A Representation of migration in Latvian mass media (2015 – 2016): Deny voice to the voiceless

The main focus of the research is the representation of migration in Latvian media. In total, 860 publications were analysed covering both Latvian and Russian speaking media content, and the following two methodological approaches were applied: the framing analysis of textual and visual content, and historical discourse analysis in order to reveal the arguments and strategies behind the justification of intolerance. 
The research data reveals that the framing of migration in Latvian media is left in the hands of politicians and officials. Economic strains and threat argumentation topoi dominate media discussions. The influence of migration is explained and approached from an economic perspective, and most frequently, the intolerance against migrants is interpreted as a failure attributed to the political elite – their inability to solve problems. Intolerance justification strategies were detected in 79% of the publications. This figure confirms that the authors are aware of intolerance not being a virtue nowadays, and the causes of it must be backed up and supported. Visual messages depict migrants exclusively as unidentifiable, dangerous, as a part of an anonymous crowd.


Introduction
In September 2015, the Latvian government was one of the last in EU to decide on the EC plan for the solution of immigrant crisis that also intended the increase of immigrant quotas (European Commission, 22 January 2016). Even though there wasn't a significant increase of refugee inflow in Latvia, the news reports from other countries didn't go unnoticed, the society of Latvia was informed about the large number of asylum seekers entering the European Union since the summer of 2015. As a reaction to this situation, some politicians and opinion leaders started an anti-immigration campaign that was built around fear (Nacionālā Apvienība, 6 July 2015), it became popular immediately within the tradition of political discourse that has strong historical roots and that have been strengthened by the recent demographic concerns.
Historically, Latvia is one of the Baltic countries that lost its independence as a result of WWII, a great part of its population became refugees or were exiled due to political persecution. Latvia re-gained its independence after the decline of USSR. The people of Latvia started a political movement -a call for freedom, and it was partially fuelled by anti-immigration arguments in 1987 (see Ardava, 2015;Kasparāns, 1 June 2018;Nacionālā Enciklopēdija, 2019). Latvia was also one of the EU countries that was hit hard by the world's financial crisis in 2008 -2009, and the austerity measures imposed by politicians initiated a new wave of economic emigration in EU (Krasnopjorovs, 2011;Hazans, 2013;Centrālā Statistikas pārvalde, 2019). It is important to recognize that even if the media representation of emigration process was in part alarming in depicting the financial and economic crisis through people's stories, it lacked an analytical or critical treatment of government actions, media were supporting government policies by not providing a wider range of arguments for a public discussion between politics and society. But what role, in general, did Latvian media play in the context of representing the immigrant crisis of 2015? What agenda did they create in the interaction with political and public agendas? And more specifically, what frames and arguments did Latvian media provide for the public discussion of migration and its political references? Our general hypothesis was that editorial strategies and journalistic work in framing and argumentation during representation of migration crisis in Latvian media was developing an exclusive and intolerant discourse because of ideological and cultural assumptions. Thus, in this research we aimed to answer the following research questions: RQ1: How conceptual frames of migration in Latvian media were constructed by linguistic and structural means of articles studied -the choice of main topic, fields of reference, word usage, choice of source of information and formulator, and headlines and leads? RQ2: What were the general discourse strategies -was the inclusion and tolerance expressed through usage of linguistic and cognitive elements, types and topoi of arguments? RQ3: How did the visual representation support the frames created and discourse in general?
Although migration has been on media agenda in various countries for decades, in 2015, as the European Union experienced the refugee crisis, issues surrounding the processes of migration and asylum seekers was a new topic on the Latvian media agenda. Since 2004, when Latvia joined the EU, the media discussion was essentially focused on the emigration of Latvian inhabitants to more prosperous EU countries.
