University of Helsinki

I INTRODUCTION

A. The Problem and Basic Concepts

Political ideology can be defined as a set of attitudes, opinions and beliefs that constitute at least some kind of system on which some political group may be considered to base its activities, (ef. Tingsten 1941, 9; Jansson 1959) In this study an attempt is made to discover those unifying factors that are fundamental to both the value content and the structure of ideology presupposed by the definition. Only the existence of some unifying factor can justify an assumption ofasystemof opinions, attitudes and beliefs instead of only some accidental grouping of human expressions.

To date hardly any systematic mapping of these basic dimensions of ideologies has been attempted. There are, of course, several more or less theoretical a priori descriptions which list and classify the topics around which ideological discussions and disagreements concentrate. In these classifications, however, the main emphasis has been on the characteristic value content of ideologies rather than on the search for basic structural dimensions in general. However, if we hope to move towards a more general formulation of theory in this field, it is necessary to have basic information about the common features of both the structure and content of different ideologies.

Naturally, this study cannot provide a definite solution to the problem of the general dimensions of the content and the structure of ideologies. But a study of this kind, based on a systematic analysis of the manifest ideologies, i.e. the language of ideology, can give some indication of the direction in which the research on ideology possibly can be developed.

In the course of this study it is necessary to resort to certain customary concepts in the field. Since the content of these concepts is often rather ambiguous, if not endowed even with multiple meanings, it is necessary to give them a short nominal definition. The following concepts are most frequently used:

Ideology. A general nominal defintion has been already given above. The terms

"language of ideology" or "language of politics" are also used in the text almost synonymously with ideology. Operationally, ideology is defined in this study to be identical with "party ideologies", as they appear in the statements of principle, i.e. the general programmes officially adopted by the Finnish political parties.

Content of ideology. This refers to the subject matter of the programmes, such as social, economic, cultural, governmental and political issues (e.g. public administration, taxation, church and religion, school policies, etc.), regarding which a stand is being taken.

Basic ideological ordinal scale (BIOS). The terms "party scale" or "party grouping along the right-centre-left dimension" are also used synonymously to express this concept. BIOS — in a multiparty system — refers to an ordinal scale based, from right to left, on the intensity to preserve, develop or change the social order and political system of the country. The intensity of this attitude is measured by content analysis of the official party programmes. (Borg 1964) The order of the Finnish parties included in the study is as follows: Kansallinen Kokoomus (Conservative Party), Svenska folkpartiet (Swedish People's Party, mostly liberal), Suomen kansanpuolue (Finnish People's Party, mostly liberal), Maalaisliitto (Agrarian Party), Suomen pientalonpoikien puolue (Finnish Small Holder's Party), Suomen sosialidemokraattinen puolue (Social Democratic Party of Finland), Suomen kansan demokraattinen liitto (Finnish People's Democratic League), Suomen kommunistinen puolue (Communist Party of Finland). This order is controlled by a study of voters' opinions on the average image of parties. (Borg 1963) The BIOS is used in two different ways; first, as an intervening variable to describe the general political nature of the variables included in this study, i.e. to show their "colouring"; and secondly, as a dependent variable which calls for specification itself. This doesn't purport that the BIOS would be a broad explanatory variable for differences in political attitudes and opinions. This is sometimes more or less implied when one tries to describe and/or explain in a general way the differences in individual attitudes or group cleavages in a multiparty system.

The structure of ideology. This refers to those common properties found in ideological statements which are present independent of content or value differences. For example such general structural properties are: whether the statements are abstract or concrete, whether they are formulated positively or negatively, or whether they try to influence behaviour more on the verbal than on the actual level, etc. It is not always possible to make a clear distinction between content and structure. For instance, the emotionality or intensity of expressions might equally well be considered as a property of the structure or of the content of ideology.

The expressive function of ideology. This means that the language of politics — as any language — can be seen primarily as a part of general meaningful human communication, whereby the source of a message wants to transmit to the receiver certain perceptions, opinions, emotions, etc.

The instrumental function of ideology. This means that the statements of a manifest ideology do not solely serve the purpose of "pure" transmission of messages, byt also are instrumental in trying to make the receiver think and/or act in a certain manner desirable to the source of the message.

The political Right. In a multi-party system this refers to those parties which take the most uncompromising stand toward communism, which manifestly represent conservative or bourgeois ideologies, and which are proverbially located "at the right Wing". In this study the Conservative Party and the Swedish People's Party are included in the Right (on the basis of their party programmes). The latter, however, is occasionally included also in the Centre.

The political Centre. This refers to those parties which hold a more moderate view of communism than the Right and do not represent conservative ideologies nor call themselves bourgeois, but parties of "the political centre". The Liberals and the Agrarians, and in certain cases also the Social Democratic Party and the Swedish People's Party, are here included in the Centre.

The political Left. It designates those parties which regard themselves as communist, socialist and anti-bourgeois. As such will be considered the Communist Party of Finland, the Finnish People's Democratic League and the Social Democratic Party of Finland. The latter, however, is sometimes also included in the Centre.

B. Data and Analysis

The documentary materials of this study consist of the basic long-range programmes
of the following Finnish parties:

— The Swedish People's Party (SPP), programme adopted in 1937 (no longer
in force, replaced in 1964 by a new programme)

— The Finnish People's Democratic League (FPDL), programme adopted in
1949 (still in force)

— The Agrarian Party (AP), programme adopted in 1950 (no longer in force,
replaced by a new programme in 1962; in 1965 the party changed its name
to The Centre Party)

— The Finnish People's Party programme adopted in 1951 (no longer in force,
replaced by a new programme in 1966; in 1965 the party changed its name
after fusion with the small Liberal Union to The Liberal People's Party)

— The Social Democratic Party of Finland, programme adopted in 1952 (still
in force)

— The National Coaltion Party (Conservative), programme adopted in 1957
(still in force)

— The Communist Party of Finland, programme adopted in 1957 (still in force)
— The Finnish Small Farmers' Party, programme adopted in 1959 (still in force).
Since we are not interested in the unique characteristics of the content of

different party programmes but in their common structural and substantive
features, the changes some of them have undergone soon after the completion
of this study have no significance from the standpoint of the results'. Party programmes

are used as the data in this content analysis rather than, for example, political speeches, other party declarations or doctrinal writings in newspapers, because this kind of principal programmes are the very link between theory and practice in politics. They most clearly demonstrate the two-fold function of ideologies — expressive, on the one hand, instrumental, on the other. Everyday political debate and its byproducts are often too diffuce and heterogenous to lend themselves to systematic analysis, while doctrinal writings are too often far removed from the expedient and instrumental function of ideology. Furthermore, political programmes are documents which in a way may be considered to endow political activity with certain norms. The programmes are submitted to the party memebrs for official endorsement; consequently, they are considered to represent and express the policy collectively adopted by the party.

