I. The Institutional Infrastructure

Norway cannot boast a long tradition of disciplined research and teaching in the field of politics. Political science (statsvitenskap) was introduced as a subject at the University of Oslo in 1947 but no distinct teaching position was established for the field before 1957.¹ The first full professor of political science was in fact appointed as late as the summer of 1965. The decisive thrust toward the development of a systematic discipline of politics had not come from the University: by contrast to the other Scandinavian countries the bulk of the Norwegian work on politics had been done in independent academic research institutions without explicit teaching functions. Four of these still dominate the field:

the Chr. Michelsen Institute in Bergen, established in 1929, active in international
political research since 1938;

the Institute /or Social Research in Oslo, established in 1950, active in political
research practically from the start and operating a distinct Division of Political
Studies since 1955;

the Norwegian Institute of Foreign Affairs, established in 1960, mainly active as
a documentation and information centre, but also engaged in long-term research
in the field of international relations;

the International Peace Research Institute, established in 1959 as a Division within the Institute for Social Research but given independent status from January I, 1966, active in a wide range of fields within the sociology of economic, cultural, ethnic and national conflict and responsible for a number of empirical studies of immediate interest to students of politics.

Any attempt to review the production of political research in Norway during the
years since World War II must concentrate on the work at these five institutions:
the rapidly growing Department of Political Science (Institutt for Statsvitenskap)

at the University of Oslo and the four independent research organizations. Work at other institutions and by isolated students must of course also be taken into account: there is an important body of high-level political analyses by Norwegian historians and legal scholars and there are also examples of valuable contributions to Norwegian political research within the dynamic Department of Social Anthropology at Bergen (567, 568). Nor can the student of politics afford to neglect the contributions from the politically committed study organizations: of these Studieselskapet Næringsliv og Samjunn has been particularly active in compiling information of potential utility in academic political analysis.

This first account in the series Scandinavian Political Studies will not go into details of all these developments: it will essentially serve as a commentary to help readers find their way in the extensive bibliography given at the end of this volume. No attempt will be made to cover all fields in equal depth: the main emphasis will be on the development of a tradition of empirical and analytical research and on the few extant attempts at systematic conceptualization and model-building. The focus will be on conditions, structures and processes within national territorial communities: readers interested in Norwegian literature on international relations are referred to the Bibliography and to the surveys already published elsewhere.² Our account will also be limited in time: it will focus on the five to six years of production since the latest presentation of Nordic studies in the German collection Politische Forschung in 1960.³

II. Models and Conceptual Frameworks

Two developments have been of crucial importance for the growth of systematic
political analysis in the years after World War II: one technical and organizational,
the other intellectual and conceptual.

The technology of the "data revolution" has produced a dramatic break with the earlier traditions of single-scholar documentary research: the "new science of politics" set itself off from history and constitutional jurisprudence through its concentration on the developing methods of sampling, standardized data gathering and automated mass analysis. But this emphasis on technological innovation did not in inself guarantee the development of a systematic discipline of political analysis: the mushrooming organizations of pollsters and social surveyors and the increasingly powerful generations of computers opened up extraordinary possibilities for research but also increased enormously the intellectual responsibilities of the analysis. It became easier to assemble data and easier to manipulate them but it also became easier to produce statistical nonsense and to get lost in sheer numerology. Happily the revolution in the technology of research was parallelled by a vigorous, if not always concerted, revival of interest in theory construction, in the formulation of models and in the systematization of the frameworks for the design of research and the strategy of analysis. Only a handful of attempts were made at the formulation of general theories but a remarkable number of "middle range" models and frameworks were suggested and explored. Social scientists were caught in a dialectic dilemma: the computer technology forced them to systematize their conceptual apparatus and to sharpen their tools of inquiry but the onrush of data generated through these very techniques set increasingly difficult tasks for the theorists and the model-builders.

Since the early 1950s this dialectic has conditioned the development of all the social science disciplines, but has found particularly clear-cut expression in the field of political analysis.⁴ On the one hand the political process itself generated vast masses of readily codable data for this new machinery of analysis; on the

other hand the conceptual and the methodological equipment inherited from a long tradition of philosophical thinking and historical exegesis proved highly inadequate in meeting the technological challenge. There was a gap between the accelerating production of new data for analysis and the only slowly developing efforts to bring these masses of information under conceptual and analytical control.

This "data-theory" gap became a concern of political analysts in most countries in the West in the late '50s and the early '60s: it was perhaps less discussed in the circles where the study of politics remained a branch of traditional history but it became an increasingly important issue among all students concerned to develop a truly comparative and cross-national discipline of political analysis.

In Scandinavia a number of attempts were made during the early '60s to take up the challenge of this gap between theory and data in the study of politics. Most of these theoretical formulations reflected developments in sociology: in fact there was a marked tendency toward a merger of research traditions at the borderline between sociology and empirical and statistical politics.⁵ Surprisingly little was done to explore the potential uses of economic models and econometric methods in the analysis of political data. The anthropologist Fredrik Barth made use of paradigms from the theory of games in his study of the Swat polity in Pakistan⁶ but there was little evidence of attempts to systematize such approaches in the study of the complex nation-states of the West. Nor was there much original work on the development of formal models of decision-making in administrative organizations: an outstanding exception is Knut Dahl Jacobsen's attempt at an analysis of a crucial phase in Norwegian administrative history through the use of models of the type suggested by Herbert Simon (302).

