Cognitive Maps, Cultural Distances and National Stereotypes in Times of Crises: Comparing Greece and Hungary

Abstract

This chapter centres on polarised rearrangements of the imageries of ‘Others’ both in the Greek and the Hungarian public, during a period witnessing the impacts of the global financial crisis and the so-called ‘refugee crisis.’ We examined varieties of imageries of ‘Others’ in crisis situations, based on an online survey that was conducted in parallel in Greece and Hungary. Concurrently, we analysed the auto-stereotypical features of Greeks and Hungarians, along with their hetero-stereotype characteristics concerning some key nationalities (Americans, Arabs, Germans and Russians) to capture emerging patterns of sympathies, perceived skills and cultural distances. We found a substantial core of positive expressive auto-stereotypes among the Greek population. Hungarian auto-stereotypes reflected somewhat more instrumental-oriented self-images. The cognitive maps outlined by the two-mode network methodology displayed groupings of nationalities with related stereotypical attributes, such as ‘Western,’ ‘Eastern’ or ‘Peripheral.’ We also examined the beliefs of certain segments of the overall population depending on their exposure to financial and refugee crises. Media analyses of various activity domains conducted correspondingly for the respective countries highlighted substantial shifts between pre- and post-crisis patterns regarding both these countries’ positions and their characteristic features, as portrayed in the Greek and the Hungarian public discourse.

1 Introduction

This chapter deals with the parallel impacts of the 2008 financial crisis and the 2015 massive flow of refugees on public beliefs in Greece and Hungary. Our analysis utilises data from a multistage research project,¹ focusing not so much on the crises situations themselves, but on the public beliefs concerning these crises. In the spring of 2014, the Peripato Research Group of the Hungarian Academy of Sciences at the Eötvös Loránd University conducted the ‘Crisis and Social Innovation’ Survey. The questionnaire contained a module on immigration and stereotypes towards some nations and further questions on various public issues (including the issue of trust) (Fokasz et al., 2017). The next stage of our research emerged from the insight that the refugee influx from Turkey along the so-called Balkan route during 2015–2016 was an unprecedented mutual experience for both Greeks and Hungarians.

Given that this ‘refugee crisis’ generally impacted public life both in Greece and Hungary, the Peripato Research Group conducted online comparative surveys in Hungary and Greece, in collaboration with EKKE (Ethniko Kentro Kinonikon Erevnon [National Centre for Social Research]) in Athens, at the turn of 2016 and 2017. To place the initial refugee problem in a broader framework, our research also dealt with the more general questions of the economic and social shocks of the last decade that had a strong impact on citizen beliefs and public images. These crises substantially affected how the position of other nations and that of their own was seen in the Greek and Hungarian public.

Based on the 2016–17 Peripato survey findings this chapter attempts to study national and ethnic images by extending a classical technique developed by Buchanan and Cantril (1953), adopting contemporary network methodology. Our central research question concerns the impact international crises have on global world images, stereotypes and cultural distances and the rearrangements of the mental configurations in the two countries. The chapter targets the general patterns that can be revealed on the basis of various stereotypes with regard to sympathies and perceived skills. Our approach pays special attention to the ‘catnet’ character of network-like entities, that is those with salient categorical properties (White, 2008), and to the interplay of these features based on the structural and cultural duality (Breiger, 2010).

In order to approach the public discourse from other perspectives beyond the survey methodology, we also conducted a socio-semantic network analysis of media contents. A two-mode network analysis (Yang & González-Bailón, 2018) was performed on textual data based on the co-occurrence of words. On the one hand, our analysis included countries and ethnic groups,² and on the other, various activity domains as properties typically attached to these nationalities.

The media research was based on our previous comparative analysis of the Greek and the Hungarian public discourse in the course of 2015 (Bodor et al., 2016). Our research revealed that the topics which prevailed in the 2015 Greek political public discourse were the debt crisis and the ‘refugee crisis.’ The detailed reconstruction of the related processes indicated that the initial prevalence of topics was connected to the debt crisis, and that a change took place with a surge of the ‘refugee crisis’ in August 2015 (Fokas et al., 2021). Some lessons from these previous studies may provide an introductory view of the settings that surrounded the present research.

As far as Hungary is concerned, during 2015, we detected a significant turn of the political agendas in the press (Bodor et al., 2016). Since the terrorist attacks against Charlie Hebdo, the Hungarian government used various ways to place the immigration issue in the focus of the Hungarian public discourse. The character and themes, and even the vocabulary of this discourse were determined by the government’s xenophobic communication during 2015–2016 (Gerő & Sik, 2020; Simonovits, 2020):

• April 2015 witnessed the start of the so-called ‘national consultation on immigration and terrorism,’ initiated by the government. The very title of the consultation clearly showed that throughout this campaign, the Hungarian government immediately associated the topic of immigration with terrorism.