The first step of analysis was carried out by content analysis. A preliminary classification of the programmes (programme statements) was carried out according to class defintions based on both theoretical considerations and empirical observations of party ideologies. After the preliminary classification, the definitions of categories were corrected and specified. This, as well as the final (second) classification, was done in its entirety by the author himself. Both intracoder and inter-coder reliability tests were carried out.

In determining the frequencies a word was used most commonly as the unit of classification. In some classifications the whole programme statement was used as the unit of measurement. Accordingly, the context unit, i.e. that portion of the text wherein the presence of class criteria was observed, was either a given programme statement or the preceeding and succeeding statements, depending on the unit of measurement.

In order to eliminate the differences in the length of the programmes, the absolute frequencies were converted into relative ones. This was done simply by expressing them in percentage of the absolute tength of the programme, i.e. of the total sum of programme statements in question. Due to the limitations of the reliability of the classification and the uncertainty of the statistical nature of the frequency distributions we chose the ordinal level of measurement. For the convenience of manual analysis Spearman's rank order correlation coefficient was used as the parameter of interdependence.

II ANALYSIS OF VALUES

A. Theoretical Background and Variables

In analyzing values expressed in the programmes the point of departure is the expressive function of language. It is assumed that "behind" every programme there is a collective subject - a source system — whose opinions, attitudes and values are expressed in the programme. Because of the nature of this study we are not interested in describing the content or characteristic features of each source system, but in identifying such general and common features and issues around which all of them, i.e. the party programmes, are constructed, notwithstanding their differing ideological content.

In order to discover these basic value dimensions we defined a total of 24 variables which signify such general aims or norms of political, economic, social and cultural activities, which the parties themselves have declared to achieve, i.e. to preserve, develop of more or less radically change. The selection of the variables (content categories) was based, on the one hand, on a priori view of the basic ideological values, and on ex post facto knowledge of interparty disagreements on the ideological level after the preliminary classification of the programmes, on the other. The variables and their general nominal definitions are as follows. Some examples of the operational definitions are also mentioned.

1. Parliamentarism. The statement contains an explicit demand for adherence to parliamentarism,
or a reference to and to the social reforms introducedit.

2. Enterprising Individual. The statement contains such appreciative references to an individual
as "enterprising", "has initiative", "resourceful", etc. Also substantives denoting the same are
included.

3. Irreproachable citizen. The statement mentions positive personal characteristics, especially from the standpoint of "preserving society". Examples of such expressions are "responsibility", "conscientiousness", "integrity", "sincerity", "modesty", "sacrifice" "chastity", "sound morals", "law-abiding", etc.

4. Social justice. The statement contains expressions emphasizing adherence to the principles
of justice, fair play or equality in political activity and in particular in deciding on the distribution
of social benefits among different social groups.

5. Constitutional form of government. The statement contains expressions either defending or supporting the prevailing constitutional form of government. In addition to expressions explicitly referring to the form of government (or its synonyms like political democracy, governmental democracy), included are also those which deal with essential characteristics of the prevailing system, like the independence of courts, legal security, etc.

6. Constitutional civic rights. The statement contains expressions either supporting or promoting
the civic rights and liberties guaranteed in the constitution.

7. Nationalism. The statement contains expressions emphasizing nationalistic way of thought,
nationalistic values, customs, culture and traditions. Also expressions placing particular emphasis
on "Finland" or things "Finnish" are included.

8. National independence. The statement contains expressions either supporting the independence
and freedom of the country, sovereignty and national existence, or expressions aiming to promote
these.

9. Patriotism. The statement contains expressions either emphasizing or promoting the meaning
of patriotism, preparedness for national defence, national cohesion, unity, strength and
perseverance.

10. General economic security. The statement contains expressions concerned with the achievement, maintenance and development of economic security. The items touched upon are, e.g. Improvement of the standard of living, safeguarding the value of money, the maintenance of employment opportunities, etc.

11. Economic productivity. The statement contains expressions either emphasizing or promoting the efficiency, profitability, competitiveness, rationalization, etc., of economic activities. As indicators are often used such expressions and symbols as "capacity", "technology", "capital accumulation", "profitable loans", "interest", "credits", "savings", etc.

12. Economic welfare. The statement contains explicit demands for achieving a state of
economic welfare, the improvement of the standard of living and earnings, and for the increase
of consumer goods, etc.

13. Social security. The statement contains expressions which emphasize the expansion of the
social security system either by legislative or other means, a guarantee of social benefits, the
improvement of housing and health standards, etc.

14. Development of cultural life. The statement contains expressions either supporting or promoting
the development of education, training, science, arts, research and intellectual or physical
culture in general.

15. Classical republican ideals. The statement contains positive appreciation of liberty (as a
general value), equality, fraternity, human rights, and similar ideals.

16. Individual contentment. The statement contains expressions either emphasizing or promoting the happiness of the individual, his inner contentment, security and welfare in this sense. Expressions that have a material emphasis or refer to economic or social security are not included.