In 2015, more than two million people applied for asylum in the EU. During the meeting of EU country representatives on the 25th and 26th of June, a decision was made that all participant countries must take part in the relocation of asylum seekers. On the 6th of July, the Latvian government accepted the national positioning project from the Office of Internal Affairs on the European Commission announcement "European Agenda on Migration" (Ministry of the Interior, 2015). On August 6, 2015, a protest took place by the Cabinet of Ministers in Riga against refugee accommodation in Latvia. Slogans heard during the protest were analysed by the Latvian Security Police, and they were evaluated as being on the verge of encouraging racism and religious disputes (Leitāns, 2015). On September 12, approximately 100 people gathered for the support of refugees joining the movement Europe Says Welcome, (Rozenberga, 2015). The participants reminded that Latvians have also once fled from the war and the consequent reprisals. A week later, there was a protest march in Riga "against forced immigrant transmission from other countries" (Skaties/LNT, 2015). One of the participants compared the duty to accommodate refugees to the Russian influx in Latvia in 1945 after the Soviet occupation (Delfi, 2015).
Paradoxically, during the period of most heated media discussions and public protests, there were no refugees of the EU quota scheme. The first two families from Syria and Eritrea, six people in total, arrived in Latvia on February 2016 (LSM, 2016). In 2016, 207 people were transferred to Latvia (Anstrate, 2016). However, by the autumn of 2016, most of the accommodated refugees had left Latvia due to the insignificant financial support (256 EUR per month) (LSM/LTV, 2016).

Literature review: media representation of migration
There a lot of studies describing complexity of the role played by media and the typical attitudes constructing image of migration and migrants. Mainly media are isolating the migrants and asylum seekers as an estranged group. Media depict immigrants as "the others" as opposed to "us", they are frequently represented as a homogeneous group, associated with 'otherness' and 'difference' (Khosravinik, 2009;Kochan, 2009;Arcimaviciene, Baglama, 2018). While migration may be associated with different groups of people and phenomena, the media representation of immigrants, refugees and asylum seekers has similar approaches, such as the generalization and dehumanisation. Research data show that media rarely mention nationality, skin tone, religion, migrant representation is characterised by "generalisation, problematization and dramatizing of migrants presence groups" (Haenens; de Lange, 2001, p.849).
An interesting aspect here is, as McKay et al. (2011) assert, that media being dependant on government announcements on asylum seeker policies brings the representation of this subject closer to Herman and Chomsky's (1988) 'propaganda model ' (McKay et al., 2011, p.610), since media, for the most part, reflect the position held by the government, and are not trying to reveal the complex reality behind it.
Comparing different studies of group representation some similarities can be recognized. Most of conclusions agree, for example, on the xenophobic discourse or racism (Wodak, 1996). As early as in 1970s researchers started to analyse this phenomenon - Hartmann and Husband (1974) had observed that the xenophobic reasoning behind one's attitude towards immigrants in the early 70s is similar to the anti-Semitic discourse in the 1920s referring to the Jewish immigration (Khosravinik, 2009). However, the current liberal and egalitarian discourses have influenced the way out-groups are being constructed -cultural evaluation came instead of one's race is in the foreground, constructing 'us' versus 'them' (Van Dijk, 1991).
Choosing migration related themes, opinions and framings, the media promote prejudicial attitudes (Cooper et al., 2017). The dedia display the discourse of nationalism with an emphasis on national rights and national identity (Saxton, 2003;Van Gorp, 2005, McKey et al., 2011, through which the physical space and prosperity of a country must be guarded. Nations are linked not only to the space they inhabit, but also to a certain national identity, which is created by the idea of a "national character" (Saxton, 2003, p.114.), and which can be affected by the accommodation of refugees. The representation of asylum seekers as "stateless wanderers" in migration debates stresses the identity instead of the human rights violation that asylum seekers have experienced (Pugh, 2004).
