University of Bergen

I have been trying, over the last few years, to devise parsimonious models for the explanation of the many striking differences in the style and structure of mass politics within Europe. In this brief paper I shall not present yet another model but shall, instead, offer some reflections on the character of this enterprise and suggest a few next steps.

As set out in several different contexts, the ambitious aim of this effort of
model-building is to account for variations in two sets of parallel processes of
change in European political systems since, say, 1789:

first, the sequences of steps in the institutionalization of formal mass democracy: the establishment of safeguards for the freedom of organized competition, the broadening of the franchise, the standardization of secret voting, the lowering of the thresholds of representation, the introduction of various measures of parliamentary control over the national executive;

secondly, the timing of the growth and the stabilization of organizations for the mobilization of mass support through these new channels, the formation and the "freezing" of organized party alternatives within each national political system.

Thus far, my efforts of model-building have focused on each set of variations in turn: I first drew up a paradigm for the study of processes of democratizatio ¹ and then proceeded, quite separately, to construct a schema for the explanation of variations in party alternatives.²

In an initial round of comparisons this seemed a sensible procedure: I was

---

* This article was originally prepared as a paper for the Round Table Conference of the International Political Science Association in Turin in September 1969. The article grows out of studies carried out within the cooperative project on "The Politics of the Smaller European Democracies" financed by the Ford Foundation and various national sources. Part of the article will be incorporated in a collective volume by Hans Daalder, Robert A. Dahl, Val Lorwin and Stein Rokkan scheduled for publication in 1971 by Stanford University Press. In revising the Turin paper for publication in Scandinavian Political Studies, I have relied heavily on criticisms and suggestions by Alain Lancelot, Val Lorwin and Richard Rose: in fact my comparison between Swedish Finland and Ulster is largely inspired by Richard Rose's pioneering study of the politics of the six counties: The Authority of Regimes: An Irish Perspective, London, Faeber, forthcoming 1971; cf. his "The Dynamics of a Divided Regime", Gov't and Oppos. 5 (2), 1970, pp. 166—192.

able to produce a set of neat historical typologies and was pleased to see how often some of the distinctions helped to account for variations in the style of mass politics.3 But this was clearly only a half-way house: the real challenge lay in the construction of a unified model for the interaction between the two processes of change.

The initial models were developmental in the sense that they generated propositions about irreversibilities in sequences of socio-economic preconditions, elite options, and outcomes for the scope and structuring of mass politics. But they were not dynamic models: they predicted to characteristics of the democratization process or to the structure of national party alternatives in the phase of full manhood suffrage, but they were not designed to account for variations in the sequences of interactions between pressures for democratization and efforts to establish organized control of the mobilization markets through the formation of parties. Any such "dynamicization" will of necessity increase the complexity of the models: if we take as our ultimate dependent variable the structure of the political alternatives open to the citizenry, the task of explanation would have to be multiplied by the number of distinct states of democratization in each national political history. In practice, given the crucial importance of the distinction between mobilization efforts within and mobilzaton efforts outside the formally established channels of participation, it will be essential to characterize each point in "political time" both by the level and the formal equality of enfranchisement and by the total level of effective mobilization, whether measured in organizational strength or in electoral turnout rates. The task would no longer be to account for variations in party alternatives at one single phase of democratization/mobilization: the first elections after the achievement of full and equal manhood suffrage. The task would be to account for step-by-step changes in the pedigree of party alternatives from one phase of democratization!mobilization to the next. This is a vastly greater challenge to comparative political analysis: the great question facing the student of comparative development is whether he can immerse himself so thoroughly in the details of national political history and still retain his primary interest in the parsimony of an overarching schema of concepts and propositions.

A Possible Procedure: Paired Comparisons of Developmental Sequences

My own strategy in this quandary is to proceed by paired comparisons of natonal developments: I start out from the distinctions in the original models and then check through for each pair how many of the actual differences in sequences and structures can be roughly accounted for in these terms and how many require further differentiations. The essence of the method is to stay at the level of cross-national comparisons of similarities and differences at every step: to reconcile the need for maximal information about concrete developments with the need for conceptual parsimony in the model of explanation.

I have not yet worked through all the possible pairs in the original set
of 16 "end-state" polities in Western Europe: 11 smaller and 5 larger. In this

paper I shall confine myself to only a few illustrative examples and then proceed
to a sketch of the possible elements of a unified model.

Denmark vs. Sweden

In my model for processes of democratization I contrasted the Danish development with the Swedish and explained the differences against the background of the heritage of protracted absolutism vs. continuous estate representation.⁴ On the other hand, in my model for the generation of party systems, I kept the two older Nordic polities in the same cell: Protestant, strong dissent but no Catholic minority, state closer to urban interests.⁵