17. Pacifism. The statement contains expressions either emphasizing or promoting the peace
movement, peace, opposition to wars, international comity and understanding.
18. Religiosity. The statement contains expressions either appreciating or promoting Christianity,
church and its activities, religious instruction, the satisfaction of spiritual needs, etc.
19. Individualism. The statement contains expressions either emphasizing or promoting individualism,
personal freedom, tolerance, intellectual freedom, etc.

20. Democracy. The statement contains explicit expressions either emphasizing or promoting democracy, popular sovereignty and popular participation or the political influence of the people. 21. Planned economy. The statement contains expressions either emphasizing or promoting longrange economic planning and (governmental) control, the role of public or private institutional guidance, and the active participation of society (the state) in the economy.

22. Capitalistic economy. The statement contains expressions either appreciating or promoting
private ownership, freedom of enterprise, free market economy, popular capitalism, etc.
23. Socialistic economy. The statement contains expressions either emphasizing or promoting
socialization, nationalization, abolishment of the private capitalistic system, etc.
24. Bourgeois society. The statement contains expressions either emphasizing or supporting the
preservation of bourgeois society (with explicit reference to it) or the present social order in
general.

Since several categories (variables) were composed by uniting originally more specific categories into more general and extensive ones, it is not always possible to calculate the reliability index directly for each variable. Consequently, only the mean value of reliability of all variables in question has been calculated. In the analysis of values the average concurrence between the author and the parallel classifier was 83.5 per cent. For most categories it was between 80 and 90 per cent, in one instance it remained below 70 per cent.

B. Results

Table 1 indicates the distribution of parties on different variables. The rhocoefficient of each variable with BIOS has been calculated. This correlation indicates the general colouring (right, centre or left) of the variables. As such it clarifies the nature of the Right-Centre-Left dimension, and is also likely to facilitate the interpretation of the results achieved by factor analysis. As Table 1 indicates, the following variables have Leftist colouring (+.63 is used as the cutting point, which in the present material corresponds to the .05 level of significance):

In all other cases the zero correlation results from the wide scattering of the parties across BIOS. Examination of the highest frequencies for the Centre parties shows that they occur only with these three variables (see Table 1). Even then the Agrarian Party deviates on the variable "individualism".

With an intervening variable such as BIOS one can give only a very rough illustration of the interdependence of the variables. In order to get a more comprehensive picture, we must examine the whole configuration of the intercorrelation of the twenty-four variables. Perhaps the most usual multivariate technique for this type of a problem is factor analysis, (eg. Vahervuo-Ahmavaara 1956, 9) The correlation matrix is presented in Table 2. Factoring was carried out by Hotellings¹ principal component method. Altogether eight factors were extracted. The highest correlation of each variable was used as an estimate of communality, and since certain communalities after the sixth factor exceeded the value of 1.0 and the eigenvalue of the sixth factor remained so low as .996, only six factors were included in the rotation. — Even so, the reliability of the last two variables leaves something to be desired. — The factors were rotated by the orthogonal varimax method. Table 3 shows the primary factors, estimates of communalities and the results of the varimax rotation. Because of the great number of the variables, it is more practical to study each individual factor in turn. In the following, the factors are ordered according to the highest loadings of each factor. However, it must be kept in mind that some factors are rather clearly bipolar.* To control the logicality of the change from the primary to the rotated factor, both loadings appear in the following scheme.

---

* Since we are here concerned with ideological values, this is to be expected also on a theoretical basis. The goal contrasts concommittant with politics resulted in both high positive and negative coefficients in the correlation matrix. Consequently, its is to be expected that at least some of the factors must be bipolar, since their interpretation must be sought in some ideological value contrasts and not only in the apperance of some ideological value in different degrees.

* Since we are here concerned with ideological values, this is to be expected also on a theoretical basis. The goal contrasts concommittant with politics resulted in both high positive and negative coefficients in the correlation matrix. Consequently, its is to be expected that at least some of the factors must be bipolar, since their interpretation must be sought in some ideological value contrasts and not only in the apperance of some ideological value in different degrees.

* Since we are here concerned with ideological values, this is to be expected also on a theoretical basis. The goal contrasts concommittant with politics resulted in both high positive and negative coefficients in the correlation matrix. Consequently, its is to be expected that at least some of the factors must be bipolar, since their interpretation must be sought in some ideological value contrasts and not only in the apperance of some ideological value in different degrees.

Factor I

The following variables have highest loadings:
Number and name of variable Rotated factor Primary factor
(11) Economic productivity —.94 —.70
(22) Capitalistic economy -.85 -.79
(24) Bourgeois society —.74 —.66
(18) Religiosity -.71 -.87
( 3) Irreproachable citizen —.70 —.91
In addition to the above, the following variables had high negative loadings
on the primary factor:
( 2) Enterprising individual —.57 -.89
( 9) Patriotism -.58 -.72
In addition to the above, the following had high positive loadings on the
primary factor:
(23) Socialist economy .52 .82
(17) Pacifism .32 .80
( 7) Nationalism .19 .76

The general line of interpretation for Factor I is quite clear; we are dealing with contrasting ideas about economic systems. Although the bipolarity of the factor is much attenuated as the result of rotation, it still remains slightly discernible. The variable which is dominant in this factor is not the clearest ideological one. "Economic productivity" has unquestionably the highest loading; it explains almost nine tenths of the whole variance of the factor. Thus Factor i could be called by this name. However, the interpretation of the factor could be more precisely formulated by saying that it deals with the contrast between capitalistic and socialistic systems viewed primarily from the standpoint of the efficiency of economic activity. If we revert here to the programme statements, we can see that in most programmes the question of the economic system is looked at just from this viewpoint. For instance, in the programme of the Conservative Party, the reason given for the Conservative's position regarding the economic system is that free enterprise and private ownership create the best conditions for work, production and savings. Correspondingly, the programme of the Social Democratic Party states that free market economy is incapable of providing the kind of guidance for economic life that would be effective and conducive to growth.

