantages of radical changes in the status quo. Low-consensus polities nuy find it difficult co profit from the advantages of incremental changes; yc( posing radically conflicting alternatives to citizens and leaders is accompanied by the irrationalities of ideological controversy.
Is it possible to have the best of both worlds? A society where disscnc is low enough to encourage a relatively calm and objective appr;iis;il of alternatives, and yet sufficient to make sure that radical alternatives will not be ignored or suppressed? Among our ten countries, if Italy lies at ihc one extreme, the United States is ac the other. Possibly some of the north-European democracies come closer to the balance we seek. Yet if their high-consensus endures and increases, will not they, too, suffer the disadvantages of weak dissent?
Dispersion, Concentration, and Majority Rule
Does the two-party system offer a solution? Probably no other cure is so often proposed for the ailments of a sick polity. Does it not solve the
4. James G. March and Herbert A. Simon, Organisations (New York, Wilcy.
rncRI unff

problem of how to balance a large measure of consensus with a satisfactory amount of rational dissent? For cannot one party embody the values of stability and consensus, and the other the values of change and dissent?
The only country where a two-party system of this kind has ever endured in a relatively clear form is Britain, which does, I believe, exhibit many of the virtues claimed for the two-party system. Should it be more widely copied?
Unfortunately, the two-party system presents two imposing difficulties as a general solution. It is evidently not viable in many countries. And even if it were, it would not in all circumstances produce the results found in Britain.
The very rarity of the two-party system, a fact heavily stressed in an earlier chapter, argues that the existence of such a system requires an unusual combination of circumstances. The absence of one or more of these circumstances greatly reduces the likelihood that a two-party system will exist, or, if it does exist, that it will have the results expected of it.
The conditions under which a two-party system would provide an optimal solution for meeting the eight goals listed earlier probably include these:
1. The parliament is sovereign in law and in fact.
2. Within parliament, the principle of majority rule is applied to decisions.
3. Among citizens, and markedly among political activists, there exists a high degree of consensus on the desirability and legitimacy of the first two conditions; on the other characteristics of the constitiAkmal system;
on the rights, liberties, and duties of individuals and groups, and on a great many social and economic goals, institutions, and arrangements.
4. On all questions about which there are conflicting views, and for which governmental action is regarded by some people as desirable, most citizens divide into only two great clusters of opinion. These opinions, though stable for long periods, are not rigidly fixed but change with time, as does the size of the two opinion-clusters.
5. There are two political parties, and neither of these parties is, or expects to be, indefinitely out of office.
In these circumstances, two unified parties, each having a program and policies directed toward one of the two great clusters of opinion and competing actively for office, meet a great many of the criteria listed earlier. As long as most conflicting opinions fall into one of the two great clusters, the two-party system would provide an outlet for expressing views, including criticizing the government (Criterion i), and opportunities for citizens to participate in political life (Criterion 2). Any fair system for apportioning parliamentary seats would insure that the government would represent the larger opinion-cluster rather than the smaller (Criterion 3). The existence of two divergent sets of public attitudes bounded bv ex-