• An anti-immigration billboard campaign³ sponsored by the Hungarian government was launched in June 2015.

• The next milestone in the government discourse on the refugee problem in Hungary was the erection of a fence along the Hungarian border on July 13th. On October 16th, when the fence was completed, the massive flow of migrants into Hungary during 2015 practically ended.

• By the Autumn of 2015, as a result of the governmental campaign, the percentage of Hungarians who believed that immigration was one of the most important issues facing the EU surpassed the EU average.

• In October of 2016, the referendum was held against the Europe-wide responsibility-sharing system. Throughout the campaign, the Hungarian government associated refugees with terrorists. Although the referendum eventually turned out to be invalid, over 98% of the valid votes agreed with the government’s proposal.

The investigation of the Greek and Hungarian dailies clearly showed (Bodor et al., 2016; Fokas et al., 2017) that the Greek and Hungarian media were different in the ways in which they assimilated and presented the ‘Others’ during the 2015 high refugee influx. Comparing the two countries’ dailies, it was especially conspicuous that during 2015 the various versions used for ‘illegal migrants’ appeared in the Hungarian dailies twenty times more often than in the Greek ones, which was an apparent indication of the different ideological and political contexts embedded in the Hungarian and the Greek media. These observations were of special relevance for initiating further studies to explore the impact of international crises on global images, stereotypes and cultural distances in these two countries. Media framing were obviously influential in the development of these cultural schemes, as to both the image of nations and of ethnic groups, and to the perception of the crises. The analyses of media contents based on text corpora of political dailies on various activity domains and countries/nationalities provided additional details for the study of changes in the public opinion related to the crisis context.⁴

By presenting an overview of the related literature, Section 6.2 reveals how public attention in Greece shifted from focusing on the financial crisis towards focusing on the migration issue, while public discourse in Hungary was dominated by xenophobic and anti-immigrant content. In this section we also present a conceptual discussion, with special regard to a social network approach (Breiger, 2010; White, 2008) applied to both the surveys on national stereotypes and the media analyses of public discourse. The data sources and the specific methods used are analysed in Sect. 6.3 and range from the classical comparative technique based on a joint approach of attributes and nationalities in the adaptation of our key survey instrument, to the related thematic domains implied by our media analyses.

Section 6.4 on the findings of both pillars of our investigations highlights the visual configurations as they emerge from the surveys on national stereotypes and the thematic patterns related to the same countries, as they manifest themselves from a corresponding analysis of the Greek and Hungarian online press. The findings are summarised and discussed in the concluding (Sect. 6.5), also involving historical and cultural-anthropological material into their interpretation within a broader perspective.

2 Related Research and Conceptual Literature

Though the recent financial and refugee crises have given rise to several studies on the development of the related public attitudes and the rearrangements of the mental configurations, the focal subject matter of our study has remained more or less unexplored. A substantial body of comparative research on the consequences of the 2008 financial collapse was produced regarding public attitudes towards international institutions and inter-state relations. Sierp and Karner (2017) have raised the emergence of a new type of essentialism connecting notions of reified collectives with their economic achievements, with special regard to those in a lending position along the North/South divide. Processes of victimisation and the generation of expressive narratives have been submitted to quantitative and qualitative analyses by Lialiouti and Bithymitris (2017), Capelos and Exadaktylos (2017) and Michailidou (2017), with special regard to the currently stereotypical features of the Greek-German relationship in the Greek public discourse. Hutter and Kriesi (2019) have pointed out a sharp politicisation of the crisis phenomena on the public agendas. Regional characteristics such as an emphasis on European economic issues with respect to the polarisation of the political landscapes in the Southern part of the continent are also relevant for our study.⁵

The latter observations already relate to the effects of the 2015–16 refugee crisis, a subject which has also witnessed a growth of publications based on surveys and media analyses. As observed by Paschou et al. (Chap. 7 in this volume), public claims-making in Greece turned towards the migration issue somewhat later than in other parts of Europe. Our related analysis of online contents (Fokas et al., 2021) indicated the parallel domination of two topics in the Greek public discourse: the debt crisis and the refugee crisis, with a shift from the former to the latter beginning in August of 2015. The analyses related to the Hungarian media revealed the domination of the political discourse by the ‘refugee crisis’ in 2015, heavily impacted by xenophobic anti-immigrant propaganda messages.

Utilising comparative material from the European Social Survey, Messing and Ságvári (2019) and Gerő and Sik (2020) pointed to an exceptional increase in non-inclusive Hungarian attitudes towards migration issues. Simonovits (2014) presented the selectivity of exclusion-oriented attitudes towards migrants for this earlier period, suggesting that Hungarians tended to distance themselves to a lesser extent from European migrants than from migrants of more remote (non-European) origins. Boda and Simonovits (2016) indicated a further decrease in the receptive attitudes in Hungary in the attitudes in the highlighted 2015–16 period, while Sik and Simonovits (2019) pointed out a sharp surge of refugee and immigrant topics in the media over a similar period.