On the whole, Factor I appears to constitute some kind of a general factor of the value contents of party ideologies, its eigenvalue is as high as 8.46, i.e. over one third of that of the total number of variables. This means that this amount, 35.25 per cent of the total variance of all the variables, is explained by this factor. Again examination of the loadings of different variables on this factor indicates that in case of ten variables more than one half of their variance is explained by this factor. In rotation, however, this number decreases to five.

Factor II

The highest loadings on factor II are:
Number and name of variable Rotated factor Primary factor
( 1) Parliamentarism —.80 —.45
( 7) Nationalism -.68 -.58
Also the following variables have relatively high primary factor loadings:
(24) Bourgeois society -.36 -.69
(14) Development of cultural life -.54 -.43
(12) Individualism .78 .68
(13) Social security .64 .72
( 2) Enterprising individual .62 .28
In addition, one factor has a high positive loading on the primary factor, viz.
(16) Individual contentment .46 .73

The interpretation of Factor II is not an easy task. The clear bipolarity of this factor becomes very evident by rotation. Consequently, interpretation must be attempted with the aid of some contrasting values relevant to the variables. Examination of the rotated factor loadings indicates that the structure of the factor has become more clear, since several ambiguous primary factor loadings approach or fall on zero. The rotation does what it ought to do! But the interpretation of the primary factor on the rotated factor would quite obviously lead to two rather different solutions. By interpretation of primary loadings the result would be some kind of a contrast between "social security" (to which the second highest loading, "individualism", can be theoretically quite easily related because of the positive component of the concept of freedom) and "bourgeois society" and its certain norms (parliamentarism) and values (nationalism). Rotation, however, emphasizes both parliamentarism and individualism to the extent that approximately two thirds of their variance can be explained by this factor. But what kind of value contrast or possibly contradiction could be found between the principles of individualism and parliamentarism? This problematic relationship is difficult to perceive and efforts in this direction could easily prove to be unproductive. We believe that the only thing that can be said is that these two concepts operate on entirely different levels: the principle of parliamentarism concerns a rather definite group norm, which in a democratic country is regarded as indispensable and which regulates the mechanism of political life on the institutional level On the other hand, individualism is a more or less metaphysical general human value on the individual level which, no doubt, often clashes with group norms, especially with those of legalistic and traditional colouring. On this basis, the interpretation is that, as with Factor I, we are here concerned with a certain type of a rather general factor (its explanatory power calculated from the primary factor is 21.75 per cent). However, this factor seems to describe more the general structure of party ideology than its value content. Accordingly, this factor may be called the dimension expressing the contrast between group norms

and individual values. This interpretation is supported by the fact that, unlike Factor I, this dimension has no clearcut ideological group or value colouring. In the initial analysis the correlation of the rankings of parties and the original frequencies of the variables with BIOS was very low; for the four variables with the highest loadings it varies between -.02 and -.36. Out of these only "individualism" has a slight center colouring.

If the above interpretation of Factor II is correct, and if the existence of this basic dimension can be verified by some other studies (e.g. of the programme literature of some other multi-party system), the "discovery" is to be regarded as theoretically significant, for the contrast between group or society and the individual has been generally considered as one of the basic problems of ideologies. Heberle, for instance, takes the view that this contrast is in a way the underlying question in all political philosophies and ideologies (1951, 31).

Factor III

The characteristic loadings on Factor 111 are:
Number and name of variable Rotated factor Primary factor
(10) Economic security .95 .57
(15) Classical republican ideals .94 .85
(23) Socialistic economy .65 .18
(12) Economic welfare .62 .72
(14) Development of cultural life -.46 -.26

The interpretation of Factor 111 is easy; we are concerned with the dimension of economic security and social equality, which is the most frequent and emphasized value among the classical republican ideals mentioned in the programmes. The dimension is a leftist-coloured one. With the exception of "economic welfare", all variables with a high loading on this factor have a high or relatively high negative correlation with BIOS. From the ordinal scale and from the original frequencies it appears that the leftist parties have clearly higher and the rightist parties lower frequencies on these variables. The programme of the Conservative Party forms an exception, because of its neoconservative stand on the development of social policies.

The eigenvalue of Factor 111 (calculated on the basis of primary factor loadings) is 3.04. This means that it explains as much as one-eighth of the total variance. The variables, "economic security" and "social equality", which are very characteristic of this factor, are almost entirely (approximately nine-tenths) explained by it. Consequently — and since the factor has hardly any bipolarity — we give the factor the above interpretation.

Factor IV

The fourth factor has the following highest loadings:
Number and name of variable Rotated factor Primary factor
( 8) National independence —.79 —.75
(20) Democracy -.75 -.65
(21) Planned economy —.71 —.49
The primary factor also has a relatively high loading on the variable:
(14) Development of cultural life —.45 -.60
( 3) Irreproachable citizen .50 .23

The fourth factor has a straightforward structure, for only three variables have high loadings. Rotation, however, almost equalizes the variable and gives the factor a slightly bipolar character. The interpretative difficulty lies in the fact that the variable "national independence" is a general basic value which does not seem to have any clear-cut ideological colouring (its correlation with BIOS is —.10) On the other hand, the variables, "democracy" and "planned economy" are clearly of leftist or of leftist-centre colouring. It can be said that the strengthening of the other end of the factor with one discernible positive loading makes the intrepretation easier, for this variable, "irreproachable citizen", has a pronounced rightist-bourgeois content (r = .81 to BIOS). Consequently, this factor contains, at least to a certain extent, ideological value contrast. Especially for the Conservative Party and the Swedish People's Party the frequencies for the three variables with highest loadings are very low. Possibly the most concise interpretation of the fourth variable can be given by calling it the dimension of national independence and democracy. This entails a slight contrast between the centre-left-coloured democratization of society (including planned economy) and the rightist-bourgeois idea of the preservation of traditional values in a legal society. The eigenvalue of the factor calculated on the basis of the primary factor loadings is 2.65, and approximately one half or more of the variance of the three variables with the highest loadings on this factor are explained by it.