By analysing U.S and EU media discourses and metaphors used in the media content of recent migration representation, researchers concluded that delegitimization is a key discourse strategy, which was implemented through ideological positioning based on the dichotomy "us" and "them" (Arcimaviciene, Baglama, 2018;Bhatia, Jengs, 2018). Indeed, some research findings confirm that asylum seekers are represented in a positive light, as an asset to the society, or else -they are being victimized (Lippi et al., 2017). For example, the framing 'gifted' represents an opinion that asylum seekers possess some necessary skills. On the other hand, victimisation tends to emphasize the protection of human rights. Lippi et al. research (2017) shows that male asylum seekers "are framed negatively, but women and children are victimized" (Lippi et al., 2017, p.15). In general, media content research shows that refugees and asylum seekers are being evaluated based on their usefulness (knowledge, skills, economic advantages) or uselessness (threat, connection with international terrorism) -the potential benefits or hazards they could deliver to a particular community.
It must be taken in account that countries share some common characteristics, but, in general, the representation of the migration in media can be different in different countries and regions because of national and regional differences and the main frames traditionally built for events connected to migration. Since 90s the perception of migration in China is related to the contribution of migrants to China's economic development and the social problems that people face when moving from rural to urban areas (Kochan, 2009). Australian national media usually frames migrants and asylum seekers negatively (McKay et al., 2011;Lippi, et. al., 2017, Cooper et al., 2017, constructing migrants as "the others" and unwanted (Saxton, 2003). However, Australian regional media provide diverse information, promoting not only the official standpoint, they also cover expert opinions, personal stories that create the (self) image of the community -the building role of local media (Cooper et al., 2017).
In other countries (Belgium and most EU countries) migration is related with societal security issues, thus, media also represent asylum seekers as suspicious, linked to human trafficking, and criminals (Van Gorp, 2005). A negative image of migrants is created by the emphasising incidents and sensationalising conflict on the news, whereas everyday life is scarcely reflected (d'Haenens, de Lange, 2001). In Ireland national newsprint media represented immigration by portrayals of illegal immigrants' and most commonly used argumentation (topos) that concentrate on the 'control' of 'illegal immigration' (Burroughs, 2015).
Considering historical and socio-economic differences in EU countries where during the time period studied in this paper the solution of the refugee problem had to be found on the bases of political dialog and common base, it is essential look how refugee crisis and migration issues are covered in so called 'new' democracies or 'new' EU member countries. According to Karner and Kopytowska (2017), in those new circumstances the largest and wealthiest EU countries have responded to the dramatic increase of number of asylum-seekers by showing ethical and practical responsibility to overcome this humanitarian crisis. But in some EU border countries and in the part of Central and Eastern European countries, the refugee crisis has resumed discussions on the creation of new borders (Karner, Kopytowska, 2017). An analysis of Romanian online media representation of crimes committed by refugees, for example, shows that the media agenda was more often focused on national security related crimes (e.g. illegal border crossing) than cases of interpersonal violence (e.g. murder, rape and suicide) (Marinescu, Balica, 2018).
The role of media in depicting migrants and migration is very important because it influences (often emotionally) not only wide publics, but also political decision makers, because they use media as political showcase both to view popular attitudes and show themselves in attractive, remarkable way building their political capital. This is why the researchers that study the representation of migration often recognize the agenda -setting model (Rogers; Dearing, 1988) as usable for this kind of media analysis. To emphasize the threat caused by asylum seekers to societies, McKey et al. (2011) propose the theory of 'moral panic' (Goode, Ben-Yehudam, 1994), it helps to depict asylum seekers as deviant and outside of the normal realms of society (Cohen, 1973). To show danger of the unpredictable changes and 'social anxiety' caused by asylum seekers, the theory of a 'risk society' is applied (Ungar, 2001). Especially, framing analysis helps to define the problem, determines its causal relationships and moral evaluation (Entman, 1993); using framing, media influence public opinion and public agenda.