However, despite more extensive research, there are no comparative studies that attempt to approach global images from a stereotypical beliefs’ perspective, as originally conducted by Buchanan and Cantril’s (1953) classical study in the aftermath of World War II.⁶ Though not directly crisis-related, the work by Fiske and her colleagues (e.g. Fiske, 2017; Fiske et al., 2007) is an important exception on the conceptual foundation of social judgment and stereotype content theory. The researchers found the emotion-driven aspect of warmth and cognition-led competence as universal dimensions of group stereotypes.⁷ In addition to various minority segments of the American population, Fiske et al. (2007) indicated generally low scores on both warmth and competence (e.g. ‘hostile,’ ‘untrustworthy,’ respectively ‘stupid’ or ‘unmotivated’), the two principal dimensions.

The study of global images also requires attention to dominance-symbolic aspects. While the warmth/competence dimensional design has proven to be a plausible starting point for the present study, it also required some additions to the dominance-symbolic aspects of global images of particular relevance to our topic. The inclusion of a typology close to the sorts of skills and assets outlined by Bourdieu (1986) resulted in a broader framework in this respect.⁸ At the same time, the dynamic elements potentially present in the Fiske model, especially the contextual component, were more directly accessible through a network-like approach that simultaneously included aspects of cultural and structural embedment.

Global images and cognitive maps are inseparable from national stereotypes, but they do not contain the richness of pictures and memories that may be present regarding some familiar macro-collectives; however, they comprise more general cues of orientation in the world. For example, Buchanan and Cantril’s above-mentioned study appeared when the Western and Eastern ‘camps’ diverged sharply and was followed by the birth of the ‘Third World’ as a notion. Though with no exact knowledge of their meanings, most people had some ideas about the ‘good’ and ‘bad’ sides and who belonged where. The ‘great transformation’ of the nineties also brought a re-arrangement of beliefs about world images. The financial and migration crises of the last decades re-organised the perceptions and the formation of cognitive maps of the world. Besides contemporary effects, however, the global images at issue are linked to the remote past in many cases.

The mental cues of navigation at issue contain beliefs of national-ethnic entities and their wider groupings (such as what a ‘Westerner’ or an ‘Easterner’ means). Alternative emphases on being ‘good,’ ‘smart,’ or ‘strong’ are of importance just as the structural aspect of ‘being alone’ or ‘together with others.’ Therefore, the sharpness of the boundaries between ‘Us’ and ‘Them’ is of primary relevance (this is an aspect also highlighted by Kalogeraki in Chap. 5, this volume).

Emerging issues of migration have also triggered a reinterpretation of several aspects of national identity. Triandafyllidou (2006) points to the role of the ‘Significant Others’ that can be perceived as a source of threat for ethnic and cultural identity and independence. The motive of differentiating from others may surge amidst the confrontation of ideologically coloured platforms in public discourse. Krzyżanowski et al. (2018) analyse robust tendencies of politicisation of the ‘refugee crisis’ in Europe, with the related features of mediatisation entailing a sharp polarisation of public opinion on migration issues and in many cases evoking traditional exclusionary patterns rather than new ways of perception of ‘Others’ outside national boundaries.

New currents of network analysis mutually linking structural and cultural aspects are key sources for our theoretical and methodological orientation. The cultural turn of the network analysis was in many respects anticipated by Harrison White’s (2008[1965]) explications on the catnet concept.⁹ Basov et al. (2020) review of the recently increased body of socio-semantic networks research was the next step. It combined the two-mode perspective of cultural and social duality (as explained by Breiger, 2010; Everett & Borgatti, 2020) with a simultaneous approach of micro-and macro elements of society. It also emphasised not only the presence, but the absence of ties in the formation of meanings. Division and discontinuity are highlighted by the conception of ‘cultural holes’ (see Pachucki & Breiger, 2010 also with a reference to transnational aspects).

Results on belief networks based on total populations may be erroneous given that different subgroups can have different thought patterns with particular logics on their own (see DiMaggio & Goldberg, 2018). Thus, we analyse the general Greek and Hungarian survey populations and various segments, with their specific stereotypical configurations. In addition to the perception of economic hardships, we introduce the degree of trust radius as another criterion of segmentation. The level of general trust is significantly related to positive attitudes towards migration (see e.g. Drazanova & Dennison, 2018), and as part of an inclusive orientation, migration attitudes are also associated with the radius of trust (see Delhey et al., 2011).

The element of national uniqueness rooted in history, both in Greece and Hungary, is a key aspect of our empirical analyses and the discussion of findings. Our triangulation efforts are not only methodological, using both survey and media analyses to study public images, but also involve contextualising our findings in light of historical-cultural evidence (see Tsoukalas, 1995 or Luhmann, 1995 for such a broader perspective).