Explicit sociological frameworks for political analysis have been sketched by
Erik Allardt in Finland, Ulf Himmelstrand in Sweden, Vilhelm Aubert, Johan
Galtung and Stein Rokkan in Norway.

Erik Allardt's continuous efforts to spell out paradigms for the analysis of forces making for conflict or consensus in industrializing societies have attracted a great deal of attention among students of politics in Scandinavia. His emphasis on the importance of theory building in the study of variations over time and across distinct territorial systems⁷ has had an increasing impact on the planning and design of inquiries into political statistics. This influence has been particularly marked in Norway. The new technologies of data gathering and data processing had made an early breakthrough in Norway: the Institute for Social Research in Oslo served as a bridge-head for a distinguished phalanx of American social scientists. The theoretical orientation of the early studies was essentially social psychological: the emphasis was on individual and small-group variables and there was little if any concern for the historical and the geographical contexts of the variations under study. This changed with the institutionalization of nation-wide inquiries: the analysts had to face up to the wide range of variations among localities and regions and were forced to look for developmental dimensions in these variations. The new analysis tasks increased the demand for macrosociological model-building: Allardt was one of the first to have taken up this challenge and his work had a distinct impact on further explorations in this direction.

This trend toward a concentration on developmental and ecological dimensions
of the national polity can be seen in the work of a number of Norwegian social
scientists.

Vilhelm Aubert has broadened his study of the origins and functions of lawyers in Norwegian society (427) into an over-all inquiry into the growth and interlinkages of the professional elites (426, 428, 429), and suggested ways of approaching the sociology of nation-building and national integration.⁸ Interestingly, this interest

in the growth of core elite of nation-builders has been coupled with an intensified concern with developments at the peripheries of the nation, in the economically and culturally backward communities of the North, the latest to enter into the national system of communication and exchange (422, 528)⁹.

This work parallels several other endeavours on the border between sociology and politics. Odd Ramsøy (84), has explored ideas set out by Edmund Leach and Talcott Parsons and tried to spell out a paradigm for the study of relationships upwards and downwards in hierarchies of systems. Johan Galtung has tried to operationalize a general "centre-periphery" dimension and to formulate hypotheses about the consequences of differences in positions on such a scale (556, 557, 559). In later papers the elements of this dimension, geographical, economic, social, cultural and political, have been analysed in further detail and a model suggested for the generation of hypotheses about the consequences of movements towards disequilibria in rankings on each of the elementary attributes (703). This style of model-building draws its strength from the important Yule-Lazarsfeld tradition of attribute combinatorics and has very direct applications in survey research and aggregate comparisons of the type pioneered by Karl Deutsch and his colleagues at Yale.¹⁰

Parallel theoretical developments have taken place within the long-term programme of electoral research carried forward since 1956 at the Institute for Social Research in Oslo and the Chr. Michelsen Institute in Bergen (473). These studies were from the outset focussed on the political effects of the historical processes of change in Norwegian society: the continuing spread of urban settlement, the exodus from the primary economy, the increase in the size and complexity of work organizations and the consequent changes in occupational structure (396). The basic design of the studies was formulated in a two-dimensional "location chart for variables" (537). One dimension was structural: from attributes of the total nation through a series of sub-system attributes to attributes of the membercitizens. The other dimension was historical and developmental: the variables were time-specified at each level of the national polity, from the total "state of the system" in its geopolitical environment to the behaviour of the constituent organizations, collective units and individual actors.

The early studies focussed on the latest phases of change in Norwegian society: the years since World War 11. This soon proved an impossibly restrictive framework of analysis: the decisive party alignments had developed during the first decades after the introduction of manhood suffrage and the current constellations in each locality could only be understood against the background of these early alignments. The result was a gradual broadening of the programme: the task was to piece together a "statistical history of Norwegian politics", to trace the spread of series of political innovations from the central to the peripheral communities within the nation (539). The rationale for this programme was seen in the theories of social, cultural and political mobilization formulated by Karl Deutsch and Dan Lerner and brought into the broader framework of a developmental model by Gabriel Almond, Lucian Pye and their associates. What was new in the Norwegian effort was the attempt to pin down a number of "threshold points" of local development through the accumulation of time series data for all distinguishable communities within the nation. This effort generated an interest in the identification of the "crucial dimensions" of the emerging Norwegian party system (612)": this interpretive model of the translation of successive cleavage lines into distinctive party alignments still calls for detailed testing region by region but at least suggests a potentially fruitful line of comparative cross-system analysis. A first, still incomplete, formulation of a basic model for such comparisons of stages in the development of party systems was publiched in 1965¹²; it will be spelled out

in greater detail in a forthcoming volume by Seymour Martin Lipset and Stein Rokkan. The model represents a first step in the "operationalization" of the politically central dimensions of Talcott Parsons' A-G-I-L paradigm: it forst suggests ways of locating the historically given party oppositions (whether local or nationwide) in a two-dimensional diagram and then proceeds to specify conditions for the development of particular types of alignments within the framework of this classification. This attempt to get beyond the traditional two-country comparisons toward a hypothetical-deductive analysis of all extant cases represents a programme of its own: in the first round it has simply helped to pinpoint the enormous lacunae in our knowledge of even Western developments, in the next it may offer some guidelines for further cross-national studies.