In the huge corpus of analysis of migration representation in media usually quantitative content analysis is combined with qualitative analysis methods to do framing and critical discourse analysis (see, for example, Kochan, 2009;Burroughs, 2014;Arcimaviciene, Baglama, 2018), which helps to identify the power relationships between the press and political actors through language (Fairclough, 2001). Media representation of immigrants and immigration as a 'problem' has been identified not only in text analysis, but in evaluating visual representation as well (Kochan, 2009;Risam, 2019).

Method and data collection
Given the differences of media formats and data sets chosen for research and aim to investigate framing in media coverage of migration during particular time period mixed data gathering and sampling methods have been accessed for the study. Research data were collected over the following time period: July 1, 2015 -June 30, 2016 covering all media types. According to short pre-research of media content, the mentioned time period reflects the most important migration related news events and public discussions. Taking in account that quality of framing and discourse analysis is depending on the diversity of available data, the sampling of this study is based on two established data sampling techniques. First, to analyse the content of the largest Latvian national media, a simple random sampling was employed. Second, in order to access research questions related data for in-depth understanding of migration representation, data purposive (purposeful) sampling method based on pre -selected criteria was used (Morgan, 1993;Krippendorff, 2004;Etikan, 2016).
According to audience data provided by Kantar/TNS Latvia, the three largest media were selected within each media segment category covering national media, providing both Latvian and Russian language content (see the media list in Annex 1). In addition, the content of national news agency LETA was chosen for this research, since its content covering local and international affairs has a significant impact on the agenda of other media.
Research data were selected by the following keywords: migration, migrants, asylum seekers, and refugees. In total, 860 publications in 12 media were chosen for the research (four dailies, two radio channels, three TV channels, two online news sites, news agency). Depending on the number of articles in each medium, the selection differed: either all articles were examined, or every 10 th article (if the number or articles exceeded 250). The selected data was analysed and measured by using SPPS software.
Applying quantitative and qualitative content analysis, texts and visual messages (photographs and videos) were assessed. The category system of content analysis included the representation of migration (topics, information sources, number of sources, genres), framing (text subject, location mentioned, the compatibility between the title and its content, words used to characterise migrants), discourse (types of discourse, attitude towards intolerance, types of argumentation, argumentation topoi, argumentation used to justify or explain intolerance). Subcategories used to quantify the latter group of categories -discourse -were elaborated drawing from the typology of topoi articulated in discriminatory discourse (Reisigl, Wodak 2001), and the typology of arguments (Walton, 2006). The overall task of the research is to test the eventual uniformity of the migration discourse, i.e., do media outlets and their information sources mention the same topics, is the supporting argumentation consequent, and are they aware of tolerance as a universal value.

Framing the migration: how did Latvian media make the nation think about migration
An important element for framing the migration is the context within which various issues surrounding migration emerged. The first factor of influence is the main topic of a media story (N=860 stories). Latvian media address migration in relation to political issues -29%, with choices of politicians and the action expected from government are linked also next most popular topics -the assistance needed/given to migrants -21%, as well as the attitude towards refugees -17%. Only a relatively small percentage of publications tend to give a negative frame in addressing crime and anti-social behaviour in the context of migration (8%); reflecting the process of fleeing a country (11%), or adding the frame of a "clash between civilisations (3%) (3% include several of the abovementioned issues, 6% discuss another topic).
The attitude towards refugees (racism, intolerance, tolerance, diversity aspects) has been discussed in all the media revised. The greatest number of publications on this topic was observed on a news program of Perviy Baltiyskij Kanal (PBK), news programmes "Latviyskoje Vremja" (Latvian Time) and on the news site Apollo (in both cases, it was 30% of all publications). In general, news sites and television looking for more emotional aspects of the story and, for different reasons, the national conservative newspaper Latvijas Avīze (Latvia's Newspaper, LA) have all devoted a relatively similar amount of materials to the attitude towards refugees. This issue was discussed at least on radio broadcasts (LR1 -6%, LR4 -5%) and in the Russian newspaper Vesti Segodnja (News Today) (4.2%).