3 Research Design

Our research design is based on a network approach using data from a comparative attitudinal survey and on the media analysis of two daily newspapers, as described in detail below.

Comparative Survey of Stereotypes:

The adapted Buchanan-Cantril questionnaire block of national stereotypes was administered at the turn of 2016 and 2017 on subsamples of 907 Greek and 1000 Hungarian Internet-using respondents aged 18–70. Our version of the Buchanan and Cantril (1953) method required some substantive and technical modifications. We sought to balance the positive (Brave, Debonair, Generous, Hard-working, Intelligent, Open, Passionate, Practical, Self-controlled) and negative adjectives (Backward, Conceited, Crafty, Cruel, Domineering, Factious, Lazy), with some increase of the latter aspect in the questionnaire. The scope of attributes was also expanded with special regard to some in-group and out-group features of national-ethnic stereotypes. Besides the feeling aspect, our version also emphasised competencies (see Fiske et al., 2007) and the knowledge styles based on cognitive-instrumental, expressive, and dominant-symbolic skills (See Angelusz & Tardos, 1995). The respondent had to select three attributes considered most characteristic of the given nationality from the provided list. In addition to Greeks and Hungarians, we selected some nationalities –Americans, Germans, Russians– based on their historical or current political roles. Due to the 2015–16 wave of migration, ‘Arabs’ were also included.

Network Analysis in a Two-Mode Approach:

Our survey tackled two types of entities: nations and attributes which constituted each other through their linkages. Their stereotyped cognitive maps in the two populations were depicted by two-mode tools of network visualisation (see Everett & Borgatti, 2020). These global images were revealed from the joint clustering of nations and attributes, outlining cultural proximities/distances and knowledge-style dimensions in a parallel way.

Network inputs as scores for a kind of cultural matrix (see Edelmann & Mohr, 2018) were generated by the attributes’ relative occurrences with the given nationalities in the two samples. Scores of similarities and differences implied the stereotypical features attributed to nations.

We aggregated multiple response contingency tables including all nation/attribute mentions. These were transformed into a network analysis format through adjusted standardised residuals by post-hoc analysis with Bonferroni adjustment (Agresti, 2019). The positively significant and the negatively significant values highlighted the salient and the non-characteristic features for each nationality, respectively.

We identified community structures (see Girvan & Newman, 2002) based on two-mode network techniques. The two-mode Factions module was used to find sub-groups with both relational and positional features of direct or indirect relations. The community groupings comprising both aspects, nationalities and attributes, provided the outlines of network patterns with categorical characteristics.

Segmentation of the Survey Population:

One selected dimension segmented the population by their perceived exposure to the economic crisis. The degree of being stricken by the economic crisis was approached by a composite index of seven variables. These related to the evaluation of the personal situation, the country’s economic situation, life satisfaction and a question on the most important problems.

Another dimension of segmentation was related to the refugee crisis, in terms of external groups (foreigners, people of different nationalities, people of different religions) with which we segmented the population based on the levels of the radius of trust. The survey questions included various target groups from the family to the ‘strangers.’ We separated the population into three segments again, from ‘trusting only the family’ to ‘trusting more distant groups too,’ referring to minimal, narrow and broad trust radius.

Media Analysis Regarding Topic Domains and Nations:

Relying on a previous media dynamics study for some political dailies by Bodor et al. (2016), our analysis of media contents extended the period covered by the study from 2004 to 2016. The database embraced the online edition of two dailies per country, Kathimerini and To Vima, from Greece and NOL and MNO from Hungary. These dailies were chosen based on their common traits. All of them had a considerable history and solid reputations. The former ones from both countries were considered right-leaning while the latter left-leaning.

The analysed themes included nations, on the one hand, and various activity domains as properties typically attached to them, on the other. We examined the same six countries used in our survey.¹⁰ We aggregated 48 thematic units (e.g. aggression, victory, crisis, technology, and celebration), emerging from searches based on selected keywords into 14 synthetic activity domains (see Table 6.1). The 14 fields of activity take more similar positions on the cognitive-instrumental, expressive-emotional and symbolic-dominance axes than the applied stereotype attributes in the survey.

Table 6.1 The list of aggregate activity domains in the media content analysis and the thematic keywords underlying the aggregation in the textual corpus

Our semantic network analysis, following Yang and González-Bailón (2018), which was based on the compilation of word co-occurrences (see Lenci, 2008) in distinct articles from text corpora, resulted in a two-mode (nation/domain) cultural matrix for our examination.