When I drew up this simplifying model of party developments, I was obviously aware of the many differences in developments and structures among the Nordic countries but I had been so forcibly struck by the contrasts between the North and the rest of Europe that I gave top priority to the explanation of these broader differences and simply packed the two older of the sovereign states of the North into one single cell in the 2 X 2 X 2 X 2 scheme of differentiation. In fact, judging from the full-suffrage structures of party alternatives in the two countries, there was very little left to explain: both Sweden and Denmark had a Conservative business-oriented party, an Agrarian party, a residual party of cultural opposition (Folkpartiet in Sweden, Radikale in Denmark), a strong Social Democratic party and a weak Communist party. But this parsimonious typology proved highly Procrustean as soon as the analysis proceeded further into the details of party development in the two countries. The contrasts in institutional heritage did not only affect the sequences of decisions on formal democratization but also conditioned the character of the mobilization drives in the two countries and produced quite marked differences in the alignments of the voters behind the party alternatives, at least up to the 1930'5. The crucial difference concerned the steps in the integration of the peasantry into the national political community: in the Swedish case the continued tradition of estate represenation eased the entry of the upper layers of the peasantry into the political community and quickly domesticated the first party of rural opposition, Lantmannapartiet ⁶ in the Danish case the rapid transition from absolutist rule to the enfranchisement of the great majority of all men over 30, set the stage for a bitter struggle between the mobilizing peasantry and their allies in the cities against a hardening coalition of Royal officials, patricians and estate owners.⁷ As a result, the two countries followed very different paths in the development of their party systems: in the Danish case the peasant-radical Venstre, established its organizational identity through the struggles in Parliament against the Royal Executive; in the Swedish case there was a bewildering variety of short-term alliances in the two Chambers of the Riksdag, while the first genuine mass organizations established themselves outside the institutional framework, in the fight for franchise reform up to 1906—09.⁸ The result was a marked difference in the initial party structure: in Denmark the old opposition party remained a predominantly rural movement and in fact was transformed into a near-agrarian

party after the split-off of the Radicals in 1905; in Sweden the integrated stratum of the peasantry joined with the urban bourgeoisie in the formation of the initial protectionist-nationalist party, variously called the Moderates and the Right, Hogern, and only later broke out to defend its economic interests in the commodity market through the foundation of an explicitly agrarian party.⁹ In the end the two party systems came to look numerically and structurally alike but the steps in the process differed markedly and some of the underlying voter alignments remained very different for quite some time. To account for such variations in the paths of development we must clearly search for ways of linking up the model for democratization processes with the model for cleavage expression. In practice this means a time-phasing of the dependent variables: we are not just interested in explaining variation at the point of final electoral saturation and "party freezing" but also in tracing changes in political alternatives from one phase of mobilization to the next.

Such paired comparisons could obviously be pursued through the entire gamut of the 16 (16—1)/ combinatorial possibilities in the original set of cases, but only a few of these confrontations have proved disturbing enough to force a reconsideration of the entire scheme of model-building. Typically the differences left unexplained in the cleavage expression model can be accounted for through the institutionalization model. This was pointed out already in the initial statement of the cleavage model in Party Systems and Voter Alignments: the Dutch and the Swiss cases were left in the same cell in the cleavage expression model but could be differentiated through the introduction of such institutional variables as the level of centralization; much in the same way, the differences between the French and the Italian party systems, both in the same basic category in the 2X2X2 schema of successive dichotomies, could be traced back to variables such as centralization and monocephality/polycephality.10 But these were essentially comparisons focused on the end-products of party political developments: the structures of alternatives that emerged during the more stable periods after the full extension of the suffrage. The real challenge lies in the construction and testing of models for the time-phases of. national political development: which cleavages find which sort of organized expression at which points in the mobilization history of each territorial system? To illustrate this style of comparative analysis, let me set out very schematically a contrast between two of the latest units to reach independent status within "pluralist" Europe.

Finland vs. Ireland

In the institutional model Finland and Ireland are bedded down in the same cell: both of them at the periphery of Europe, both of them for centuries subject territories under representative regimes, both grown out of a long struggle for national identity against powerful oppressors, both latecomers to the community of sovereign political systems. In the cleavage model, the two bedfellows fall miles apart: Finland in league with the Protestant countries of the North,

Ireland at the opposite end of the scheme as a thoroughly Catholic country largely untouched by secularizing movements. Both countries went through violent upheavals in the transition to independent nationhood: both were baptised in traumatic civil wars. But the battle-fronts differed markedly in the two cases: in Finland the civil war of 1918 was essentially an expression of deep distrust between social classes, in Ireland the civil war of 1922—23 was exclusively a conflict over the definition of the national territory and the defence of national cultural identity. This contrast in conflict alignments found clearcut expression in the party systems of the two nation-states. In Finland the bitter class conflict in the rural economy had produced a remarkably strong Socialist party at the first full-suffrage election in 1907 and the civil war had not only deepened the cleavage between the propertied and the propertyless, but also added a lasting split between moderate Social Democrats and extremist Communists, for long periods conspiring against the national system in underground organizations ¹¹ By contrast, the Irish struggle for national independence produced a simple two-and-a-half party system: the primary party-forming cleavage centred on the acceptance/rejection of the 1922 Treaty and the boundary between the Republican South and the Unionist counties in the North, while the potentially disruptive conflict over class issues in the economy proved surprisingly weak and produced only small and marginal Labour parties in the Southeast and in Belfast ¹² Marx and Lenin and a host of other international revolutionaries had predicted an anti-imperialist class uprising in Ireland:¹³ the long history of economic grievances against an alien class of owners and employers, the stubborn tradition of conspiratorial politics, the openness to ideological influences from abroad, all these factors seemed to offer great opportunities for a revolutionary working class movement. In actual fact, the struggle over the strategy of national independence was to divide the Irish citizenry much more profoundly than the class issues: "proletarian socialism was distracted by the pull of nationalism, and whenever a choice had to be made inside this framework, socialism invariably lost."¹⁴

How are we to account for this contrast between these two latecomers on the
European political scene?