tensive consensus would facilitate rationality in political discussion and decision-making (Criterion 4), by insuring that alternatives would be posed, providing a reasonably clear choice to voters, and enabling a rather high degree of coherence of policies and programs. Rationality would also be enhanced because an opposition's past experience in office and its expectation of future responsibility as the governing party would encourage its members to avoid demagogic and irresponsible appeals for unworkable and unrealistic solutions. The existence of widespread consensus and the experience and expectation of governing would help to minimize the resentments of the opposition and the need for coercion (Criterion 5) and also to insure that conflicts would be peacefully resolved (Criterion 6), Policy questions uppermost in the minds of any large group of people would almost certainly be brought forward by one of the two panics, and in due time each party would have an opportunity to enact its own solution (Criterion 7). Finally, all these conditions taken together would
surely go very far toward creating widespread loyalty to democracy ;uid constitutional government (Criterion 8).
It is easy to see why the two-party model, especially in the idcali/.cJ form in which it is often described, has charmed so many political observers. Yet the conditions I have just specified are an unusual combin;i-rion; they have not always existed even in Britain.
In the first place, as we have already seen (Chapter 12), if a society is polarized into highly antagonistic camps, then the two-party system might
actually increase the intensity of conflicts by wiping out the nicdi;iuiiy center.
Second, opinions may fall into more than two clusters, as they gcncf.itly do in Belgium, Holland, Italy, and the United States. In these circumstances, there would have to be more than two parties; or, as in the United States, the two parties would not be highly unified. Moreover, if a system of disciplined parties existed, ic could produce flagrant conti'-ulic-tions with several of our criteria. Specifically, the application of the prin1 ciple of majority rule and parliamentary sovereignty could lead to minority government, negate majority rule, and thus violate Criterion 3. For if " faction in one party, even if it were a majority faction, coulil use the instruments of party discipline to impose its policies on that party, anil *l that party had a majority in parliament, then the policies adopted by tht government and agreed to by the parliamentary majority might we" "< policies preferred only by a minority and opposed by a majority 01 W whole country. When there are more than two large clusters of opinion* it would be necessary, in order to satisfy our criterion, for different fWr' Jority coalitions to form on different issues. Thus a mulriparry ^^ w two heterogeneous parties without strong discipline would be preicwrtft Third, the government and the opposition parties might not alicnut?""

omce. In this case, an opposition might find demagogic and unrealistic appeals increasingly attractive. Even if, as Nils Stjernquisc's essay shows. the problem of a permanent opposition is not at all peculiar to a two-party system, there is nothing inherent in the dynamics of a two-party system that guarantees an alternation between the two parties.
A system with two disciplined and strictly competitive parties, one controlling the government and the other providing a concentrated focus for opposition, is not always, therefore, a desirable solution. The circumstances under which it is the optimal solution may be, in fact, rather uncommon. The typical solution of democracies is not concentration but dispersion, not strict competition but bargaining and coalescent strategies. Given the conditions of political life in most countries, quite possibly this solution is preferable; for it is often possible where the other solution is nor, and it may come somewhat closer to satisfying our various criteria,
Majority Rule, Minorities, and Organized Pluralism
Every solution to the problems of opposition that focuses upon party systems runs the danger of neglecting a palpable fact of political life:
iimny important decisions are not made in parliament. To the extent that :m opposition concentrates on elections and parliamentary action^ it may
be powerful in unimportant encounters and feeble or even absent when key decisions are made.
As the extent of governmental intervention and control in social and economic affairs has expanded, the work of parliaments has also multiplied. But even with the enormously increased work load of parliaments .Hid the much greater weight and range of the effects that laws passed by parliament now have on social and economic behavior, the relative im-I'on.mce of parliaments in making important decisions has not increased
in tlic same proportion; the "decline of parliaments" has become a familiar topic and a source of concern.
No single curve could summarize the historical changes in the power tif rurious parliaments. But in a number of countries two kinds of developments have helped to increase the relative importance of other sites. One 's a pronounced growth in many Western democracies in the power of a plebiscitary executive who acquires great political resources by winning a "-"'onal election. Although this development is clearest Jn the United Sr-ircs and the Fifth French Republic, the rise or highly disciplined parties 'us led by a different route to similar results in Britain, Norway, Sweden, Austria, and a number of other democracies. In France and the United o"oires, die Constitution, Jaws, and political practices grant extensive dis-crctJonary power and authority to an elected chief executive; in the other "untnes, if a party or coalition wins a majority of seats in parliament, ^"V d'scipline insures that it will form a government whose policies