Media Segmentation:

Corresponding to one key aspect of our survey analyses, we distinguished a pre-crisis and a post-crisis corpus of the newspaper texts. Constrained by our media study’s period limits, the former was based on data from 2004 to 2008 and the latter from 2009 to 2016. Our study focused on the complexity of global images and in-group/out-group distinctions, which were highlighted under the impacts of the financial and migration crises of the last decades. Employing a contextualised adaptation, it revived the approach of the classical comparative study by Buchanan and Cantril (1953) following the aftermath of WWII concerning stereotypes of friends and foes in the public discourse of several nations. Based on the underlying conceptual framework with a duality of national-ethnic communities and attributes mutually defining each other, it made use of the contemporary apparatus of social network analysis. Beyond Greek and Hungarian populations, two cases heavily impacted under the recent crises, our analyses also differentiated between various segments of these publics differently exposed to the recent emergencies (as distinguished by the perception of the economic crisis and the radius of trust towards ‘Others,’ including other religions and nationalities).

4 Findings

Scholarly and everyday experiences all suggest that the financial and refugee crises of the last decades were robust enough to bear upon the perceptions of the world outside and the relations of ‘Us’ and ‘Them.’ The analyses below outline how this potential was realised in the formation of public opinion in Greece and Hungary.

4.1 Global Images in the General Public in Greece and Hungary

Based on the two-mode data emerging as the frequent occurrences of the nations/attributes’ matrix, we found that Greek auto-stereotypes were very strong and mostly positive. In contrast, Hungarian auto-stereotypes reflected an instrumental-oriented self-image as an ideal. The Greek auto-stereotypes were much stronger than their hetero-stereotypes. Based on these stereotypes, it seems that Greeks had a definitive idea of the kind of people they are. It is also noteworthy mentioning that of the four stronger stereotypes, two were positive (intelligent and generous), while two were rather negative (lazy and crafty) (Fig. 6.1, left side). Based on a typology of attributes according to their differentiation into instrumental, expressive and dominance-oriented traits, the prevalence of the characteristics to the right side of the origin indicated an auto-stereotypical profile of the pronouncedly expressive character.

Fig. 6.1

Spider charts of national stereotypes. (based on odds ratios; Greek and Hungarian baseline samples)

The auto-stereotypes of the Hungarians (Fig. 6.1, right side) were much less salient than of the Greeks, and one negative attribute (factious) dominated all others. Hungarians’ stereotypes concerning Greeks were also quite strong. Hungarians also had a definitive idea as to the kind of people Greeks are. Furthermore, two attributes (generous and lazy) assigned to Greeks by Hungarians were also strong among the auto-stereotypes of Greeks.

Greeks’ and Hungarians’ stereotypes concerning Germans also shared some characteristic similarities as well as differences. Attributes such as ‘self-controlled’ and ‘hardworking’ were frequently mentioned both in Greek and Hungarian samples. The differences between the samples were also of interest. Hungarians found Germans ‘intelligent,’ while Greeks did not share this positive image. Also regarding non-frequented characterisations, according to Greeks, Germans were definitely ‘non-open,’ ‘non-generous,’ and ‘non-brave.’ It seems that Greeks also had a definitive idea of the kind of people the Germans were not. It remains a question as to how much of this negative attitude was due to uneasy memories from the past or to more recent impressions like those regarding the German role in the treatment of the Greek debt crisis.

The visual patterns outlined by the Greek survey manifested three subgroups of nations, the ‘Western-type’ German–American group, the ‘peripheral’ Hungarian and the ‘Eastern-type’ Greek-Russian-Arab subgroups, framed by some characteristic attributes (Fig. 6.2).¹¹ The abbreviation ‘N’ at the beginning of expressions (e.g. ‘Nopen’) denotes the typically non-chosen/non-characteristic traits in the figures of two-mode networks.

Fig. 6.2

Two-mode network pattern by the Greek baseline sample

The profile of the latter grouping was mainly affected by the Greek auto-stereotypes with mostly expressive features such as ‘brave,’ ‘passionate,’ ‘open’ and ‘generous.’ In addition, ‘intelligent’ was also a distinctively positive attribute, while ‘crafty’ was ambivalent in the Greek self-image with some positive nuances. Two positive attributes, i.e. ‘generous’ and ‘passionate’ and one negative, i.e. ‘lazy’ connected Greeks and Arabs, while ‘brave’ connected Greeks and Russians.

Regarding the out-group relations, Greeks and Hungarians were connected by the ‘non-domineering’ and ‘hardworking’ attribute ties, while the latter represented a link between Greeks and Germans. Germans and Americans were both characterised as ‘conceited’ as well as ‘cruel’ and ‘domineering,’ which may sound like a kind of shorthand evaluation of Western-type societies by Greeks.

Based on various criteria, we also applied the results based on total populations to various subgroups. The first of these dimensions segmented the population by their perceived exposure to the economic crisis. The labels for the suffering and thriving segments of the poles of crisis perceptions borrowed from the classic Cantril ladder approach and we found that the ‘suffering’ segment was more prevalent in the Greek sample, while the ‘thriving’ one was relatively larger in the Hungarian case.