III. Mass Reactions and Electoral Behaviour

The breakthrough towards a "new science of politics" invariably occurred at the lowest level of the system, at the level of the "unit citizen" (537) defined in electoral law. The enforcement of the principle of "one man, one vote, one value" set the stage for the emergence of a systematic discipline of political statistics: the economist could count physical and monetary units, the political scientist could count enfranchized citizens and their votes.

In most Western countries the earliest studies in political statistics concentrated on the data produced through the electoral machinery: the official counts. In Norway there was no tradition of electoral geography or quantitative ecological analysis before World War II: there was a curious attempt at an "ethnic" interpretation of the regional contrasts in Norwegian politics in the 'nineties (612) but there was no Siegfried, no Tingsten. The breakthrough toward systematic electoral analysis came in the late 'forties and then through the introduction of political sample surveys, not through the analysis of official records.

Private organizations for the conduct of nation-wide sample surveys emerged shortly after the end of World War II and were soon producing large quantities of information about reactions in the Norwegian electorate. The Norwegian Gallup Institute established a regular service of monthly polls and fed information about the results to the headquarters of the parties willing to pay. Academic social scientists soon took an interest in this new source of data and tried to make use of the facilities of the survey organizations. A first academic survey of the national electorate was conducted as early as in 1949. This was to be the first of a long series. The mushrooming Institute for Social Research trained a number of young academics in the techniques of field work and data processing and organized a number of sample surveys, some local, some nation-wide. Most of the nation-wide studies were carried out in co-operation with the Gallup Institute: plans for a governmental survey organization came quite late, by 1964. The majority of the surveys were not explicitly designed as inquiries into political processes in Norway but they invariably included a series of questions about public issues and electoral behaviour. Of immediate interest to political analysts were the studies of attitudes on international issues carried out within what was to become the Institute of Peace Research (555—59, 703). Of great importance for an understanding of changes in the Norwegian electorate in the early 'sixties were the surveys of attitudes to unionization carried out by Egil Fivelsdal (431, 432) and Sverre Lysgaard (416). Political analysts also benefitted greatly from the study of attitudes to the use of alcohol carried out under the direction of Sverre Brun-Gulbrandsen (445)¹³: these added important details to the growing body of statistical information on regional contrasts in political culture.

Systematically designed surveys of political opinions, attitudes and behaviours

have been organized since 1957 under a joint programme of electoral research at
the Institute for Social Research and the Chr. Michelsen Institute (473).

The first reports on the surveys carried out at the time of 1957 election were published in 1959 but it took several years before the general results of the analysis work could be presented to the public. A major event in the history of the programme was the publication in 1964 of the principal report on the local surveys conducted in the Southwest of Norway: Henry Valen's book with Daniel Katz (516), was the first broad presentation of facts and findings from a pioneering study of party activities and voter reactions in Norway. The parallel analyses of the nation-wide data collected in 1957 have not been presented in any single report but have found their way into a variety of technical reports (cf. the list published in 473, also 611, 612 and for interpretive accounts the volume Party Systems and Voter Alignments to be edited by S. M. Lipset and S. Rokkan, and Stein Rokkan "Norway: Numerical Democracy...", see footnote¹*). The pressure of other work did not make it possible to organize a further nation-wide survey at the election of 1961, but some secondary analysis was subsequently carried out on the basis of Gallup data for the dramatic period of mobilization against entry into the Common Market from 1961 to 1962 (612). By the time of the next national election, however, there was so much interest in public circles for this type of research that grants were made to the programme both from the Government and the Research Council. A nation-wide panel survey was carried out before and after the election of 12—13 September, 1965, The focus was again on the political effects of the processes of change in Norwegian society and on this occasion there were for the first time since the war really significant movements in the dependent variables: a marked increase in turnout and a sizeable cut in the strength of the governing party. To explore the effects of one major change in the technology of electioneering a study of differences between communes with and without television coverage was built into the design: this necessitated a separate interview operation in TV-less districts of the North. The first reports on these studies are likely to be ready by mid-1966.

The joint programme t^went far beyond the organization of such sample surveys, however. A central element in the programme was the development of an ecological archive of coded information on the political, economic, social and cultural structure of each commune in Norway¹⁴. This archive has cost a great deal of work and is still in full development: the aim is continuous expansion, not only through the addition of fresh data from new elections and official enumerations but also through compilation of time series back to the earliest partisan elections. The idea is to create a permanent facility, to be used not only in advanced research but also for purposes of graduate training. The first analysis reports exploited very few of the analysis possibilities inherent in this body of coded materials but plans have been worked out for much more complex computer manipulations. One early report focussed on turnout and party membership and brought in an important developmental variable: the politicization of local elections through the breakdown of the traditional territorial oppositions and the entry of nationally organized parties (cf Rokkan and Valen in publication 536): this kork has subsequently been followed up in a detailed analysis of turnout among younger voters (613) and an historical study of rates of politicization15. Another early report focussed on the sources of the remarkable contrasts in political alignments between East, West and North in Norway (612): this has recently been followed up in an initial analysis of the results of the 1965 election¹⁶.