Factor V

On this factor the distribution of the highest loadings is as follows:
Number and name of variable Rotated factor Primary factor
( 4) Social justice .84 —.01
(16) Individual contentment .72 —.08
Furthermore, there is a relatively high loading of the primary factor on the
variable:
(10) Economic security .13 .51
( 6) Constitutional civic rights —.67 -.48
( 7) Nationalism -.66 .18
(17) Pacifism -.61 .12
(12) Economic welfare —.54 —.07

The factor has a pronounced social-ethical content. The variable with the highest loading describes a social-ethical norm or value on the group level, and the second highest that on the individual level. A great deal of the variance of these two variables can be explained by this factor; seven tenths of social justice and approximately one-half of contentment. Therefore, we are justified in calling this dimension the factor of social justice. However, the considerable changes in loadings from primary factor to rotated factor demand a closer examination of the consistency of the interpretation. To return to the ordinal scale and frequencies, the 4th and 16th variables that characterize this factor are of distinct centre colouring: all the parties with centrist position on BIOS have considerably higher frequencies on these variables. The positive end of the factor is therefore rather homogenous. This is also true of the negative end of the factor; both variables 7, 17 and to a great extent also 12, are characterized by leftist colouring. In the case of those variables which are located on the negative end of this factor the part of their variance remaining outside this factor does not cluster around any other factor but is very much scattered. Thus, we cannot get any help for a more precise interpretation in this direction. Returning again to the intercorrelations of the variables we find that the high correlations in the expected direction fall among the variables 4, 6, and 7, and between 4 and 16. On the other hand, from the standpoint of factor interpretation, there are certain inconsistencies in the relationships among variables 17, 12, and 10. It is very probable that the loadings of these variables contain a fair amount of random variance. The eigenvalue of this factor, calculated on the basis of the primary factor loadings is only a scant 1.51, which shows that the share of this factor in the total variance is rather small.

Factor VI

The highest loadings on the sixth factor are as follows:
Number and name of variable Rotated factor Primary factor
( 5) Constitutional form of government .82 .10
( 9) Patriotism .62 .00
( 6) Constitutional civic rights .54 -.23
(13) Social security -.47 -.20
(21) Planned economy —.45 -.13

The interpretation of this factor is easy since the sth variable distinctly dominates the whole dimension; over two-thirds of its variance falls on this factor. Of the other loadings only the "patriotism" variable is significant to the content of the factor; all other loadings are low or their variance is more pronounced on some other factor. In this case the bipolarity of the factor is also without significance. The intercorrelations of the variables indicate that the correlations of the sth variable with the others is in the expected direction; the same holds true also for the 9th variable, but thereafter inconsistencies become evident. The difference between the primary loadings and the results of rotation

leads us to expect that the loadings of these variables must include a fair amount of random variance. However, since the core of the factor is distinct and lends itself to sensible interpretation, we shall call this factor the factor of legal political system. This is associated with the idea of patriotism of rightist-bourgeois colouring.

The above six factors explain over 90 per cent of the variance of the 24 variables included in the analysis, which must be considered as a very high level of explanation. The level achieved e.g. in psychological intelligence tests has generally remained at 50 to 60 per cent, while in sociological studies based on questionnaires one often has to be satisfied with 30 to 40 per cent. (Markkanen 1964) As a rule, if Pearson's correlation is used one might well be satisfied when over one-half of the total variance is explained. On the other hand, in certain studies a percentage of over 70 has been achieved.

In considering the very high explanatory power of the factors extracted in this study two reasons which render this study unsuitable for comparison with the great majority of other factor analyses should be kept in mind. First, Spearman's rank order correlation coefficient was used as the parameter of dependence. This results in high communalities, in the same way as in using tetracoric or contingency coefficients. In mathematical sense, the structure of the resulting matrix is very different from the matrices based on Pearson's correlations (they are not scalaric in nature, see e.g. Vahervuo - Ahmavaara 1955, 181). Another, and a very important difference between this and most other factor analytic studies is that the point of departure was the classification of written materials. "The response to the questions" — if we compare classification in content analysis to responses to questions in surveys — is that of only one individual and done according to rather strict and unifying class definitions and coding process. As a result, the effect of spurious factors on the outcome has been obviously a great deal less than in opinion or attitude studies based on surveys.

11l ANALYSIS OF NORMS*

A. Theoretical Background and Description of the Variables

While the explicit point of departure for the analysis of values was the expressive function of language, in this analysis the emphasis is on the instrumental function. The idea is that the party programme is a collection of norms and rules governing human activities, whereby the collective source system "behind" the programme attempts to influence the receiver's conceptions, action and valuations in a certain direction. As such norms cannot be divorced from values and valuations, nor can the study be restricted solely to "actions", to which the concept of norm mostly refers. In the following we assume that norms may be of different levels as well as of varying form, character and intensity. The analysis presented here utilizes the basic concepts which have been developed and used in the so-called general norm theory, i.e. in the exposition of the logical

---

*) I wish to thank prof. G. H. von Wright for reading this chapter and giving me some valuable advice about the terminology in this field.

*) I wish to thank prof. G. H. von Wright for reading this chapter and giving me some valuable advice about the terminology in this field.

*) I wish to thank prof. G. H. von Wright for reading this chapter and giving me some valuable advice about the terminology in this field.

characteristics of norms, (von Wright 1958-59 and 1963) The most essential
distinctions and classifications of general norm theory are as follows:

1) There is a distinction between norms concerning what ought to be and what one ought to do. In the following we call the former "norms of being" and the latter "norms of doing". The former may also be called "ideal norms" or "-rules", and the latter "norms of action" (von Wright 1963).

2) The separation of the different elements of norms. These are:

a. The content of the norm (the area of behaviour to which the norm applies).
b. The general character of the norm (permission, prescription, prohibitions.
c. The conditions of application of the norm whether it is applicable to
situations "everybody", "always" or "always when" "only in this specific
situation", etc.).
d. The authority of the norm (who orders to "be" or to "do").
e. The subject of the norm (who is being ordered to "be" or to "do") and
f. The intensity of the norm (how forcefully the order is given).