The ‘Eastern-like’ category of nations and attributes also appeared in the ‘suffering’ segment of the crisis perception; the positive attributes of this grouping were joined by ‘intelligent’ used for the Russians too (Fig. 6.3). In the grouping including Hungarians and Germans, as a difference from the baseline setup, the neutral or sometimes positive image of the Hungarians was replaced by affiliation with a German profile, seen hereby especially negative. ‘Cruel’ became a joint attribute of Germans and Hungarians, turning the peripheral yet positive Hungarian baseline image into negative. Having suffered most from the post-2010 debt crisis, this Greek segment seems to have attributed their troubles to the severe expectations of austerity attached to Germans. Americans constituted a one-nation category with a more balanced image and a kind of mediating role.

Fig. 6.3

Two-mode network pattern by the Greek ‘suffering’ segment of the crisis perception

According to the Hungarian sample as compared with the Greek global images, the most significant difference was related to the close position of Greeks with Hungarians, mostly due to linkages by attributes like ‘generous’ and ‘non-domineering’ (Fig. 6.4). Among the Hungarian auto-stereotypes, we found two negatives: ‘factious’ and ‘backward.’ The first one was a distinctively negative attribute of Hungarians while the second one connected Hungarians with Arabs and Russians with regard to out-group relations. According to the Hungarian sample, ‘hard-working’ as well as ‘non-lazy’ connected Hungarians and Germans while among Hungarian auto-stereotypes ‘open’ was assigned to both Greeks and Americans. Somewhat differing from the pattern by the Greek sample, the ‘Westerner’ sub-grouping was embodied with mostly positive, respectively non-negative, attributes (except for ‘conceited’ in this case, too). At the same time, the contrary stood for the ‘Easterner’ counterpart, especially as far as negative instrumental features were concerned.

Fig. 6.4

Two-mode network pattern by the Hungarian baseline sample

As a further aspect of segmentation of the general populations besides crisis perception, the survey also showed that in both Greek and Hungarian cases, the radius of trust was drastically reduced for all groups other than ‘the family.’ There was somewhat more differentiation in the Hungarian attitudes regarding such external groups (as with religious ones, contrary to nationality and political). The somewhat symbolic ‘unknown’ category reached the lowest level of trust in both cases.¹²

Trust radius as a segmenting criterion resulted in more differentiation in the Hungarian than in the Greek sample and was closely related to attitudes regarding immigration (Fig. 6.5). Respondents who were more open to other groups and external relations positioned themselves in the same grouping with Arabs besides Greeks. This kind of self-image suggested a sense of community with small countries and peoples. In the global pattern of this segment, the further sub-groups of ‘Westerners’ (Americans and Germans) as well as of the Russians were all represented with mixed attributes (the former had more positive ones corresponding to the general pattern of the Hungarian case). This ‘broad trust’ pattern can be compared to the self-image of the Greek ‘suffering’ segment, however, with a more moderate negative counter-image of the European centre.

Fig. 6.5

Two-mode network pattern by the Hungarian ‘broad trust’ segment of trust radius

4.2 Global Images in the Greek and Hungarian Media

With a focus on how the Greek and Hungarian media assimilated the 2008 financial crisis, we also examined the changes in the public images of the respective six countries as compared to the pre-crisis years. In correspondence with our survey approach, we conducted media analyses of the online press highlighting these countries/nationalities and various activity domains typically attached to them with respect to characteristic skills and knowledge styles.

Figure 6.6 shows, based on the Hungarian media, how the nations’ position changed in the post-crisis period (marked II on the visual schemes) as compared to the pre-crisis years (marked I in the network diagram) –in a network configuration derived from the two-mode data-matrix of nations and thematic domains.¹³

Fig. 6.6

A two-mode network configuration of nations and thematic domains based on the analysis of Hungarian media

From the countries/nationalities included in the study, the position of the Greeks and Germans changed significantly between the two periods. There was a slighter but readily noticeable shift of the Arabic image, while the position of Hungarians, Americans and Russians essentially remained unchanged.

The significant shift in the position of Greeks was due to no longer being linked to basically positive attributes like ‘invention,’ ‘culture & arts,’ ‘tradition,’ ‘contests,’ and ‘tourism’ while controversial or negative issues apparently related to the debt crisis such as ‘finance’ and ‘crisis’ became dominant. The German shift was brought about by a strong focus on the crisis aspects in the Hungarian media between 2009 and 2016, while ‘invention,’ ‘culture & arts,’ ‘tradition’ and ‘tourism’ fields of activity more or less related to science and learning became less cultivated. The relative stability of the pattern was particularly true for the self-reference of the Hungarian media which probably had to do with a sort of path-dependence in domestic reporting.

Figure 6.7 shows, based on the Greek media, how the positions of countries/nationalities changed in the post-crisis period compared to the pre-crisis one.