Other reports have dealt with the role of the mass media in Norwegian
politics. The early study of readers and voters17 have been followed up in close
co-operation with the Institute for Press Research at the University of Oslo: a

report scheduled for publication in the near future focusses on the impact of
television on turnout at the dramatic local elections of 1963.

The political functions of the press have also been extensively studied through quantitative content analysis: the principal investigators in this field have been Per Torsvik and Svennik Høyer. Torsvik has completed an extensive study of content categories in the Norwegian press and Høyer has made interesting attempts at analyzing the political style of editorials as a function of audience composition and community dominance (582—84).

The Norwegian programme of electoral research has been carried forward in close co-operation with social scientists in other countries. A number of international working conferences have been organized (542) and efforts have been made to integrate findings for Norway into a wider context of comparative political analysis. Two explicitly comparative articles have been published (541, 467) and a number of ad hoc confrontations of crossnational data have been attempted. Stein Rokkan has written several programmatic statements on the potentialities of comparative electoral research and has taken the lead in organizing an International Guide to Electoral Statistics: the first volume of the Guide, covering fifteen countries of W. Europe, is scheduled for publication in 1966.

IV. Parties, Interest Groups and Popular Movements

The bulk of the studies carried out within the joint programme of the three institutes have been focussed on the explanation of reactions at the lowest level of the political system: the behaviour and the attitudes of the "unit citizens" on the electoral rolls. In recent years, however, there has been a marked tendency to devote greater resources to the study of variables at higher levels of the system: to treat such variables as dependent and worthy of explanatory efforts in their own right rather than as independent and intervening variables in the study of the mass electorate. In a mass market the behaviour of consumers can only be understood against some information about the product, the retailers, the wholesalers and the advertising operations. In a mass democracy, the behaviour of the rank and file voters cannot be understood without detailed knowledge of the alternatives marketed, the policies and the candidates promoted by each party, the local and the regional networks, the central structure, the communications machinery. But political research must go beyond pure market research: the alternatives offered to the consumers are not finished products but living human beings continuously interacting with others in a variety of organizational networks. These networks of interaction set complex tasks for the student of politics: no single approach can give all the answers. Statistically oriented students will tend to start out from the discrete units in each network: the candidates and the representatives, the party personnel, the local organizations, the press. The historically oriented scholars will tend to focus on central policy-making in the legislature and the party bureaucracy and trace repercussions down-ward in the system.¹⁸

In the joint research programme at the Institute for Social Research and the Michelsen Institute attempts have been made to explore several approaches to the study of such networks of interaction, but the bulk of the work has concentrated on unit statistics. Henry Valen has established an archive of data about candidates at the Storting elections of 1957 and 1961 and has also collected information about party organizations and party membership in each commune. The first analysis to be published on this basis focussed on the relationship between local politicization and candidate recruitment (Rokkan and Valen in publication 536). Further, much more detailed analyses of candidate characteristics have been circulated in stencilled reports in Norwegian: some of the findings will be found summarized

in Valen's article in this volume¹⁹.Following up this work at a different level of the system, Torstein Hjellum20 has analyzed the composition of the lists of candidates presented at local elections in the west of Norway and tried to pinpoint the effects of the emergence of partisan conflicts on the recruitment of leaders from different social strata: this work links up with Francesco Kjellberg's study of the professionalization of local politics in an isolated community in the North.²¹

The statistical studies of party entrenchment and leader recruitment all have an important historical dimension but they will clearly have to be supplemented by detailed studies of developments at the æntre of each party organization. Henry Valen has worked extensively on the development of the Norwegian party organizations and has prepared a broad presentation of the available information on constitutional structure and decision-making procedure in a compendium for students (515). But much still remains to be done on the history of Norwegian party developments. Academic historians have recently begun to take a serious interest in the emergence and early structuring of the parties. Rolf Danielsen has dealt in some detail with the development of distinct political alliances and "protoparties" in the 'seventies and' eighties (191, vol. II) and has analysed the changes in the party systems after the introduction of manhood suffrage in 1898 (502). He has also given us a broad overview of findings and problems in the study of developments in this build-up phase under the system of indirect elections (501). Ulf Torgersen has gone into further detail in a series of analyses of the growth of the urban parties in the 'eighties and 'nineties and has looked into principles and procedures of member recruitment (514) as well as the class composition of the electoral colleges (513). This set of studies have recently been followed up through an analysis of changes in the systems of representation on the national party conventions from the early period of indirect elections through the brief period of run-off majority elections to the period of PR elections after 1920. Torgersen's historical studies of party structure go further in sociological depth than any others so far carried out and constitute important elements in a broad programme of studies of changes in the recruitment and functions of Norwegian elite groups in the decades before and after the extension of the suffrage (228, 436, 448).

The period of run-off majority elections from 1906 through 1918 has been opened up tor detailed research by Tertit Aasland.²² Her recent study of the party labels of the candidates and the second-round coalitions in the five elections held in this period is of great interest: it points to a number of challenging tasks of statistical analysis and raises important questions about variations in party strategy under the impact of changes in the rules of the electoral game. The internal history of the parties during this period has as yet been very poorly explored. Tønnesson's study of pressure politics during World War 123I²³ adds insight into the changes in the structure and ideology of the Right, but very little is as yet known about the dynamics of the processes that led to the emergence of the Agrarian party in the years from 1915 to 1918 (cf. 504).