cannot:, for all practical purposes, be defeated by opponents in parliament:. Although the development is highly uneven and the pattern is markedly different from one country to another, the importance of the legislature as a sice for encounters between opposition and government is reduced to the extent that a plebiscitary executive (whether president or cabinet) has acquired the power to make key decisions without much restraint by parliament.
The other development that creates a powerful rival for parliament is the one stressed by Stein Rokkan in his chapter on Norway: the evolution of national bargaining among employers, trade unions, and other interests has led to a process for making decisions of great economic and social importance over which parliaments and sometimes even executives exercise scant control.
What these two developments have in common is the creation of highly strategic sites outside parliament-rivals to parliamentary power- Where they differ is in the importance of national elections. For if the rise of a plebiscitary executive has reduced the relative influence of parliament, that development has, if anything, made elections even more crucial. Yet because the concrete alternatives open to voters are few and simple in comparison with the great range of problems confronting a modern scntc, an election furnishes a vague mandate at best; and at worst it simply allows
winners and losers to provide their own interpretations of the election returns.
Bargaining with the executive, bargaining among private and public bureaucracies, negotiations among the great national associations, all these provide ways of supplementing, interpreting, offsetting, and even ncg.itni^ the election returns. This is exactly the source of both the advantsigc.s' aiitl the dangers of organized pluralism and national bargaining.
Organized pluralism meets many of the criteria for opposition in n democracy that I proposed at the beginning of this chapter. For cx;unpk, because it often enables key groups to arrive at decisions they find more acceptable than decisions imposed by legislature or executive order, li-ir* gaining is an instrument for gaining consensus and enlarging the iirca w Freedom in Political Obligations. It provides additional sites for cfl^cuvc expression of views, dissent, criticism. Often it insures that the speciali/w knowledge of the groups most deeply involved in some activity will M brought to bear on a solution. Yet if it has many advantages-nnd it ti t" any case inevitable in every modern libertarian industrial society-orgift* ized pluralism creates two problems that have not yet been solved "nVr where. For one thing, since all resources except the vote nrc ""^R111"/ -^ distributed, some minorities (one thinks of the uneducated poor In W ^ United States) may not have much in the way of political resource*._ T-^ bargain with: they have the ballot-and little else. In addition, (o w^

extent that parliament is excluded from the process and elections provide only a vague and rather uncertain control over national leaders, there is no political institution in which majorities weigh heavily that can control
the great bargained decisions by means of public review, appraisal, opposition, amendment, or veto.
Perhaps organized pluralism would weaken democracy in a small city-srate, as Rousseau and his admirers would argue, for it encourages a citizen to take only a fragmentary view of his interests. And faction has always been the mortal disease of the city-state. Yet in the modern nation-state, it is difficult to imagine an alternative to organized pluralism that would
not leave the piebisckary executive and the official bureaucracies without effective oppositions, criticism, and control,
What is not yet perfectly clear, however, is how organized pluralism and national bargaining are to be reconciled with systems in which political equality and majority rule are major principles of legitimacy.
Oppositions in Western Democracies: The Future
There is a tension, then, among our goals, a tension that seems to be
inescapable. The demands imposed by the values of democracy are extraordinarily severe.
To one who accepts these values, one perennial problem of opposition . is that there is either too much or too little. The revolutionary parts of the world have a surplus of poverty and a deficit of order. The authoritarian countries have a surplus of order and a deficit of political freedom.
Some of the Western democracies are achieving a mounting surplus of riches and consensus.
To be concerned over the decline of structural oppositions in most \Vcstern democracies may very well be an anachronism, a throwback to nineteenth-century styles of thought, on a par with a nonrational faith in the virtues of a balanced budget or the conviction that a seven-day workweek is indispensable if the working classes are not to become dissolute from having too much spare time on their hands. Should we not begin instead to adjust our minds to the notion that in the future-or at least in that short-run future into which it is not wholly senseless to extrapolate present trends-a great many Western democracies will have rather high levels of agreement and not much structural opposition?
That a large number of democracies have won the battle for tlie allegiance of their citizens among all social strata is, surely, a satisfying- vic-^ry to anyone who believes in the values of a democratic polity. Yet it ls difficult to disregard the sense of disquiet that follows hard upon one's au'arcness that severe criticism of social and economic structures has all "Ut disappeared from the political life of many Western democracies- or else has become a monopoly or political forces like the Communists