Fig. 6.7

A two-mode network configuration of nations and thematic domains based on the analysis of Greek media

The relatively stable Greek position was due to its stable relation to basically positive aspects like ‘development,’ ‘tourism,’ ‘culture & arts,’ ‘tradition’ as well as ‘protest’ and ‘finance.’ It does not mean any lack of thematic changes in the Greek media (Fig. 6.7). Generally, an array of activities from ‘invention,’ ‘technology’ to ‘contests’ or ‘corruption’ were relegated to the back by the domination of the elements of the crisis situation.

5 Discussion and Conclusions

The concluding section of this chapter sums up the evidence found for the periods under study, implying significant impacts exerted by the financial and refugee crises on cognitive maps and world imageries in the Greek and Hungarian public. These findings highlight the representation of ‘Others’ as a source of threat, particularly in the sharply politicised Hungarian public discourse; furthermore, the heavy influence of the perceptions of the economic crisis on attributing the troubles to the ‘domineering Westerners,’ with special regard to the most hard-hit segment of the Greek public. Our discussion contextualises these observations in the frames of long-range cultural-historical processes embedding the related attitudes up to 2017.

5.1 Uniqueness and Belonging

Cognitive maps emerging from national stereotypes called for a renewed application of the catnet concept introduced by Harrison White. The catnet-like nation/attribute two-mode network configurations revealed a dominant ‘Western-like’ component with both cases studied. The content of such a centre-block differed, however, by the specific attributes attached to them. While the Greek cognitive maps merged Americans and Germans into one subgroup mainly for negative traits of dominance, the same nation-set was embodied by the Hungarian pattern with a positive image of modernism.

Being a key outgroup aspect of the Greek pattern, no positive attributes were found connecting the ‘Western-like’ and the ‘Eastern-like’ clusters. The Greek pattern embodied only the national self-portrait with a longer list of positive attributes. This corresponds to the idea of ‘uniqueness’ widely shared in Greece. Also, the ‘Eastern-like’ nation/attribute set appeared in its core with the Greek profile.

The crises of the last decades brought the age-old issues of ‘Who we are’ and ‘Where we belong’ to the forefront both in Greece and Hungary. ‘We belong to the West’ stated the prime minister of Greece, Konstantinos Karamanlis, in 1979. Presenting the opposite position, Andreas Papandreou, a later prime minister, declared: ‘Greece should rather belong to the Greeks.’ A similar in-between situation was depicted in Hungary with the metaphor ‘Ferry-land.’ All this corresponds to the Hungarian historian Szűcs (1983) about Europe’s historical regions presenting East-Central Europe, including Hungary, as an intermediate one stranded between Western and Eastern Europe.

Influencing the public discourse and public opinion, self-positioning along West/East and North/South axes has been a major element of political discourse in both countries. The content analysis carried out on Greek newspaper texts data agreed with the survey-based findings of stereotypical images. The domestic reports cultivated a positively shaded theme structure (like culture, tourism, traditions and science) for the whole period covered. The noticeable shift in the images of Arabs and Russians in the media compared to the pre-crisis period also coincided with tendencies in survey related results.

The Hungarian stereotypes presented a wider dispersion than the Greek ones. The Hungarian respondents were most definitive with respect to negative features, particularly towards the Arabs and then Russians. The Greek global images’ network pattern was more polarised with regard to their cultural and political affiliations.

The contents of the Hungarian online press also exhibited some matches with the survey findings, especially with regard to the auto-stereotypes. The thematic profile of the domestic focus was close to the Greek media self-portrait. An additional similarity was that this pattern remained throughout the period. However, contrary to the initial period when the domains outlined for the Greeks and Germans were close to the domestic aspects, a significant departure occurred after 2008, resulting in a somewhat isolated domestic position.

The Hungarian configurations suggested some distancing from the East and even more from the South, while the Greek ones rather manifested opposition to some centre-image of big powers. They both referred to a semi-periphery situation, though the Greek ‘suffering’ segment even shifted to a periphery standpoint. Though a hostile image of the Centre was present in both crisis-hit cases, the self-image of the Hungarian segment was more isolated even if accompanied by some sympathy towards those from the East and South. This feeling implied some expressive attributes perceived as kin manifesting a catnet-like global pattern, sort of ‘we and those left behind.’ Also, the Hungarian segment of ‘sufferers’ exhibits a different pattern of attribution than the Greek one. The latter tends to rely on a mechanism blaming others, while the former seems to acknowledge its own part (as shown by the presence of backward in the self-portrait).

5.2 A ‘Southern’ Model

The Greek public opinion displayed particular institutional distrust and resentment following the financial crisis since 2008. However, this negative turn of attitudes had less impact on subjective well-being (beliefs of self, personal confidence, level of happiness). The term ‘tightness’ of belief organisation (Martin, 2002) helps interpret these results.