The further developments from the intro/duction of PR to the entry of the Labour party into the national establishment have given rise to a great deal of excited speculation and controversy but have not as yet been subject to systematic scrutiny. Langfeldt's work on the "Soviet" phase in the history of the Labour party (508) adds much to our knowledge. So does Roset's study of the first Socialist Government in 1928 (511). A couple of Master's thesfes have been written on the "Radical Right" and its offshoots in the 'thirties (498, see also 423) but these do not go far towards filling the gaps. Nils Ørvik's history of the conflicts

over foreign policy (743, vol. I) gives fascinating details about the tense years from 1928 to 1933: the trend towards a polarization of the political forces, the traces of Verfemung and militarization. But solid monographs on this phase of near-disintegration are still missing: the early work of Sten Sparre Nilson on the background of the Quislingites have not been followed up despite the great interest in the biographical details of their leader.

The internal histories of the parties after the five years of German occupation have as yet attracted little scholarly attention. Thomas Wyller has written a vivid narrative of the crucial events of the summer of 1945 and has raised a very important question for further research (110): the established parties had lost face in the early months of occupation in 1940 and did not really count in the resistance front, yet the first result of Liberation politics was the reestablishment of all the parties largely on the same lines and along the same dimensions as before the interregnum.

The intriguing history of the Communist party, the only one of the pre-war parties to organize for resistance on its own, is only poorly known: Jahn Otto Johansen's account²⁴ does not go far beyond a documentary summary. The Swedish political scientist Ingemar Glans has written an important thesis on the Left Socialists in Norway and Denmark but this unfortunately has not yet been made available to other scholars.

The history of the Labour party after the accession to power has so far mainly been described in official accounts (517): the only academic study of internal developments in the party organization is the one by Fredrik Hoffman of the short-lived split in the party ranks over the atomic rearmament of Germany.25 An American student (505) has collected a great deal of information on the efforts towards mergers and co-operative arrangements among the four non-socialist parties but made very little headway towards an interpretation of his findings. An American analyst of comparative political finance, Arnold Heidenheimer, has dealt in more detail with the parapolitical organization Liberias and has added an interesting perspective: this study has not yet been published.

The detente between the parties during the period from 1945 to 1961 has been the subject of a great deal of discussion (Torgersen in publication 536).²⁶ Ulf Torgersen has recently gone into this process in some detail and has tried to throw light on dilemmas of party strategy in a phase of de-ideologization.27

All students of political parties have in one way or another had to face up to the existence of the parallel network of interest organizations and popular movements. At the level of unit statistics the linkages between the two channels of influence can be studied through the collation of information about joint positions and memberships. Henry Valen has gone far in this direction in his study of the roles of party personnel in voluntary associations (517, 615) and Arne Stokke has looked into the organizational ties of the Stortingsmenn (193). This field is still wide open, however, < and other approaches need to be explored. Stein Rokkan has on several occasions called attention to the challenging tasks of research inherent in the "two-tier" system of public decision-making (396)²⁸ and has organized studies along these lines at the Chr. Michelsen Institute. Jorolv Moren has established a basic file of information on Norwegian organizations at the Institute and is currently editing a Register of such data. A major concern in this effort of stock-taking has been the exploration of channels of representation and influence from the organizations to the central bureaucracy. To gain some perspective on Norwegian developments a general overview of trends towards the institutionalizatioh of functional representation has been prepared for the majority of Western European countries: the first volume of this study was published in 1962 (295).

V. The Central Institutions of Government

A crude map of the territory for disciplined political research may be generated
from two dimensions in an EASTON-type model of a differentiated polity:

Most of the efforts within the empirical-statistical school of politics have focussed on the feedback units: their economic, social and cultural situation, their opinions and attitudes, their alternatives of action and their actual decisions (537). A few studies have cut across several levels of the polity but have then almost invariably focussed on only one single among a great variety of analytical tasks: the recruitment of incumbents to roles in higher-level units and the

Intel-linkages among units and across levels through cros^cutting careers and through role cumulations. In Norway, such studies can clearly be multiplied in many direction and will no doubt contribute significantly to our knowledge and understanding of processes of change in the polity from one phase of nationbuilding to another: analyses of such unit statistics will add precision and depth to the uncertain impressions of the traditional historians, technically unable as they have been to master the vast masses of potentially important data. But purely statistical studies of recruitment channels and career stages are not likely to pay off intellectually before they are built into a wider framework of research on the structural contexts, the norms of procedure and the strategies of decision-making in conflicts between units at different levels of the system. Historians have made great contributions to our knowledge and our intuitive understanding of the processes of change in the building of the Norwegian polity but have felt under less obligation to systematize their findings and to check through the consistency of their modes of reasoning: they have been fascinated by concrete sequences of decisions and have developed their conceptualizations ad hoc. To make progress toward the systematic study of nationbuilding it will be essential to develop a mixed strategy: efforts must continue toward fuller technical and analytical control of the data masses but the procedures of interpretation must be subjected to detailed scrutiny through the accumulation of contextual knowledge and through the development of theoretically grounded models applicable across a wide range of concrete situations.