and the Radical Right: whose allegiance to democratic values is, to say the least, doubtful. If the growth of extreme dissent can endanger a democratic system, universal but quite possibly superficial and irrational "consensus" may also be undesirable, for reasons I have jusc been exploring.
But is the trend evident in so many Western democracies over the past several decades toward greater consensus likely to continue? Might it level off, or even be reversed?
No way of conjecturing about the future, as Bertrand de Jouvenel has remarked, is more compelling than the temptation to project recent trends, "to suppose that tomorrow is going to differ from today in the same way that today differs from yesterday." ° And as M. de Jouvenel also reminds us, it is the extrapolation of recent trends that has so often led men who understood their own time well to miss completely the large changes and abrupt reversals that transform the future into something radically different from the past. Suppose, then, that we make two assumptions about the future; that there will be no holocaust, an assumption without which it would be futile to speculate about the future of politics;
and that Western societies will continue to develop greater affluence, higher consumption, reduction of poverty, wider educational opportunities, and steadily increasing technical and technological resources. Under these conditions, what can we conjecture about the future of oppositions in Western democracies?
To begin with, even the growth of affluence does not automatically wipe out conflicts over the distribution of the national income and opportunities of all kinds. There is no reason to assume a decline in the fainili.ir conflicts among different interest groups, each striving to insure that its members gain a satisfactory share. Indeed, since a just or satisfactory share probably cannot be denned so as to command general assent, ;uul since any particular allotment reveals itself more and more clearly nowadays to be a product of political decisions and less and less an act of Goii. nature, or the inexorable operation of economic laws, conflicts over ihe distribution of income might, if anything, become more numerous even if less intense.
International policies will also remain, surely, subjects of conflict. Since there is small chance that international politics will diminish in impon.uK'c and salience in the next half century, and since judgments of alternative policies and proposals will necessarily rest on highly controversial a&^o-ments of very great risks, gains, and costs, a variety of foreign policies, military affairs, treaties, regional and international organizations and alliances all promise a steady flow of internal conflict.
In many cases, conflicts over international politics and the distribunoft ,.. of national income and opportunities may not generate anything moff -

than the kinds of policy oppositions with which we are already familiar in Western democracies. To this extent the future looks rather like the immediate past that has been described in the essays here. Yet this may not be the whole picture. For these cwo kinds of conflicts need not necessarily entail only narrow group interests or technical matters. The more the distribution of incomes and other opportunities is thought to be subject to determination by government, the more relevant may become the ancient and evidently inextinguishable controversies over the issue of equality versus differential rewards. And conflicts over international affairs will in some cases involve nothing less than alternative views of how the nation, the civilization, even the species itself are most likely to survive or perish.
Yet neither of these kinds of issues inevitably entails structural oppositions. Are there possible sources of alienation in Western democracies that might foster new structural oppositions? Since alienation has lately become a fashionable topic, let me hasten to add that I do not mean to imply anything about social or psychological alienation, whatever these may be. I speak only of political alienation. A citizen is alienated from his political system to the extent that he has unfavorable feelings, evaluations, and attitudes toward it. I assume that a citizen might be alienated from the political system in which he lives without necessarily being neurotic, rootless, excessively anxious over his social standing, or otherwise much different in personality and social characteristics from his fellow citizens. In short, I wish to leave completely open the murky empirical question
of how political alienation may be related, if at all, to strictly social and psychological factors.
Among the possible sources of alienation in Western democracies that:
may generate new forms of structural opposition is the new democratic Leviathan itself. By the democratic Leviathan I mean the very kind of political system the chapters in this book have described, a product of long evolution and hard struggle, welfare-oriented, centralized, bureau-cr.uic, tamed and controlled by competition among highly organized elites, and, in the perspectives of the ordinary citizen, somewhat remote, tlisrani, and impersonal even in small countries like Norway and Sweden. The politics of this new democratic Leviathan, as we have seen so often in the past chapters, are above all the politics of compromise, adjustment, negotiation, bargaining; a politics carried on among professional and quasi-professional leaders who constitute only a small part or the total citizen "ody; a politics that reflects a commitment to the virtues of pragmatism,
moderation, and incremental change; a politics that is un-ideological and <vcn anti-ideological.
The traditional opposition to the new democratic Leviathan has come "om critics on the Right. In most European countries this traditional fpposiEion has been greatly enfeebled. In the United States, I suggested