The above "classes" were used as variables in the analysis of programme statements. It has not been possible to utilize all of the classes as such, but they have been adapted to suit the material at hand. Furthermore, certain variables have been included in the analysis on the basis of the results of the preliminary classification of the programmes.

To examine the general character of the norm formulation in the light of permissions, prohibitions and prescriptions would be too formal in this study, for the nature of the language of politics, which primarily aims at effects, would not be brought out sufficiently. Consequently, one of the main subject of this analysis of norms is to investigate whether the programme statements aim at normative influence at the level of action (that is why I shall call them actual norms) or the level of valuations and attitudes (I shall call them verbal norms), or whether they are purely statements of fact. The conditions of application of the norm is again examined from the standpoint of the totality of the norm, and intensity from the standpoint of its absoluteness, efficiency and moderation. Altogether fourteen variables are included in the analysis, and their more detailed definitions appear in the following list.

1. Norms of being. — This includes statements expressive of "ought to be", which are of overtly normative character and in which reference is made to some ideal goals, ends of action, which shall or must "be", "prevail", etc. In party programmes such expressions very often deal with values of qualities. Often the verbal indicators are of the type "must be", "ought to be", "has to be", "must be permitted to do", "it must be possible", "such and such principle must be considered", etc. In this category are also included such formally non-normative statements whose contents are nevertheless, explanations, or interpretations of some more common norms of being. Such are, e.g. most of the statements about the development of capitalism in the programme of the Communist Party of Finland, for they are basedon a very definite ideal in contrast with the prevailing economic system. Examples are: "The best conditions for work, production and savings are created by free enterprise and private ownership." "Parliamentary authority must be inviolable."

2. Norms of doing. - In this category belong all those statements concerning the "ought to do" in which the content and / or aims of political activity and decision are spelled out. As a rule, such expressions are in the form of orders, promises, demands or obligations, wherein the verbs "to do", "to act", "to make" appear in different ways. For example "Annual leave must be lengthened and working hours cut down without lowering ages' .

3. The dfffuseness of the norms. - The relevance of the entire citizenry in the action of
valuation is emphasized by such words as "all", "everybody", "whole", "nobody", etc.
4. Absoluteness of the norm. - The following expressions are to emphasize how positive and
unconditional the party's position or demands are: "absolutely", "inevitably", "without reservation",
"unconditionally", "unfalteringly", "always", "constantly", etc. These statements operate
primarily on the verbal leve».

5. Effectiveness of action. - The effectiveness, extent, forcefulness, impact, etc., or the activity
either pursued or promised by the party is emphasized by phrases such as "categorically",
"increasingly", "overtly", etc. These statements operate primarily on the action level.
6. Importance of the norm content. - The significance of the content of the norm is
emphasized, e.g. by verbal indicators like "important", "valuable", "essential", "decisive",
"foremost1, "fundamental", etc.

7. Moderation of the norm or action. - Demands, positions or valuations are presented in a moderate tone that emphasizes fairness, consideration, factuality, etc., using e.g. the following expressions; "with moderation", "not unreasonably", "not one-sidedly", "not excessively", "satisfactorily", "if possible", etc.

8. The population as subject of the norm. - Here are included, in addition to the expression
"the population", also corresponding expressions designating a collectivity, such as "society",
"country", "state", "nation", etc.

9. The party as the subject of the norm. — In addition to the proper name of the party, also
synonymous expressions are taken into account.

10. The party as the authority of the norm. — By this is meant that the party demands that some
one (e.g. the state) take certain action. Verbal indicators are the same as in class 9.
11. Norms on actual level. — An attempt is being made to influence the reader's actual behaviour
through the norm in a certain direction. The expressions are primarily of three different types:
demands, commands and obligations. For example: "As a nation, we have an obligation to
extend educational work to so great a proportion of the people as possible".
12. Norms on verbal level. - An attempt is being made to influence the reader's or listerner's
opinions, ideas and attitudes in a uniform manner. As a rule, such statements take the form
of promises, arguments or certain types of interpretations of social problems and their solutions.
The promises indicate the future or planned course of action regarding certain matters. The
arguments, as a rule, focus on the significance of certain values, matters or events, or the
effectiveness of certain procedures to attain the avowed goals. The interpretations, in turn,
tend to explain the reasons and origins of certain events that have taken place or they may
attempt to forecast future events. — For example: "Our party endeavours to safeguard our
nation's independence and freedom as a sound democratic country". "It has become increasingly
clear that the interests of the basic classes in capitalistic society, workers and capitalistists,
are irreconcilable".

13. Statements of fact. - A statement of fact does not purport to influence either attitudes or actions, it only expresses some generally known empiriealfaet, some general characteristic of the party, etc. - For instance: "The wars (Finnish-Russian wars 1939—44) resulted, not only in the decrease.' of cultivated area by 10 per cent, but in the settlement of several hundreds of thousands of refugees, which constituted the greatest agricultural problem in the post-war era". "The purpose of the Swedish People's Party is to unite in a patriotic spirit the Swedishspeaking population of Finland for the safeguarding of its position and rights in its native country".

14. Conditions of application. — The norm or some aspect of the norm (the object of valuation or action) is specified and singled out as particularry important, privileged and specific, e.g.: "As a Swedish party the Swedish People's Party especially underlines the significance of the fact that Swedish-speaking youth will receive its education and training at Swedish-speaking institutes of learning".

The classification of programme statements and their analysis took place in accordance with the same principles as those employed in the analysis of values. Since some of the above variables have been formed by combining subgroups of the original classification into larger units — to obtain sufficient frequencies without disturbing the content classes as far as logic and content are concerned — it is not possible to show the reliability for each variable separately. The average reliability (the amount of concomitant codings at both time or by

different coders) of the classification was approximately 80 per cent (re-testing
by the author resulted in 82.5 per cent and the parallel classification by a team
of coders in 75.4 to 83.9 per cent).