The Greek self-portraits and some ‘Eastern-like’ affinities were mostly based on the expressive traits of ‘connectivity’ and ‘self-representation’ (like ‘passionate,’ ‘debonair’ or ‘brave’ connecting to Arabs, respectively to Russians). These findings agree with Tsoukalas’ (1995) insight that the Greek national self-portraits mirrored not so much norms and values as attitudes, habitual traits and feelings.

The appearance of ‘lazy’ among the ‘self-related’ attributes may seem curious in light of the Greek public’s deep resentment regarding such a portrayal of Greeks in the European media after the 2010 debt crisis. The parallel presence of ‘hardworking’ in the ‘suffering’ population segment suggested that the traditional Greeks’ self-image element of a leisurely lifestyle has been joined by the emphasis on solid work attitudes.

The combination of the attributes ‘intelligent,’ ‘crafty’ and ‘lazy’ referred to a rule-evading habitus regarding the role of law and norms in social life. It also suggested a public awareness of the problem of rent-seeking behaviour, putting a burden on mutual confidence. As data on trust radius also suggested, trust in Greece typically relied on direct interaction and embedded relations.

In some modification of his functional differentiation theory, Luhmann (1995) outlined a ‘Southern’ model whereby universal principles of an impersonal character, such as normative contractual obligations, played a secondary role relative to the network symbolism of reciprocity.

Both the large gap between personal and impersonal trust and the particularly low level of institutional trust¹⁴ correspond to Tsoukalas’ (1995) further observation that reciprocity and expressive self-representation in Greece were stronger than normative rules, which were interpreted as a deficit in freedom.

5.3 Public Beliefs of Themselves and ‘Others’

Trust radius, another criterion of our segmentation, was closely related to attitudes regarding migrant and outer groups. Both the Greek and the Hungarian data scored relatively high on the non-receptive traits in the European comparison. This resemblance relied, however, on a different base of experiences. Based on more extensive immediate contacts, Greeks manifested more irritation by everyday inconveniences like competition in the labour market. In turn, Hungarians’ resentment against migrants and refugees was more related to abstract attitudes, a tendency likely related to a massive flow of propaganda. The Hungarian study population was significantly differentiated by trust radius. Those most receptive to other groups tended to have a ‘small-country-like’ attitude relative to hostile forces and had positive feelings for people of remote origins. Agreeing with the observations of Fiske et al. (2007), Triandafyllidou (2006) and Krzyżanowski et al. (2018), those with a narrow trust radius, who sharply differentiated between in-group and out-group relationships, tended to be more aversive to nationalities of more distant and less familiar origins.

The baseline pattern of Hungarian self-positioning clearly exemplified an intermediary image between West and East. A ‘Western-like’ cluster was seen more positively, mainly in instrumental regards and an ‘Eastern-like’ one was regarded with more criticism. The joint position with Greeks in the self-portrait was both an expression of friendly feelings towards ‘small countries’ and some appreciation of ‘Southern-like’ expressive contact-creating skills somewhat perceived as kin.

The portraits of ‘this is the way we are’ and ‘those are the ways they are’ evolve amidst an interplay of structural and cultural aspects (White, 2008). Larger groupings of national-ethnic entities get combined for perceived geographic-historical-political proximities, just like the ones for stereotypical features of skills and knowledge styles. Conforming to this conceptual duality, our two-mode network approach permitted us to reveal national-ethnic groupings and the clustering of related attributes in a complex joint analysis. Contrary to most studies of national stereotypes, our research focused not so much on distinct ethnic/national specificities, but on more general configurations exhibiting specific cultural distances. So, categories, like the ‘Westerners’ and the ‘Southern-likes’ can be interpreted in terms of certain country nodes and in terms of more or less distinctive characteristics attached to them in public beliefs.

Our segmentation’s aspects also pointed to some limitations of the stability of cognitive maps. Differing images in a given society facing each other may loosen the sharpness of stereotypical contours. However, in cases of sharply differing ideological platforms, and especially in times of crises, even an opposite tendency can occur when views of the world become even more rigid. Our analyses shed light from various angles on a polarised rearrangement of the imageries of ‘Others’ both in the Greek and the Hungarian public in the context of the developments of the financial and the refugee crises of the last two decades. Public opinion was examined amidst crisis developments in Greece and Hungary, with a focus on national stereotypes, cognitive maps and in-group/out-group social distance. The comparative analysis unravelled contrasting global images in the two countries offering new insights on native perceptions of refugees and migrants.

We expect that the combination of the catnet concept, of our theoretical basis, and of the two-mode approach, as a methodological innovation, may open the path to further research areas. On the one hand, these may be concrete empirical areas such as those of networks of public discourse or the reconstruction or interpretation of the formation of political camps. On the other hand, employing this dual approach may significantly contribute to the understanding of how micro- and macro-level phenomena may lead to divergent social-organisational patterns.

The catnet concept may furthermore serve as a comparative approach of social mechanisms, and thereby also contribute to a deeper understanding of the specificities of the societies of Southeast Europe.