David Thornley has focused his analysis of Irish developments on the timing of British responses to the three great cries for social revenge against the Establishment in Ireland: the cry for justice for the Catholic religion, the cry for land, and the cry for cultural and political independence. If the rulers had had to face all the three claims at once the result would have been extreme polarization and very likely a disastrous revolutionary war. As it happened, the three challenges could be met one after the other: the peculiarities of Irish liberation history can essentially be explained through the time-phasing of mass mobilization and elite response. The mobilization of the Catholic bourgeoisie and the politicization of the clergy under O'Connell brought forth a stream of concessions: the British rulers not only conceded full professional and political privileges to the Catholics but also abandoned their control of the Church tithe and

of primary education. In a further phase, the tactical alliance between the Land League and the Irish Parliamentary party brought forth a rash of concessions on tenancy conditions and the sale of holdings and changed the entire structure of the Irish rural economy, over less than thirty years, from one highly concentrated in largely absentee-owned estates to a highly dispersed system of small peasant properties. The peasants remained poor and socially underprivileged but they owned their land: this proved decisive in the final phase of the process of liberation, the struggle over national independence from 1916 to 1922. "The folkmyth of the nineteenth century is the struggle of the Gael to regain the land; the reality of its conclusion was a small farm pattern which . . . could be sustained only by the safety-valve of emigration and by subsidization . . . This uneconomic but atavistic clinging to peasant ownership is perhaps the greatest political influence bequeathed by the nineteenth century to the twentieth, perhaps more decisive even than any issue of nationality or religion. The achievement of successive English reform administrations, from Gladstone's to Balfour's, was to convert the great bulk of the Irish peasantry from social revolution to social conservatism. In so doing they largely dictated that the pattern of the final, national, revolution should be socially conservative."¹⁵

The initial conditions in Finland were in many ways similar to the Irish: the Swedish bridgehead in the Sounthwest corresponded to the English-Scottish in Ulster and the Pale, a Swedish administrative, professional and cultural elite confronted a Finnish peasantry much in the way the English confronted the Irish, there were considerable concentrations of land and commercial-industrial capital in Swedish and other alien hands, at least partially comparable to the concentrations in English hands in Ireland. But there were also important differences:

first of all, the initial mobilizing cleavage within Finland was linguistic, not religious: the Fennomans fought for their cultural identity against the Swedish "ascendancy"; the first Irish mass movement gave clear-cut primacy to the claim for religious rights and in fact retarded the awakening of linguistic consciousness ₁₆

secondly, the Swedish bridgehead in the Southwest was culturally unified while the colonial settlements in the Northeast of Ireland were deeply divided ethnically as well as religiously: there were, to be sure, distinctive class differences between the established elite families and the farmers and fishermen of Swedish Finland but there were no historically rooted ethnic-religious divisions similar to the ones separating the Presbyterian Scots from the Anglican landowning families;

thirdly, there was no Finnish equivalent to the Irish Famine and there was
very little overseas emigration to slow down the growth of the rural proletariat
before the onset of mass politics;

and finally, the sudden change-over in 1809 from provincial status within
Sweden to dependent status within the Russian Empire not only created a double
front in the Finnish fight for cultural identity but also blocked the possibilities

of effective legislative action to take the brunt off the many social and economic inequalities that divided the national community; by contrast, the Irish remained throughout their liberation struggle dependent on the reactions in one political centre, Westminster, and in fact were able to exploit the great economic resources of this centre in their efforts to break up the large landed estates and to achieve greater protection for peasant property.

These differences in cleavage structure and international context go far to
explain the differences in the mobilization processes in the two countries.

In Finland the initial opposition centred on the issue of language rights but the system was never fully polarized on this issue. The threat of Russian domination created an additional cleavage line and brought the constitutionalist Young Finns into an alliance with the Svecomans against the cautiously compliant Old Finns: in fact the pressure from the East increased the willingness of the Swedish elite to yield on the rights of the Finnish language and led them to agree in a series of measures of equalization before the great surge of mass mobilization in 1907.

By contrast, in the Irish liberation process the issue of the national language came up only in the last phase of the struggle and even then was overshadowed by the purely political fight for territorial sovereignty. The Church had given priority to its own cultural and economic claims, and the Famine, the waves of emigration to America and the institutionalization of seasonal movements of workers across the Irish Sea had gradually undermined the position of the native Gaelic language and made English de facto if not de jure the dominant medium of communication within the independent Republic. Finland had been Christianized from Sweden and had been converted to Lutheran Protestantism with as little resistance as any of the other provinces of the Kingdom. Ireland was a stronghold of Catholic Christianity long before England and had proved highly resistant to all efforts of conversion under Henry VIII, under Cromwell and under William of Orange. The Finns mobilized to defend their language and their inherited culture against the Swedish ascendancy but found a common ground with their masters in the Lutheran religion: in fact, the Church was quick to open up channels of recruitment from the ranks of the Finnish peasantry and soon became an important agency of national mobilization and integration. In Ireland there was no such bridge-building: after the defeat of the Rebellion of 1798 and the consolidation of the Orangeist alliance of the Protestant proletariat and peasantry with the English gentry elite, the counter-mobilization was spearheaded by the Catholic bourgeoisie under O'Connell and the fight for Irish independence was defined as a crusade for the restoration of the rights of the Ancient Church.