The escape process is in focus in 63 out of 860 publications. This issue was discussed the most on the electronic media (especially, on radio LR1 (21%) and the news site Delfi (19.4%)). This topic is not widely discussed in the Russian-speaking media: LR 4 discusses it on 5% of the content, Vesti Segodnja -3%, whereas PBK news program "Latviyskoje Vremja" does not reflect it at all.
Aid or assistance were discussed the most on news sites and TV news. Crime and antisocial behaviour among migrants were addressed the most on news sites and newspapers. The greatest number of articles related to migration crisis and focusing on crime and antisocial behaviour in particular was found on the Internet news site Apollo -35%, whereas on Tvnet these types of publications were not observed at all.
Comparing publication headlines and leads (N= 860 stories), it was observed that most headlines were relevant to the content -75%; even if some cover only one aspect of the article -21%. Almost all headlines match the content in publications by the news agency LETA (95%) and radio channels' (94%), but the least compatibility was observed in newspapers (58%) and TV (70%), respectively, in newspaper articles (39%) and TV news leads (30%) mostly just one content aspect has been framed.
For the most part, Latvian media focus on migration related to Latvia (47%) and EU countries (39%). Few publications mention the conflict regions that encourage migration (3%). Most media speak about migrants as a group -91%, 6% of the articles mention asylum seekers as individuals, 3% mention them both as a group and as individuals.
One of the most crucial issues concerning the so called "migrant crisis" is the necessity of assigning an adequate term to migration. The research marked the words used in media publications describing migrants and migration. The research observed 1130 linguistic units for the description of migration. The word 'refugees' is used the most often -53%, 26% asylum seekers or beggars, 6% immigrants or immigration; 6% migrants or migration, illegal immigrants -5%, other words -4%.
In one-source publications, mostly the attitude towards migration is neutral -47%, 21% of the remarks were negative, 22% -positive; 3% -ironic, 7% -mixed attitude. In two-source publications, for the most part, one of them has been the journalist (78%) and an international institution official (24%), member of the Parliament (19%), expert (14%). The attitude towards migration in these is similar to one-source publications: a half of experts, international institution and local officials perceive migration as a positive phenomenon, the other half -as negative, whereas inhabitants' attitude is mostly negative.
The choice of sources reveals that migration is dealt with on a political and expert level; either journalists have not used refugees as source or the media have not had an access to refugees. The real actors -NGO representatives, who acquired resources to partake in asylum seeker accommodation and integration, have provided an insignificant part of the information.

Argumentation model: how did Latvian media influence the attitude towards migration
To clarify this question, we have applied historical discourse analysis methodology used to study the rhetoric of racism and antisemitism (Reisigl & Wodak, 2001). Like Reisigl and Wodak we do not share the Foucauldian and postmodernist treatment of discourse and power "which reify or personify language and discourse as autonomous, collusive actors which steer the speakers and hold the reins"; also, historical discourse analysis does not accept Teun van Dijk's socio-cognitivism, which overemphasizes "a top-down causality of opinion making and manipulation". We presume that the media outlets siding with various ideological stances and representing different editorial policy strategies or even different journalism cultures would produce different reports about the refugee crisis. To expose heterogeneity of the refugee discourse we rely more on a quantitative analysis permitting the cross tabulation of data.
Arguments are statements used to convince or persuade people about the correctness of an opinion. Arguments are grouped in two categories. Rational arguments are premised on logic and sure facts, examples, testimonies or generally accepted information. Plausible arguments are statements that 'appear to be true in a normal type of situation that is familiar both to the participants and the onlookers' (Walton 2006, p.71). Following Douglas Walton, we call them plausible rather than fallacious arguments because in some cases they might be accepted as sound statements producing a sought persuasive effect. Examples of plausible arguments are appeal to expert opinion, argument from popular opinion, from analogy, from commitments, from ignorance, ad hominem arguments. In this content analysis, an argument from analogy was treated as a distinct subcategory because we presumed that historical events, cultural characteristics would be extrapolated to current and future events.