ill Chapter 2, a Radical Right has become alienated from the c,\istini> political system, for the principal leaders in the American system--whether in the Administration's coalition or nominally in opposition-accept policies, express views, and engage in conduct that the Right passionately rejects as evil; being a minority and unable to win national elections (and not many state or local elections) the Right has steadily suffered the humiliation of political impotence and rejection.
Is it likely that the wheel of history may make a full turn, that opposition to the democratic Leviathan may arise from a new quarter? There are already faint signs, not only in the United States but in high-consensus European systems like Sweden, Norway, and Britain, that many youn^ people, intellectuals, and academics reject the democratic Leviathan-not because it is democratic but because, in their view, it is not democratic enough: this new Leviathan is too remote and bureaucratized, too addicted to bargaining and compromise, too much an instrument of political elites and technicians with whom they feel slight identification. Political isolation, alienation, and rebellion among youth, intellectuals, and academics are not, of course, new. Yet in the past half century the Left has, on the whole, sought to channel these feelings into support for policies and programs that have encouraged, not retarded, the development of the new democratic Leviathan. Is it not possible, however, that political alienation will increase, and that a new Left-if one can stretch traditional terms to cover the case-might channel these feelings into radical efforts (the shape of which we cannot foresee) to reconstruct the Leviathan to a more
nearly human scale? None of the three possible sources of future oppositions that I have
been describing can be reduced, I imagine, to strictly technical questions-Although most issues involving the distribution of incomes and opportunities, international politics, and the democratic Leviathan have strictly technical aspects, few of them can be settled by strictly technical answers. The position one takes must depend in part on nontechnical factors -on values more implicit than explicit, psychological orientations and predispositions, identifications, feelings of hate, hostility, fear, Jealousy, pride, self-confidence, respect, solidarity. If factors of this kind are to play a part, then there is good reason for expecting that political conflict will encourage the birth of new ideologies. For political elites and involved citizens alike will sense a need for broad, integrated views of the world to provide guidance, validity, and authority for their judgments on specific issues. And it will be surprising if these views of the world do not differ considerably in their perspectives, goals, evaluations, and assumptions about the nature of man and society.
Yet should these conflicting issues and ideologies develop it seems unlikely that they will be strongly associated, at least in the democracies

examined in this book, with the familiar social and economic characteristics that have done such yeoman service in social theory in the recent past. For differences in political demands will probably become in large measure detached from general socioeconomic factors. At one end of the scale, new oppositions may reflect no more than conflicts among shifting coalitions of interest groups. At the other extreme, conflicting demands are likely to be attached to relatively durable orientations, perspectives, "mentalities," political philosophies, or ideologies that are related only casually to the kinds of social forces that have played such an important part in the political life of "Western democracies in the past century.
In short, differences in basic political ideas and evaluations are likely to become more and more important in explaining differences in political behavior and therefore in patterns of opposition. Yet these crucial differences in political ideas and evaluations will probably be less and less trace-able to differences in social and economic characteristics. In this sense,
political ideologies, far from waning, will be ascendant. To be sure, the traditional ideologies that have played so great a role in
Western politics in the past century show every sign of being well on the way to ultimate extinction. But democracies have not eliminated all causes of political conflict; and if we agree with James Madison that "the latent causes of faction are sown in the nature of man," then democracies will not and cannot eliminate all causes of political conflict. If democracies cannot eliminate all the causes of conflict, is it not reasonable to expect that with the passage of time the clash of governments and oppositions, indeed of one opposition with another, will generate-and will be generated by-new political perspectives that we cannot now accurately
foresee?

• [ На главную ]
• [ Библиотека ]
• [ Reader ]
• [ Программы ]
• [ Абитуриенту ]
• [ Фотоальбом ]
• [ О ФП ]
• [ Гостевая книга ]
• [ Ссылки ]
• [ Обратная связь ]