This process of ethnic-religious polarization proved fateful for Ireland: it not only defined the situation for the negotiators who reached the territorial compromise of 1921, it also froze the conflict structure within the Northern enclave and determined the alliances in the next phase of violent confrontation, the riots in Derry and Belfast from October 1968, onwards. The Orange-Catholic

polarization made it practically impossible to organize politics on class lines in Ireland: the Land Acts and the remittances from the overseas emigrants took the brunt off the grievances of the rural proletariat and the only solid basis for an industrial labour movement was, ironically, found in the centre of the Northern enclave, in Belfast.¹⁷

This development contrasts dramatically with the Finnish: there the cultural polarization has lost in intensity and left the field open for a polarization on class lines. The Reformation settlement had left few opportunities for religious revolt and the gradual Swedish concessions on the linguistic front had softened the initial oppositions in the system. But nothing of any consequence had been done on the economic and the social fronts before the sudden surge of mass mobilization after 1906. The result was a "revolution of rising frustrations"; the swelling rural proletariat of crofters and landless labourers and the rising industrial working class were given the vote in 1906 and mobilized so quickly that they could give the Socialists a clear majority in the Diet only ten years later. But this great electoral victory did not lead on to effective action: the Socialist efforts to introduce reforms in the structure of the rural economy were thwarted again and again, first by the Czar, later, and this proved fatal, through the decision of the Kerensky government to dissolve the Diet and call new elections in July, 1917.¹⁸ This was the immediate background of the strike in November and the Socialist coup in January, 1918: the chasm between the "Reds" and the "Whites" in Finland had deepened to the point of open violence, and the victory of the Bolsheviks in Petrograd and Moscow soon triggered off a savage civil war.

The Finnish civil war differed from the Irish on three counts:

first, in the class character of the alignments — it was not a fight over diverging
strategies in the handling of an alien enclave but a struggle between
social classes within the once subject population;

secondly, in the international context of the alignments — in the Finnish case the underprivileged party sided with a dominant power outside, in the Irish case the defenders of the established elite could count on support from a powerful metropolis;

and thirdly, in the character of the outcome — in Finland full sovereignty without any territorial division, in Ireland a compromise solution guaranteeing independence within the British Commonwealth for 26 of the counties while defining the Northern Six as a territory with limited self-government under the United Kingdom.

In all this the timing of external developments, internal mobilization drives and elite reactions is clearly of crucial importance. To add visual clarity to the comparison of time phases of development it might be useful to contrast the two countries diagrammatically as set out in Fig. 1.

The diagram helps to bring out the contrasts in the time-phasing of the decisive elite reactions and the waves of mass mobilization in our two countries: it helps us to compare sequences in political rather than in purely chronological time.

In the Irish case the concessions to the Catholic bourgeoisie and to the peasantry came before the final mobilization of the broad masses and military organizations by the parties and the paramilitary organizations from 1905 onwards, but the compromises reached in this process set the stage for markedly different outcomes in the territories separated from each other under the Treaty of 1921. In the independent Republic the civil war over the Treaty polarized the electorate between appeasers and intransigents, between doves and hawks: a very direct parallel to the split over the strategy to be pursued against Russia between the constitutionalist Young Finns and the compliant Old Finns. But this was essentially an electoral polarization, not an organizational and military one: it did not for very long affect the over-all consensus on the legitimacy of the established political system.¹⁹ The intransigents under de Valera stayed out of the system for five to ten years after the civil war: the great majority of them divorced themselves from the underground activities of the I.R.A. and re-entered the system in 1932.²⁰ But this left the Northern territory as divided as ever: the Catholic population remained second class citizens under the dominance of the Orange-Unionist alliance and the Catholic clergy and their allies in the Nationalist party tended to acquiesce in keeping their constituents isolated from the rest of the community rather than to force any change in the system that upheld the many blatant inequalities in housing, job opportunities and political representation. This polarized conflict structure proved remarkably stable for decade after decade throughout the rural areas and in the towns and the smaller cities. Only in Belfast, the most industrialized and the most secularized of the communities, was there a basis for an effective cross-cutting of communal and class cleavages, but even there it was hard to find an effective opposition against the Unionist-Nationalist deadlock. The Labour movement was split between two parties: one, largely led by Protestant non-conformists, identified with London and the broader British movement; the other, essentially manned by secularized Catholics, looked to Dublin and rallied to the Republican cause. But these outsiders at the centre of the enclave proved crucial in setting of the first serious wave of mobilization against the established system of polarized pluralism: the riots of 1968 and 1969 were as much a challenge to the appeasers in the Church and in the Nationalist party as an attack against the Unionists, the Paisleyites and the Protestant police force. Tragically, these valiant efforts to break through the inherited barriers of the religious-ethnic communities have tended to strengthen the alliances they sought to weaken: the Paisleyites have made heavy inroads on the Northern Ireland Labour Party and the attempts by such groups as People's Democracy to mobilize support for a non-sectarian Socialist platform seem doomed to failure.