As in the value analysis, we are not interested in detailed description or comparison of different party ideologies, but in their common structural characteristics. Consequently, the same methods of analysis were used here as in the previous section.

B. Results

Table 4 shows the distribution of parties along the different variables, as well as their correlation with (BIOS). The variables measuring the components of intensity, the "absoluteness of the norm" and the "effectiveness of action" have a strong negative correlation with BIOS; in other words, they were of leftist colouring. Also the "totality of norm situation" seems to be a primarily leftist coloured variable. Only the "norms of being" is a variable of rightist colouring. From the standpoint of the contents of party ideologies this is consistent, because in programmes this kind of norms most often are dealing with those values which one wants to maintain and preserve. — Variables 7 (moderation) and 13 (statements of fact) are characteristically of centre colouring, their correlations with BIOS being almost zero. Moreover, all the parties whose programmes place them in the Centre have the highest rankings on these variables. Variable 2 (norms of doing) is also characteristically of centre colouring, but in this case the Social Democratic Party is replaced by the Swedish People's Party as a member of the Centre.

Table 5 shows the inter-correlations between all the fourteen variables measured by Spearman's rho-coefficient, and, at the same, the results of the factor analysis carried out on that basis. As in the factor analysis of values, the loadings of both the primary factor and the rotated factor appear in the same scheme in order to facilitate the interpretation. Both the factoring and rotation techniques are the same as used above. — The great number of variables makes the attempt to interpret the factors directly from the matrix too difficult. The bipolarity of certain factors also complicates interpretation. Consequently, we again resort to an interpretative scheme based on the highest loadings. The characteristic loadings for the first factor are as follows:

Factor I

Number and name of variable Rotated factor Primary factor (11) Norm on actual level -.98 -.96 ( 8) The population as subject —.92 —.83 ( 6) Importance of content -.67 -.68 (12) Norm on verbal level .89 .91 ( 9) The party as subject .88 .96 (10) The party as authority .52 .62

This factor almost coincides with the 11th variable, and the Bth variable, too, is almost entirely explained by this factor. The intercorrelation between these two and the 6th variable shows that they all are pronouncedly positive and consistent in their differences of size. However, since the factor also has a highly loaded positive end, it cannot be ignored in the interpretation; at least its conceptual consistency with the contents of the factor must be controlled. The intercorrelations indicate that the variables at the positive pole contain no inner inconsistencies.

Factor I depicts very accurately the general structure of the normative sentences contained in the programmes. The normative sentences have a twofold pragmatic characteristic: they either aim at influencing the actual behaviour of the subject, i.e. political activity proper, or they try to influence the opinions and attitudes of the object, i.e. it operates on the verbal level. It is these two levels and their logical relationship, which this factor describes. We may call this dimension the factor of the general structure of norm formulation.

The contents of the factor may be further clarified by extending it to cover the nature of the language of politics and political activities in general. If politics is defined as the attempt to excercise influence upon, and to participate in the making of decisions in matters of common civic interest, then entire Factor I, but especially its negatively loaded end, may be said to correspond with a high degree of accuracy to this definition. The actuality of the norm specifically means the endeavour "to influence and participate". The fact that the population appears as the subject of these norms, as well as the emphasized importance of the content point out the significance of the issue for all citizens. The positively loaded end of the factor, in turn, describes the form the attempts "to influence and participate" in the actual political process takes on the party level. The parties appear either as the authorities behind the norm or as subjects making promises or demands and thereby attempting to influence the opinions and attitudes of people in a manner conducive to party goals. The conceptual difference between the two poles of the factor could be perhaps best described by saying that the negative pole represents the general contents of political activity, while the positive pole represents the view held by political groups of such activities. This interpretation is supported by the fact that none of the variables characterizing this factor have any ideological colouring, instead, their correlation with BIOS is either zero or very low (see Table 4). This factor explains approximately one-third (32.79 %) of the total variance of all fourteen variables.

Factor II

The highest loadings occur on the following variables;
Number and name of variable Rotated factor Primary factor
(5) Effectiveness of the norm —.86 —.51
(4) Absoluteness of the norm —.86 —.58

This factor does not have any bipolarity, for the highest positive loading is .38. The interpretation of the factor is very simple because both variables 5 and 4 are concerned with the intensity of the norm, and because the loadings of other variables are either zero or very low. Factor II is the factor of intensity of the norm. The factor explains one fifth (20.21 %), calculated on the basis of primary factor loadings, of the variance of all variables. As noted above, the intensity of the norms is generally higher in the leftist programmes than in the bourgeois.

Factor III

The highest loadings occur on the following variables:
Number and name of variable Rotated factor Primary factor
(14) Conditions of application -.89 -.77
( 2) Norms of doing -.84 -.59
(10) The party as the authority -.63 -.48

The interpretation of this factor comes directly from the contents of the two variables with highest loadings, consequently the name, the factor of specified norms of doing is appropriate. This dimension is descriptive of the characteristic party interests and their related demands for action (doing). The appearance of the party as the authority behind the norm is thus quite logical, provided that group interests are openly expressed. The factor explains 14.71 % (calculated on the basis of primary factor loadings) of the total variance of the variables.

Factor IV

This factor has two high loadings of opposite polarity:
Number and name of variable Rotated factor Primary factor
(3) Totality of norm situation -.86 -.61
(1) Norms of being .72 .41

The factor is bipolar but the negative pole has a considerably higher loading. The intercorrelations (Table 5) indicate that neither variable correlates significantly with other variables. The norms of being do have a slight negative correlation with the norms of doing and intensity. Thus, the factor is obviously descriptive of a conceptual dimension with an internal contrast, which is unaffected, with two exceptions, by the variables involved in this analysis. Consequently, on the basis of the available information it is impossible to say positively anything else about the contents of this factor but that it deals with the norms of being, i.e. the ideals of politics. It is not logically incompatible that the "totality of norm situation" occurs on the same factor with the "norms of being". On the contrary, this is quite natural. The characteristic of totality quite frequently concurs with attitudes toward political ideals. However, it is not easy to say anything certain about the manner in which this element internally contrasts with

the norms of being. Of course, one can easily conjecture that the logical structure of the norms of being includes a certain type of requirement for total boundedness, which might be called some kind of authoritarian idealism. This interpretation is supported by the fact that totality as a rule is most prevalent in those programmes which can also be considered to contain more authoritarianism than others. However the grounds are not sufficient to name the factor on this basis; several more variables would be needed to validate it. — The explanatory power of the factor, too, is a little under 10 per cent.