The incidence of types of arguments is summarized in Figure 1. Rational argumentation dominates in the whole sample but there are dissimilarities across the media. Plausible arguments (including argument from analogy) dominate in PBK, mixed arguments in Vesti Segodnja (Figure 2). Plausible argumentation is widespread also in different other media LETA, NRA, TV3 and Apollo. A large share of mixed argumentation is present in LR1 newscasts. Arguments from analogy are popular in LA, Tvnet and PBK. It seems that the types of arguments are rather a kind of editorial decision and it depends on the type of media outlet and its ideology. Public service media (LR1, LR4, LTV1), for example, show a bigger share of rationality, however, also mixed arguments are not rare there.  By stating their arguments journalists and their sources expect that the audience will make sense about the refugee crisis and the refugees. To support a desired conclusion the authors, provide specific conclusion rules or topoi (Figure 3) that connect the arguments with the claim. Topoi point at the place where the conclusion is to be drawn from. In other words, topoi are the contexts in which the audience is expected to make sense of the mentioned arguments.
In all the sampled media the topos of burdening dominates -it was used in 27 percent of publications. If the refugees burden the state or institutions, the burden is to be removedthis is the message's logic. Appeal to remove the burden is addressed to politicians who decide on admission quotas. As it was shown, already the political topics dominate in the media and mostly these reports cover the discussion of admission quotas. Therefore, we conclude that the burdening topos frames the domestic political rivalry suggesting the audience a simple conclusion: 'some politicians are unwilling to remove the burden we have warned them about'. The burdening topos dominates in dailies Diena, NRA and LA, and news agency LETA -here it almost twice the frequency of other topoi (33, 26, 24, 40% respectively). Taking in account the usage of similar topoi -finances (12%) abuse (9%), and uselessness (3%), almost half of all complaints are somehow related to economics. Refugees are not those who could have boosted Latvian economy but, on the contrary, they deprive the local residents of their share of resources, they abuse the provided material aid and they are useless as labour force. In a country recovering slowly from the financial crisis, such economic issues arrange a salient frame manipulating the attitude to the admission of refugees.
Danger is the second most popular topos discrediting the admission (21%), especially it is pronounced in Latvian-language portals Apollo, TVnet, Russian-language TV channel PBK and newspaper Vesti Segodnja (55, 46, 40, 38% respectively). A very different image of reality is presented in the evening newscasts of the public media -Latvian Television and Latvian Radio (LR1 in Latvian, LR2 in Russian). LR1 and LTV stressed humanitarianism -appeals to avoid unequal treatment and discrimination (appears in 48% and 22% news stories respectively). Often these channels talk about our duty to find solutions to the problems -this is the topos of responsibility (29% and 41%). In the internet portal Tvnet.lv it is even more pronounced (every second publication); however, taking in account a high occurrence of the topos of danger, responsibility in this case might refer to finding solutions how to stop the admission of refugees. LR1 and LR4 journalists (36% and 32%) developed the topos of finance. In this case, its usage was not discriminatory -the journalists discussed the sources of subsidies to the refugees. Both radio channels were concerned with the legal framework: the topos of law was included in 30 and 45% or news stories respectively.