In Finland the colonial enclave was kept within the national territory: the civil war did not leave a time-bomb to be set off two generations later. There, as in Ireland, the final denouement came after a series of concessions: the Swedish-Finnish elite had responded through measures of conciliation much in the same way as the English in Ireland. But the Russian masters had been less con

sistent: they had first opened up the floodgates for mass democracy in 1906 and then stopped even very modest efforts to temper the social and the economic inequalities of Finnish society. The victory of the Bolsheviks brought the system to the brink of disaster: the still only weakly integrated national community was torn apart between the mounting frustrations of the working class and the greatly aroused fears of invasion and occupation within the property-owning peasantry and the commercial and professional middle class. This proved as deep a division as the one over the Treaty in Republican Ireland but it did not polarize the population so thoroughly and for such a long time as the division between Catholics and Protestants in the Northern enclave. Finland was torn throughout the thirties between a Communist underground party and a violent nationalist movement²¹ and it took close to twenty years to heal the deep wounds of dissension brought about through the Soviet victories in 1940 and in 1944 and through the attempted coup in 1948.²²

The Contours of a Unified Model

These illustrations of paired comparisons of steps in the development of mass
democracy should help to establish one general conclusion: a unified model of
explanation must work with sets of time-specified variables.

In the original models there were essentially three time phases:23
The institutional model spans the longest gap in time: it seeks to explain

variations in the process of democratization through an analysis of the early conditions of nation-building and periphery — centre dependence. In this respect it parallels Barington Moore's model for the explanation of paths to political modernization:²⁴ Moore focuses on the crisis of the seventeenth century and tries to predict from variations during this early phase of state consolidation to variations in the character of mass politics three centuries later. But his independent variables are essentially economic: he focuses on the alliance options of the agricultural, the commercial and the military/administrative elites in each territory.

The cleavage model works with variables closer to each other in time. The only explicitly stated precondition variables are the centrality/peripherality of the territory and the outcome of the Reformation struggles: these variables control the ranges of possible elite options during the struggle over mass education after the French Revolution. The other variables bear on choices in new situations: the threat to the national commodity market through the pressures from the overseas producers, the need for regulation and protection of rights in the new industrial labour market.

The difference between the two models is easily accounted for: the institution model characterizes variations in the inherited structures of elite interchange within a territory and predict from these to the reactions to demands for mass participation; the cleavage model sorts out the possible identity bases for drives of mass mobilization. In the institutional model, the emphasis is on the established traditions of interaction, in the cleavage model on reactions in the face of swelling tides of cultural, economic and political mobilization.

In their original version, both models were highly selective in their choice of explanatory variables: the principle was one of strict parsimony. But it soon became clear that Occam's razor had shaved off too much: new variables had to be added to fit the empirical cases better.²⁵ In my efforts to link up the two models in a "dynamicized" one, I have started at the other end: I established a long list of theoretically possible system variables, proceeded to produce tables of combination, and the checked out which combinations helped to differentiate between distinct paths of development toward the end-state at full mobilization. This proved very cumbersome and I am far from satisfied with the result. I have therefore confined myself in this round to a simple presentation of the principal categories of variables: I hope later, after further discussion with colleagues, to publish a full version of this new model and to analyse the fit with historical sequences case by case.

To "dynamicize" the model we shall not only have to time-specify the independent variables but also the dependent ones. This was partly done in the institutional model: it predicts the number of steps in the decisions on the franchise and the final timing of full suffrage. But it was not done in the cleavage model: there the dependent variable was only established for one point in time, the period immediately after the universalization of the suffrage after World War I. To time-specify the expression of cleavages we cannot work exclusively with chronological time: we have to introduce some measure of "developmental time". For each "identity-building" or "identity-reducing" decision of the elite or the counter-elite, we have to specify: How wide was the franchise and how equal? How far had the accessible population already been reached through efforts of party organization? How open was the rest of the population to mobilization efforts? Take the Irish-Finnish comparison again: the Land Acts of 1881—87 were initiated under conditions of restricted suffrage and at a very early phase of peasant mobilization and thus prevented the buildup of a strong identity base at a later phase of mobilization; the abortive Finnish bill of 1914 came seven years after universal suffrage and at a time of widespread peasant mobilization. In this comparison, no account is taken of the level of "mobilizability" in Deutsch's terminology: the openness of communication structures, the spread of the money economy, the extent of geographical and ideological mobility.26 On this score, Finland and Ireland did not differ very much at the time, but such considerations will count heavily as soon as we move into comparisons with the highly urbanized and industrialized polities of Europe: consider for instance the difference in the political significance of a 40 °/o vote for Socialism in a country still 70 °/o agricultural (Finland around 1910) and a 40 °/o vote for Socialism in a country only 10 °/o agricultural (Britain after World War I). Such differences can be studied in great analytical detail if we focus on the successive time-phases of development and build into our model information about the mobilization conditions at each step of cleavage expression.

Thi would give us a total of four basic categories of variables in the model:

This is a demanding list of variables and it wil not always be possible to assign each case to one definite category. In some ways this is an enterprise similar to the Banks-Textor coding operation for current nation-states:³⁰ it differs from this much-discussed effort, first, in the emphasis on time-specified codings over several centuries of political development, secondly in its emphasis on cyclical linkages over time within a unified model of possible paths of development. Possibly the nearest parallel is with Guy Swanson's analysis of the political conditions for the emergence of Calvinist and Lutheran Protestantism:³¹ this is based on codings for 41 units on 14 variables and seeks to establish statistically that the Roman Catholic Church was most likely to survive the Reformation in what Swanson terms centralist or commensal (rule by council without representative functions) regimes. In our terms this would be an example of an analysis of linkages among the independent variables in the initial check list: there are a great number of obvious redundancies in this list and the task is to sort out, through detailed scrutiny of all pairs of cases, the minimum sets of variables required to account for the final outcomes. There is always the risk that this will get us lost in petty details of national history, but this is a risk we must face if we are to get beyond the proliferation of abstract paradigmas and airy models. We shall always have to face alteritas in our efforts to approach veritas.