Factor V

The following two variables have the highest loadings on this factor:
Number and name of variable Rotated factor Primary factor
(13) Statements of fact -.88 -.52
( 7) Moderation of the norm —.81 —.46

The content of this factor is very clear-cut, we are here concerned with the factor of verification and moderation. In the programme statements this is achieved by refraining from attempts to influence strongly in either direction and by making more or less neutral and moderate statements of fact (it should be noted, however, that influence may also be excercised by the suitable selection of facts to be presented), and by using mitigating, restrictive or conciliatory expressions in making demands. The ordinal scale (Table 4) show that these characteristics are more prevalent in the four parties in the political centre (in its broad meaning) than in any of extremist parties. The eigenvalue of this factor is low, 1.17, which means 8.36 per cent of the variance of all variables.

All in all, the above five factors explain (calculated by the primary factor loadings)
85.2 per cent of the total variance. This must be considered a high value,
although it is not quite as high as that attained in the analysis of value.

CONCLUSIONS

In the preceding pages we attempted to explicate, on the one hand, the values or value dimensions that form the underlying bases of the ideological content of the party programmes and, on the other, the norms evoked in order to realize and sustain these values. The basic theoretical frame of reference was that each party programme reflects — or perhaps better said — represents a certain value universe that is sustained by different norms for regulating human behavior in politics. The empirical data of the content analysis consisted of the basic party programmes of Finnish political parties in effect in 1962.

By means of factor analysis of the results of the content analysis the following
six values or value dimensions were obtained:

1. Capitalistic vs. socialistic economic system

2. Group norms (values of a political system) vs. values of an individual
3. Economic security and social equality

4. National independence and democracy
5. Social justice

6. Legal (prevailing) political system

A closer inspection of these reveals that the first two dimensions reflect those general value controversies that repeatedly form the focus of both theoretical and practical political discussions. Factors 3 and 5 also represent very general values — especially emphasized in modern societs — but in the programmes they are connected with more concrete and realistic demands and goals than is the problem of the type of economic system or the individual-society controversy. Factors 4 and 6 describe the central values associated with the Finnish nation and the Finnish political system.

In the analysis of norms following five factors were extracted:

1. The general structure of norm formulation (verbal vs. actual norms)

2. The intensity of norms

3. The specified norms of doing

4. The norms of being (the ideals of politics)
5. The verification and moderation of demands

The examination of the norms followed the basic consepts and distinctions developed in the logical analysis of norms. Our factor analysis indicates that it is possible to find in programme statements many of the elements found in logical analysis, like "norms of being", or political ideals, and "norms of doing", or goals of action. On the other hand, different norm elements do not seem to be distinguishable to the same extent as in case of logical analysis of norms. In normative statements on political issues certain elements seem to be so closely interconnected that we can actually speak about a standard formula peculiar to political language: verbal norms vs. actual norms. This formula of expressions at the same time describes the nature of political vocabulary and political action, as the interpretation of the first norm factor indicates (cf. p. 113). Of the different norm elements only intensity could be separated as a distinct dimension. An interesting finding was that the analysis yielded in a way two factors of intensity: in addition to the actual intensity factor the fifth factor, the verification and moderation, also reflects intensity.

On the whole, the research reported here is mainly exploratory. To the best knowledge of the present author no corresponding analysis of the content and structure of party ideologies has been carried out so far. It seems, however, obvious that more general and more exact conceptualization cannot be achieved before the concepts and phenomena are reduced via empirical research to a

simple form that reveals the nature of their mutual relationships. Factor analysis is one possible multivariate technique for creating a subtle and simple system of description. As to the content and structure of ideologies, it would be not only beneficial but also necessary that comparable studies would be carried out for other political systems — of course, preferably for other multi-party system — in order to obtain some basic comparisons. Similarly, materials from different sources, such as political speeches, the party press, and other political propaganda, should also be analyzed. On the basis of these results one could then gradually create some reliable overall picture of the content and structure of political ideologies.

References

Borg, Olavi, "Monipuoluejärjestelmän ideologisten peruserojen kuvaamisesta"
(On Concepts Describing Basic Ideological Cleavages in a Multi-Party System),
Politiikka (1), 1963, 1-13.

Borg, Olavi, Suomen puolueideologiat (Finnish Party Ideologies). Vammala:
Werner Söderström, 1964.

Fruchter, Benjamin, Introduction to Factor Analysis. New York: Van Nostrand,
1954.

Heberle, Rudolf, Social Movements. New York: Appleton — Century — Crofts,
1951.

Jansson, Jan-Magnus, "The Role of Political Ideologies in Politics", International
Relations, (April, 1959), 529—542.

Markkanen, Touko, Sosiaaliset ja psykologiset faktorit väkijuomien käytön selittäjinä
(Social and Psychological Factors in Alcohol Consumption). Alkoholipoliittisen
tutkimuslaitoksen tutkimusseloste N:o 12, 1964.

Theand Closed Mind. New York: Basicl96o.

Tingsten, Herbert, Idékritik. Stockholm: Albert Bonniers, 1941.

Vahervuo, Toivo — Ahmavaara, Yrjö, Johdatus faktorianalyysiin (Introduction
to Factor Analysis), Porvoo: Werner Söderström, 1958.

von Wright, G. H., Lectures on norms and values delivered at the University of
Helsinki during the 1957—58 term and on theory of norms during 1958—1959.

von Wright, G. H., Norm and Action. A Logical Enquiry. International Library of
Philosophy and Scientific Method. London: Routledge and Kegan Paul, 1963.