The extensive studies by Aubert and Torgersen of the recruitment of elite personnel in Norway exemplify possible approaches to such a mixed strategy: they have both accumulated large masses of unit statistics but are essentially concerned to fit their numerical findings into broad theoretical interpretations of the characteristic trends of development in Norway.29

Another outstanding example of "cross-level" analysis is Knut Dahl Jakobsen's meticulous case study of the conflict between administration and Parliament over the organization of services for agriculture from 1874 to 1899.³⁰ The theme of the study is a standard subject of administrative history, the development of a distinct unit of the national bureaucracy, but the unit is not studied in isolation but in the wider context of over-all changes in the balance of political forces and in cultural modes of interaction. The entire enterprise is designed not as a contribution to "linear history" but as an attempt at the development of a model for research and potentially of a theory of the sources of variations in the strategies of bureaucratic units under changes in the levels of pressure from clients and alternative services. The study is of great potential value in the planning of comparative research on the functions of bureaucracies and representative organs in the process of nation-building: it points to crucial issues in the study of the emerging nation-states in the developing areas of the world.

Jakobsen's study suggests a cyclical movement of expansion, contraction and "de-traction" in the growth of bureaucracies under the pressure of new demands from lower levels in the system: once new channels of influence and action have been institutionalized the leeway for administrative discretion increases and the chances for new disturbances decrease. Interestingly, this line of cyclical interpretation has recently also gained ground among professional historians. In a brilliant and provocative operett of Norwegian history since 1814 Jens Arup Seip has suggested a three-stage dialectic: first nation-building under the aegis of a* concerted body of appointed officials, next -a sixty-year period of multiparty competition and mass mobilization, and finally a return to the rule of the central bureaucracy in the form of the "one-party state" (107). This academic

tour de force obviously did not pretend to be based on detailed research but ft does suggest important tasks of longitudinal analysis. Ulf Torgersen has on several occasions (109, also cf. article in publication 536) pointed to the great flexibility of the Norwegian administrative elite, its capacity for survival. This is clearly a theme of great importance in the analysis of recruitment traditions: at which points in time were the sons of the "official" families most likely to opt out of the normal career and what were the typical alternatives?

However far the social scientists plan to go in the production of time series analysis, our knowledge and understanding of the processes of national development will essentially reflect the work of professional historians. This is not the place for a detailed review of historical research on facets of nation-building in Norway. We shall confine our attention to a few outstanding examples of analyses and interpretations of immediate interest to the student of politics.

The most important historical contribution to the study of central institutions of the Norwegian polity was no doubt the four-volume history of the Storting published on the occasion of the sesqui-centennial celebrations in 1964 (191). The two first volumes, by Alf Kaartvedt and Rolf Danielsen, are at high level scholarship and represent admirable efforts of systematization. These are not "linear histories": the emphasis is on organizational structures and procedures, on the alternatives under consideration and on the forces making for decisions in the one or the other direction. Professional political scientists might in fact have produced volumes much along the same lines: less readable perhaps but with more detailed statistics. The third volume covers the period from 1906 and is at a distinctly lower level of scholarship: given the very scanty production of historical monographs for this period this was perhaps only to be expected. The fourth volume is a grabbag of contributions and deserves little notice. Tertit Aasland's study of the legislative record of women members of the Storting is of considerable interest, however. This is one of a small handful of statistical explorations of legislative behaviour in Norway: a few others have been initiated by the Institute for Peace Research (192)³¹.

The volumes on the Storting deal in some detail with the legislative-executive balance but obviosly say very little about the actual procedures of the Cabinet. A well-informed overview of recent developments in the relationships between the two branches of government has recently been presented by Stavang (266). A penetrating analysis of the juridical intricacies of the various forms of delegation of legislative authority has been given by Opsahl.32 A useful rundown of information about the organization and the routine operations of the Cabinet is found in a volume by a former official of that organ, K. Bloch (154).

A history of the central administration is nearing completion: in this Jakobsen has analyzed the crucial changes under the impact of democratization in the second half of the 19th century. Debes has discussed salient features of Norwegian administrative history and sought to interpret them in the light of recent work in the theory of organizations (333).

Aubert's pioneering studies in the sociology of law have stimulated a great deal of interest in the role of the judiciary in Norwegian politics. Aubert has shown how content analyses of court decisions can contribute to our understanding of the politics of the courts (226). Torgersen (228) has concentrated on the politics of recruitment and has given us an illuminating comparison of the Supreme Courts in Norway and the U.S. The historian Jens Arup Seip has recently opened up a fascinating discussion with the jurist Andenæs over the political role of the Norwegian judiciary (227): this controversy offers exellent examples of the importance of explicitly generated conceptual models in th«» conduct of concrete inquiries.₃₃

The Cabinet crisis in the summer of 1963 prompted a great deal of academic discussion of constitutional issues (194, 262, 267) but no systematic reinterpretation of the political functions of the Basic Law of 17 May 1814: the most comprehensive attempt at an evaluation of developments in constitutional theory is probably the one due to Torstein Eckhoff (106).

There is still no adequate academic presentation of the essentials of Norwegian political institutions. The American political scientist Storing has written a handy little textbook (108) but this is very traditional in its approach and will have to be rewritten in the light of the detailed research under way. Ulf Torgersen has published the first version of what promises to become a very useful compendium for students (109). One of the leaders in the comparative politics movement, Harry Eckstein, has recently focussed his attention on Norway and has written a "theoretical case study" on the basis of the research carried out in the joint Oslo-Bergen programme and a series of personal interviews: this intriguing volume should be out during 1966.