---

NOTES

1 My first attempt in this direction was presented at the IPSA Round Table Conference in Ann Arbor in 1960: "The Comparative Study of Political Participation", printed in A. Ranney (ed.). Essays on the Behavioral Study of Politics, Urbana, Univ. of Illinois Press, 1962, reprinted in S. Rokkan. Citizens, Elections, Parties, New York, D. McKay, 1970. A further step was taken in the article on "Electoral Systems" prepared for the International Encyclopedia of the Social Sciences, vol. V, New York, Macmillan, 1968, pp. 6_—21. The first full-fledged model was published in "The Structuring of Mass Politics in the Smaller European Democracies", Comp. Stud. Soc. Hist. 10(2), pp. 173_—210, later, in an extended version, in Citizens, Elections, Parties, op.cit.

* My attempt at a mapping of cleavage systems grew out of years of work on the sociology of elections in Norway, see especially "Regional Contrasts in Norwegian Politics", originally published in E. Allardt and Y. Littunen (eds.). Cleavages, Ideologies and Party Systems, Helsinki, Westermarck Society, 1964, later reprinted in E. Allardt and S. Rokkan (eds.). Mass Politics, New York, Free Press, 1970, and S. Rokkan. "Norway: Numerical Democracy and Corporate Pluralism" in R. A. Dahl (ed.). Political Oppositions in Western Democracies, New Haven, Yale University Press, 1966. The first statement of a generalized model for W. Europe was presented in the Introduction to S. M. Lipset and S. Rokkan (eds.). Party Systems and Voter Alignments, New York, Free Press, 1967. This model was spelled out in further detail in the second half of "The Structuring of Mass Politics" op.cit.

---

* My attempt at a mapping of cleavage systems grew out of years of work on the sociology of elections in Norway, see especially "Regional Contrasts in Norwegian Politics", originally published in E. Allardt and Y. Littunen (eds.). Cleavages, Ideologies and Party Systems, Helsinki, Westermarck Society, 1964, later reprinted in E. Allardt and S. Rokkan (eds.). Mass Politics, New York, Free Press, 1970, and S. Rokkan. "Norway: Numerical Democracy and Corporate Pluralism" in R. A. Dahl (ed.). Political Oppositions in Western Democracies, New Haven, Yale University Press, 1966. The first statement of a generalized model for W. Europe was presented in the Introduction to S. M. Lipset and S. Rokkan (eds.). Party Systems and Voter Alignments, New York, Free Press, 1967. This model was spelled out in further detail in the second half of "The Structuring of Mass Politics" op.cit.

3 In the first round, the models were tested only against aggregate information for total national political systems. This obviously led to difficulties in cases of federal systems: the German Reich and Switzerland clearly require detailed treatment territory by territory. I was therefore very much encouraged by the critique offered by Peter Merkl, "Political Cleavages and Party Systems", World Politics, vol. XXI, no. 3, 1969, pp. 469^—485: his pinpointing of weaknesses in the interpretation of German developments in fact encouraged me to differentiate the model further for territories within larger political systems. A parallel proposal for France was made by Alain Lancelot at the Turin Conference.

4 See "The Structuring of Mass Politics", op.cit., pp. 182—188, especially Table 2.

5 See "The Structuring of Mass Politics", op.cit. Tables 5 and 6.

8 The early history of party formation in Sweden is well summarized in Per Hultquist. "Sverige", in Framveksten av de politiske partier i de nordiske land på 1800-tallet, Oslo and Bergen, Universitetsforlaget, 1964, pp. 151_—189. Thoroughgoing analysis of electoral developments and party-building efforts can be found in G. Wallin. Valrorelser och valresultat. Andrakammarvalen i Sverige 1866^—1884, Stockholm, Christophers, 1961, and in Sten Carlsson. Lantmannapolitiken och industrialiseringen, Lund, Gleerup, 1953.

7 Early Danish developments are summarized in Kjell Winding. "Danmark", in Framveksten ... op.cit., pp. 81^—101. Two major analyses of the early Danish electoral statistics have been completed in recent years by a historian and a sociologist: Vagn Dybdahl. Partier og erhverv. Studier i partiorganisation og byerhvervenes politiske aktivitet 1880^—1913, Aarhus, Universitetsforlaget, 1969, and Erik Høgh. Vælgeradfærd i et samfund under forandring, Manuscript, Institute of Sociology, University of Copenhagen, 1966. These studies are of particular importance in a comparative perspective because of the availability of individual-level electoral data from the period of open voting: on this point see the Introduction to S. Rokkan and J. Meyriat (eds.). Internatinal Guide to Electoral Statistics, The Hague, Mouton, 1969.

8 The rise of the Swedish suffrage movements has been described in great detail in T. Vallinder. I kamp for demokratin. Rostrattsro'relsen i Sverige 1886^—1900, Lund, Gleerup, 1962.

