This Master's thesis was titled "The possibilities for change in the national identity. Media construction of national identity in post-war Estonia, Finland, East and West Germany, Austria and Switzerland". The thesis strove to increase our knowledge about the chances and causes for changes in national identity. It aimed to offer empirical level comparisons of changes in the production of national identity during 1945-1995 in the media of six European countries - Estonia, Finland, East Germany, West Germany, Austria, and Switzerland. In order to conduct empirical comparisons a theoretical model was created, based on the constructivist paradigm.

Based on constructivism and ideas borrowed from the works of Bernhard Giesen it was hypothesised that national identity is formed by cultural codes; i.e. specific model of identity construction is influenced by general codes of collective identity that pose the framework for interpretation of reality by people. Empirical analysis is based on the assumption that there are specific dimensions (primordial, political, traditional, historical, cultural) inherent to national identity and that given the existence and coexistence of those dimensions it is possible to hypothesise about the codes for constructing the national identity for particular nations. It was also hypothesised that given the appearance of particular codes one can conclude changes in the production of the national identity.

The empirical data relating to the theoretical construct come from data collected in the framework of international project Media construction of national identities in post-war Europe. Many methodological problems stemmed from the fact that the analysis was essentially secondary analysis as the theoretical construct was created only after the collection of data.

Analysis presents the general frequencies of covariates as well as latent class analysis. The results of the latent class analysis are divided into three groups according to the categories of the latent variable of 'production of national identity': 'no production of national identity', 'positive production of 'us'-feeling', and 'critical presentation of 'us''. Countries in the analysis divided into different groups according to the style of journalism.

The latent class analysis can be summarised by three remarks. First, it appears that in terms of the relationship between production and non-production of national identity there is notable decrease of the production of national identity in the Finnish and Swiss media, while there is rather an increase in other countries. Second, the proportion of critical statements is increasing in all analysed countries. Third remark consists of a number of observations regarding the effects of style of argument on the content. The political and traditional dimension always co-occur in Austrian and Swiss media while in the other countries a style is also used where these two dimensions have not been connected. In Austrian, Swiss, and Finnish media there is a particularly low usage of the historical dimension; Austrian and Swiss media employed the style where the historical dimension was lacking altogether.

The results were approached by both 1) the explanation through the story of birth of the nation, 2) the explanation through the current situation of the nation. Very rare occurrence of historical dimension in Austria and Switzerland and very high occurrence of historical dimension in both Germanies can be explained via the story of birth of the nation. The same approach can also explain the coexistence of political and traditional dimensions in Swiss and Austrian media, while not in Estonian, Finnish, and German media. Swiss and Austrian nations are so-called 'state-nations' that identify themselves via state.

Situation can condition whether national identity is constructed or not. Given secure feelings about the place of state or nation in the world, there is no need to look for justification or acknowledgement (Switzerland, Finland). It is particularly interesting to follow which means of the production of national identity are employed by media, should the need for justification or acknowledgement arise. Based on this study it can be claimed that German media turns to distancing from history or identifying with history, with the stress on the incorporation of modern culture; Austria finds recognition in modern culture foremost; Estonia and Finland raise national pride by referring to traditions and language, factors that are likely to unite people. Accentuating the unity of people and calling to solidarity can be found in Swiss media

The third observation stemming from latent class analysis, increasing proportion of critical statements in all of the studied countries points foremost to more open atmosphere in the European public sphere. The presentation of sides where state was described only positively and the other side only negatively, common during the cold war, has been replaced by inclination toward criticism and greater openness.

Two levels for possibility of changing national identity were proposed: changes in the dominants of dimensions and changes in the dominants of codes. The change on the level of codes has to be considered more profound as dimensions react to historical situations but do not necessarily bring about changes in the codes shaping national identity. Based on the empirical comparison of the six European countries it can be claimed that the dimensions producing national identity are relatively stable.

Aside from the substantive results the methodological and theoretical problems are treated in the discussion. The most important methodological problem of the thesis relates to the question of equivalence of research objects and measured variables. The equivalence of measured variables has not been assessed in the framework of the international project therefore it is difficult to assess the extent the results are truly comparable. Based on personal experience of taking part in the international project I claim that two of the substantive results from the latent class analysis are based on comparable data: the production or lack thereof of national identity and the raise of criticism in the media of the six European countries. Substantive results can be tested qualitatively with the same sample or by conducting a specifically for this purpose tailored study.

The starting point for the theoretical discussion was the question about the nature of the media analysed. It was supposed that among the different opportunities for communication media represents the traditional code while the daily interaction is conditioned rather by the primordial code and the communication mediated by internet is conditioned by universalistic code. I believe that the dimensions for the production of national identity are conditioned by the code. Thus media analysis points primarily to traditional dimension, less so to primordial dimension. The thesis that studied media could not answer questions about the changes in the primordial dimension or about the connection of this dimension with other dimensions. It is acknowledged though that from the point of view of tolerance between nations the key must be in the primordial code.

Looking at the future of Europe in the context of cultural codes then the question will arise: if European Union takes over the political dimension from nation states then nation states must search for recognition via other dimensions. Every country studied has an alternative dimension to rely on - some are more accustomed to using it, others less so. The key question shall be, which is the relationship of these new dimensions with the primordial dimension? To what extent are states inclined to justifying their existence by differentiating from the 'other'?

Discussing the potential changes in the Estonian identity it was concluded that given a stable development one could expect the strengthening of the political dimension in the Estonian media. The traditional dimension will not disappear either.

This thesis asked the question about the possibilities for changes in the national identity. There is no single answer this study could offer. Though, it is hoped, it has posed material for discussions over possible changes in identity, new empirical knowledge and new theoretical approaches.