VI. Comparative Cross-National Research

All political research proceeds by comparisons: comparisons over time, comparisons of the characteristics and behaviours of actors and collectivities, comparisons of localities, comparisons of total political systems. Whatever its method, any systematic treatment of politics must resort to comparative analysis. It has been stated that "structural-functional analysis" is simply another term for "sociological analysis": it might be said with equal justification that "comparative politics" is just another term for "political science". This is true up to point but there is nevertheless a distinction: the expanding comparative politics movement has brought into analytical focus a broader range of units than was usual in earlier treatments. Traditional political analysis tended to stick to the one nation or the one cultural area: cross-system comparisons rarely went beyond simple juxtapositions. The comparative politics movement is vastly more ambitious: the aim is analysis across all extant units of territorial governance. The Almond-Pye Committee of the Social Science Research Council has developed a series of models för the comparative analysis of nation-building: these are not limited to one particular area of the world but are designed as paradigms for analyses across any set of systems.34 Karl Deutsch's "Data Program" is equally global in scope: the data archive set up at Yale is cross-national counterpart of the cross-local ecological archives set up in country after country in recent years.³⁵ This energetic thrust toward world coverage may boomerang but the movement will certainly have made its mark: it will at least have brought the possibilities of systematic comparisons within such areas as Western Europe into the focus of scholarly activity.

Norwegian political analysts have taken an active part in this movement. J.ohan Galtung (555) and Stein Rokkan (24) have made a number of efforts to make use of survey facilities in cross-national research. Rokkan and his colleague Henry Valen have also been concerned to advance cross-national comparisons of the findings of electoral research (467, 541, 539). Over the last decade The Chr. Michelsen Institute in Bergen has taken on an important role in the advancement of such research within Western Europe. The economist Just Faaland was among the leaders in the organization of a comparative study of the formation of economic policy in nine countries of the West and enlisted the co-operation of political scientists in this enterprise (293). The Institute has also served as the secretariat of the international Committee on Political Sociology (14) and has recently become the headquarters of a co-operative study of

"The Politics of the Smaller European Democracies".³⁶ This collective enterprise will hopefully cover all the five Nordic countries as well as the three Benelux units, Ireland, Switzerland and Austria: the idea is to assemble for each of these countries the basic information on political developments, structures and processes and to proceed to a series of contextually controlled comparisons. This study will offer a great opportunity for Nordic research: the Nordic participants will be forced to review with great care the possibilities of meaningful comparisons within the most homogeneous areas and to move step by step towards wider European contexts of any such comparisons. This is perhaps the "middlerange" strategy most likely to pay off in the long run: to start out with comparisons within the most homogeneous areas and to move step by step towards global comparisons of the type advocated by Gabriel Almond, Karl Deutsch and other oioneers of the new science of world-wide politics.

---

NOTES

1 See the offical account in Universitetet i Oslo 1911—1961. Oslo, Universitetsforlaget, 1961, vol. I, pp. 138-141.

2 Accounts of Nordic research on international relations will be regularly given in the bi-annual journal Co-operation and Conflict: a review of Norwegian studies is given by Nils Ørvik in vol. I no. 1, pp. 79—84. The activities of the Nordic institutions active in international relations have recently been described by two official committees:

1) Nordisk Udvalg vedr. forskning af international politik. Forskning af international volitik. Stockholm. Aoril 1965. 99 TOTO.. (Nordisk udredninesserie 1965:4):

2) Fredsforskningsutredningen. Internationellt fredsforskningsinstitut i Sverige. Stockholm, Norstedt och Söner, 1966, 61 pp., (Statens off. utredningar, 1966:5). Details on the activities of the International Institute of Peace Research in Oslo are found in its quarterly Journal of Peace Research.

3 Stein Rokkan and Henry Valen, "Parties, Elections and Political Behaviour in the Northern Countries: A Review of Research", and Jan-Magnus Jansson, "Der Staatsapparat als Gegenstand der politischen Wissenschaft in den nordischen Landern", respectively pp. 103—136 and 137—151, with a Bibliography pp. 237—249 in. O. Stammer Hg. Politische Forchung. Cologne, Westdeutscher Ver 1960.

4 For a recent discussion of data-theory dialectics see the chapters by S. Rokkan, K. Deutsch and C. J. Friedrich in R. L. Merritt and S. rokkan eds. Comparing Nations, New Haven, Yale University Press, 1966, pp. 3—72.

5 The single most important manifestation of this trend is the volume of papers from the International Conference on Comparative Political Sociology at Tampere in 1963: Erik Allardt and Yrjö Littunen eds. Cleavages, Ideologies, Party Systems, Helsinki, Westermarck Society, 1964.

6 Fredrik Barth Political Leadership among Swat Pathans. London, Anthlone Press, 1965 (2nd ed.), 141 pp. See also: Fredrik Barth "Segmentary Opposition and the Theory of Games: A Study in Pakistan Organization", Journal of the Royal Anthropological Institute 89, Part I, 1959: 5—22.