9 This contrast between Denmark and Sweden was first made in the concluding chapter of E. Thermaenius. Lantmannapartiet, Uppsala, Almquist & Wiskell, 1928, see also Herbert Tingsten. Den svenska socialdemokratiens idéutveckling, Stockholm, Tiden, 1941, vol. I pp. 19 —24, and D. Rustow. The Politics of Compromise, Princeton, Princeton University Press, 1955, p. 41. For a detailed account of the tension between the inherited agrarian anti-capitalism and the rising forces of industry, see Sven Anders Soderpalm. Storforetagarna och det demokratiska genombrottet, Lund, Gleerup, 1969. This analysis is particularly interesting because of its detailed documentation of the steps through which the final phase of formal democratization was reached through an alliance between the Social Democrats, the Liberals and the leading industrialists: Marcus Wallenberg was the Swedish Bismarck, but his action in favour of universal suffrage was basedon faith in the moderation of the industrial workers, not in the docile conservatism of the peasantry.

10 See S. M. Lipset and S. Rokkan. "Cleavage Structures, Party Systems and Voter Alignments" in Party Systems . . . op.cit. pp. 39^—40.

11 On the history of the Communist organizations in Finland see especially John H. Hodgson. Communism in Finland, Princeton, Princeton University Press, 1969, and the chapter by B. Matti in Åke Sparring (ed.). Kommunismen i Norden, Stockholm, Aldus, 1965; English ver sion in W. E. Griffith (ed.). Communism in Europe, Cambridge, M. I. T. Press, 1966, vol. 11, Ch. 17.

---

11 On the history of the Communist organizations in Finland see especially John H. Hodgson. Communism in Finland, Princeton, Princeton University Press, 1969, and the chapter by B. Matti in Åke Sparring (ed.). Kommunismen i Norden, Stockholm, Aldus, 1965; English ver sion in W. E. Griffith (ed.). Communism in Europe, Cambridge, M. I. T. Press, 1966, vol. 11, Ch. 17.

12 For an analytical account of the background of these developments see the first volume in the series "The Politics of the Smaller European Democracies": Basil Chubb. The Government and Politics of Ireland, Stanford, Stanford University Press, 1970. Data on the class basis of voter alignments in the Republic and in Northern Ireland have until recently been very scarce. A Gallup poll for the Republic reported in Nusight (Dublin), October 1969, pp. 78^—84, give these distributions by major occupational categories: Fianna Fail Fine Gael Labour Higher and Interm. Managers, Proprietors 42.0% 43 15 Clerical staff, lower adm. staff 53.5% 28 18.5 Skilled workers 45.0% 22 33 Semi-skilled, unskilled, pensioners 50.0% 17.5 32.5 Farmers over 30 acres 44.5 % 53.5 2 Farmers under 65 acres 65.0% 29 6 Corresponding figures for Northern Ireland will be presented in the forthcoming volume by Richard Rose; preliminary figures from his 1967 survey are given in R. Rose and D. Urwin. "Social Cohesion, Political Parties and Strains in Regimes", Comp. Pol. Stud., vol. 11, no. 1, April, 1969, pp. 7—67, Appendix Table 14, p. 64.

13 For a detailed review of the comments of Marx and Lenin on the prospects for an Irish revolution see Nicholas Mansergh. The Irish Question, London, Unwin University Books, new and revised edition, 1965, ch. 111.

14 David Thornley. "Historical Introduction" to Basil Chubb, op.cit., Part ii.

15 Thornley, ibid. Emmet Larkin has assembled detailed statistics to show that the dispersal of the large Protestant-owned estates not only increased the proportion of above-subsistence Catholic peasants but at the same time vastly strengthened the economic position of the Catholic Church and allowed it to dominate the educational system of the later Republic: "Economic Growth, Capital Investment and the Catholic Church in Nineteenth Century Ireland", Amer. Hist. Rev., vol. LXXII, no. 3, 1967, pp. 852—884.

16 On this point see especially Art Cosgrove. "The Gaelic Resurgence and the Geraldine Supremacy", in J. W. Moore (ed.). The Course of Irish History, Cork, Mercier, 1967. On O'Connell's fight with the 'Young Ireland' nationalists over the language issue see O. D. Edwards et al. Celtic Nationalism, London, Routledge, 1968, pp. 110_—113.

17 The rhetoric of the leaders of the Ulster riots from 1968 onwards reflects the frustrations of this criss-crossing of conflict lines: perhaps the most revealing document is Bernadette Devlin's The Price of My Soul, New York, Knopf, 1969. These quotes will suffice to give the flavour of her argument: "The problem of Northern Ireland, I decided, was not partition. If we took away partition, what did we join? If we had a truly free Ireland on the other side we would have something to join, but what was the point of ending partition merely to alter the bundaries of injustice?" (p. 90). "Paisley fools them (= the protestant workers) into thinking that their strength as the working class lies in beating down the Catholics, and as long as he can keep them away from — and in fact turn them against — the Catholic working class, he need never fear the unity of the proletariat. This influence reaches over the Protestant defenders of the Northern Ireland system to the working-class Catholics who, deprived of the support of their fellow-workers, seek strength in an all-class Catholic Alliance. Discrimination against Catholics in the system helps widen the division in the working class, and has effectively been used by the Orange Order-manipulated leaders of the state. The tragedy of the situation is that by aligning themselves with those who work against their interests but share their religion, the working class of my country, Protestant and Catholic, perpetuates its own misery" (p. 107). For an interpretation of the Free Presbyterian leader Paisley as the typical "poor-white", petitbourgeois demagogue opposing the granting of full citizenship to the lowest strata, see Peter Gibbon. "The Dialectic of Religion and Class in Ulster', New Left Review, No. 55, May/June, 1969, pp. 20^—41. For fuller details on the situation of the Catholic and the Protestant working class in the six counties, see D. Barritt 6c C. Carter. The Northern Ireland Problem, London, Oxford University Press, 1962; A. T. Q Stewart. The Ulster Question, London, Faber, 1967, and the forthcoming book on Ulster politics by Richard Rose.