7 See particularly his papers in E. Alladt and Y. Littunen eds. Cleavages, Ideologies, Party Systems, and his contribution "Implications of Intra-country Variations and Regional Imbalances for Cross-National Research" pp. 233—248 in R. Merritt and Stein Rokkan eds. Comparing Nations. s

8 Lecture at the Nordic sociological meetings in Oslo in May, 1965, as yet unpublished.

9 Cf. also Francesco Kjellberg "Politisk lederskap i en utkantkommune", Tss. for Samfunnsforskning 6 (1), 1965: 74—90.

10 For a presentation of "The Yale Data Program", see R. L. Merritt and S. Rokkan eds. Comparing Nations.

11 See also Stein Rokkan "Norway: Numerical Democracy and Corporate Pluralism", pp. 70—115 in Robert A. Dahl ed. Political Oppositions in Western Democracies, New Haven, Yale University Press, 1966.

---

11 See also Stein Rokkan "Norway: Numerical Democracy and Corporate Pluralism", pp. 70—115 in Robert A. Dahl ed. Political Oppositions in Western Democracies, New Haven, Yale University Press, 1966.

12 Stein Rokkan "Zur entwicklungssoziologischen Analyse von Parteisystemen: Anmerkungen fiir ein Hypothetisches Modell", Kölner Zeitschrift fiir Soziologie und Sozialpsychologie 17 (3), 1965: 675—702.

13 Also: Sverre Brun-Gulbrandsen "Hjemmebrenning i Norge", Tss. for Samjunnsforskning 6 (1), 1965: 1—30.

14

15 Torstein Hjellum. Partiane i norsk lokalpolitikk. (The Parties ih Norwegian Local Politics) Oslo, Institute of Political Science, University of Oslo, 1965, 102 pp. typewritten.

16 S. Rokkan, H. Valen & A. Amundsen "Stortingsvalget 1965" (The 1965 Parliamentary Election). Samtiden 74 (8), 1965: 463—80.

17 S. Rokkan and P. Torsvik "Der Wähler, der Leser und die Parteipresse" Kölner Zeitschr fur Soziol. (2) 1960: 278-301, cf. the general discussion in 592. For further details see Per Torsvik "Massemedia", pp. 10—1 — 10—54, in Det norske samfunnet. Oslo, Institute of Sociology, 1966, mimeo.

18 A remarkable example of systematic model-oriented historical analysis of this type is Björn Molin's Tjänstepensionsfrågan. Göteborg, Akademiforlaget, 1965, 213 pp., cf his article in this volume pp. 45—64.

19 Above pp. 121-166.

20 T. Hjellum. Partiane .... op. cit.

21 F. Kjellberg "Politisk lederskap .... op. cit.

22Tertit Aasland "Valgordningen 1906-1918", (The Electoral System 1906-18) Historisk Tidsskrift 44 (4) 1965: 267—97.

23 Kåre D. Tönnesson "Et department 'med det rette hjertelag for næringslivets vel'" (Business interests demands for changes in the central administration around World War I), Historisk Tidsskrift 44 (1), 1965: 1—16.

24 Jahn Otto Johansen "Norge" pp. 15—60 in Åke Sparring ed. et al. Kommunismen i Norden. Oslo, Dreyers Forlag, 1965.

25 Fredrik Hoffman. Påskeoppröret 1958 (The Easter uprising 1958). Oslo, Pax, 1 f\£*£f

26

27 Ulf Torgersen Partines dilemma Oslo, Institut for Samfunnsforskning, 1964, 16 pp. mimeo.

28 See further S. Rokkan "Norway: Numerical...." op. cit.

29 See particularly Aubert's study of the changing functions of lawyers in Norwegian society (427) and Torgersen's reflections on Norwegian political culture in Det norske samfunn, op. cit. Ch. 6. Another example of such contextual interpretation of recruitment statistics may be found in the Galtung-Ruge report on recruitment to the Foreign Office (344).

30 Cf. the summary in this volume pp. 59—93.

31 Ottar Hellevik "Stortingets utenrikspolitiske spørrevirksomhet" (The Parliament's use of "Questions" on matters concerning foreign policy). Tss. for Samfunnsforskning 6(2), 1965: 91-112.

32 Torkel Opsahl. Delegasjon av Stortingets myndighet. (Delegation of Parliamentary Powers). Oslo, Johan Grundt Tanum, 1965, 399 pp.

33 Johs. Andenæs. "Høyesterett som politisk organ". (The Supreme Court as a Political Instrument), Hist. tss. 44(1), 1965: 66—91; also "Jus og politikk: Svar til Jens Arup Seip" (Law and politics: An answer to Jens Arup Seip), Hist. tss. 44(3), 1965: 298-308.

34 See especially Gabriel Almond "A Developmental Approach to Political Systems". World Politics. XVII (2), 1965: 183-214, and Lucian Pye. Aspects of Political Development. Boston, Little, Brown, 1966.

36 For a presentation and discussion see R. L. Merritt and S. Rokkan eds. Comparing Nations op. cit., and S. Rokkan ed. Data Archives, op. cit.

36 For a first presentation of this enterprise, see Val R. Lorwin "The Comparative Analysis of Historical Change: Nation-Building" in the Western World". Int. Soc. Sci. J., 17(4), 1965: 594-606.