---

17 The rhetoric of the leaders of the Ulster riots from 1968 onwards reflects the frustrations of this criss-crossing of conflict lines: perhaps the most revealing document is Bernadette Devlin's The Price of My Soul, New York, Knopf, 1969. These quotes will suffice to give the flavour of her argument: "The problem of Northern Ireland, I decided, was not partition. If we took away partition, what did we join? If we had a truly free Ireland on the other side we would have something to join, but what was the point of ending partition merely to alter the bundaries of injustice?" (p. 90). "Paisley fools them (= the protestant workers) into thinking that their strength as the working class lies in beating down the Catholics, and as long as he can keep them away from — and in fact turn them against — the Catholic working class, he need never fear the unity of the proletariat. This influence reaches over the Protestant defenders of the Northern Ireland system to the working-class Catholics who, deprived of the support of their fellow-workers, seek strength in an all-class Catholic Alliance. Discrimination against Catholics in the system helps widen the division in the working class, and has effectively been used by the Orange Order-manipulated leaders of the state. The tragedy of the situation is that by aligning themselves with those who work against their interests but share their religion, the working class of my country, Protestant and Catholic, perpetuates its own misery" (p. 107). For an interpretation of the Free Presbyterian leader Paisley as the typical "poor-white", petitbourgeois demagogue opposing the granting of full citizenship to the lowest strata, see Peter Gibbon. "The Dialectic of Religion and Class in Ulster', New Left Review, No. 55, May/June, 1969, pp. 20^—41. For fuller details on the situation of the Catholic and the Protestant working class in the six counties, see D. Barritt 6c C. Carter. The Northern Ireland Problem, London, Oxford University Press, 1962; A. T. Q Stewart. The Ulster Question, London, Faber, 1967, and the forthcoming book on Ulster politics by Richard Rose.

18 On the land question the standard authority is Eino Jutikkala. Bonden i Finland genom tiderna, Helsingfors, LTs forlag, 1963.

19 For a discussion of conditions of legitimacy with particular focus on the Irish situation see Richard Rose. "Dynamic Tendencies in the Authority of Regimes", World Politics, vol. XXI, no. 4, 1969.

20 See especially Frank Hunger, "The Legitimacy of Opposition: The Change of Government in Ireland in 1932", Paper, Annual Meeting of APSA, 1966.

21 On the Lapua movement see especially Marvin Rintala. Three Generations. The Extreme Right Wing in Finnish Politics, Bloomington, University of Indiana Press, 1962, and his chapter in H. Rogger & E. Weber (eds.). The European Right, London, Weidenfeld, 1965. For an assessment of these Finnish movements in a comparative perspective see Ernst Nolte. Die Fascistischen Bewegungen, Munich, DTV, 1966.

22 For an analytical account of the crisis in Finnish politics after 1945 see Krister Wahlback. Fran Mannerheim till Kekkonen, Stockholm, Aldus, 1967.

23 This restructuring of the earlier model was first suggested in my paper "Models and Methods in the Comparative Study of Nation-Building", Acta sociologica, vol. XII, 1969, pp. 53—73.

24 B. Moore. Social Origins of Dictatorship and Democracy, Boston, Beacon Press, 1966; cf. the discussion of parallels with other models in S. Rokkan. "Models and Methods in the Comparative Study of Nation-Building", op.cit.

25 See Party Systems and Voter Alignments, op.cit. pp. 41_—46, and the addition of the territorial — cultural variable into the "core" model in "The Structuring..." op.cit. Tables 4 and 7 and pp. 202_—204.

26 For further development of this theme see S. Rokkan. "Electoral Mobilization, Party Competition and National Integration", in J. LaPalombara and M. Weiner (eds.). Political Parties and Political Development, Princeton, Princeton University Press, 1966, pp. 256^—265.

27 This is a factor given much prominence in Otto Hintze's work on the survival of representative institutions in Europe, cf. "Typologie der standischen Verfassungen des Abendlandes", Hist. Zeitschr., 1930, pp. 229—248.

28 This is obviously a very crude classification. For an interesting attempt of detailed classification for the 16th century, see G. Swanson. Religion and Regime, Ann Arbor, University of Michigan press, 1967, Ch. 11l and Table I, pp. 233—241.

29 The terms "monocephalic — polycephalic" are borrowed from Juan Linz and A. de Miguei. "Within-Nation Differences and Comparisons: The Eight Spains", in R. C. Merritt & S. Rokkan (eds.). Comparing Nations, New Haven, Yale Univ. Press, 1966, pp. 267_—310. For a general review of the relevant data for Europe see R. Dickinson. The City-Region in Western Europe, London, Routledge, 1967.

30 A. Banks & R. Textor. A Cross-Polity Survey, Cambridge, M. I. T. Press, 1963; cf. more recently J. Blondel. An Introduction to Comparative Government, London, Weidenfeld, 1969.

31 G. Swanson, op.cit., Table I. A similar attempt has been made for the German cities in B. Moeller. Reichsstadt und Reformation, Giitersloh, Mohn, 1962.