Manipulation with numbers was typical of Tvnet (59%), Apollo (35%), Delfi, PBK (23%), Vesti Segodnja (23%). Decontextualized references to numbers of refugees, amount of government subsidies etc. amplifies the scale of the problem but does not explain nether the causes nor the consequences of the crisis. Name-interpretation was popular in PBK (27%) and LR1 (21%) news. In this case the literal meaning contained in the name assigned to the group is ascribed to the persons in question. Naming strategy employed by PBK carried a negative interpretation ("illegal migrants") whereas LR1 neutral or positive ones ("shelter seekers", "women and children"). Tvnet.lv (59%) and LTV (52%) more often than other media invited to accept the reality and find specific solutions. Table 1 summarizes Top 3 topoi used in each sampled media. In case of coincidence of occurrences, more than three topoi are listed. The topoi of burdening and danger are common to all newspapers as well as the news agency LETA. Latvian Radio refers to 'objective' factors like those that financial and legal framework required for the admission. Topoi of danger, numbers and reality are common to the Internet portals. Commercial television newscasts mention danger, burdening and numbers. In the content of public LTV1 more estimable stories about Latvia's role in the refugee crisis resolution was found: journalists invite their viewers to accept reality, be humane and take on responsibility for those who suffer. The category "Discourse type" discloses whether the opinions expressed in a publication suggest that the refugees be excluded from 'us' categorically or there exists a possibility of inclusion. Both discourse types are equally frequent in the media (31% and 33%). Every fifth publication (19%) represents opposite opinions, and in many cases, the coders were not able to include the discourse type in a defined subcategory (18%).
Especially PBK, NRA and LETA (60, 50, 49%, Figure 4) promote exclusion. This discourse type is dominating in LA and Vesti Segodnja however; it occurs less than in three mentioned media outlets (35 and 39). Again, three public television and radio channels take a pro-refugee stance: exclusive discourse type was found in 3-7% of the news stories while the confrontation of opposite opinions (21-40%) dominates. The latter tactics was found also in the daily Diena (25%). A striking characteristic of the LTV-1 newscast Panorāma is the outspokenly inclusive discourse (51%).  The media report opinions held by their selected sources -be they tolerant or intolerant. The editorial office can manifest its own attitude ( Figure 5) to the expressed propositions using a variety of techniques: framing the attitude in the title and lead, adding a suggestive illustration, putting the interviewee supportive or challenging questions or reminding about the unacceptability of racist statements. LR1, LR4 and Apollo used to explicate their disapproval of intolerance (54, 47, 54% of all publications in which the editorial office expressed its attitude). Approval of intolerance is a characteristic of PBK, LETA,TV3 and Vesti Segodnja (67,43,43,and 41%).
In contemporary diverse and pluralist societies, intolerance is not an acceptable value. The information sources use to mitigate their negative attitude to the refugees. We have found the mitigation strategies in almost half of the sampled articles, audio and audio-visual reports. At the beginning, we assumed that, given the popularity of the idea of cultural nationalism in Latvia, the incompatibility of cultures would be among the most popular mitigation strategies. Nevertheless, the occurrence of the topoi (Figure 6) pointed at the concerns with the issues of economy, while the dominating mitigation strategy turned out to be the political elite's incompetence resolving the refugee crisis (41%). The result can be explained by the fact that the sampled media framed the refugee crisis as the political contestation of admission quotas and Latvian politicians' failure to stand for the country's interests in the European Commission. This mitigation strategy dominates in Diena, LETA, Vesti Segodnja, TV3, LR4, LR1, PBK, and LA (Figure 7). Lack of resources was mentioned in 16% of the cases, which does not fit the frequency of the economy related topoi. This implies that the information sources did not intend to clarify whether the country possessed a physical capacity to provide shelter but rather they used an opportunity to attack politicians who lacked competence even to explain the current state of affairs. The threshold of tolerance ('there are phenomena we cannot accept') was mentioned in 17% of all mitigation strategies used. The incompatibility of ideologies, religions, cultures, and mentalities was referred to in 16% of the cases. Lack of information about others and ghettoization ('refugees will create closed groups rather than integrate in our society') were the least frequent mitigation strategies (13% and 8%).  Thus, the mitigation strategies were found in 79% of the sampled publications. Despite the large occurrence of intolerant and exclusive discourse, this indicator can be treated as a positive trend. Namely, the information sources are aware that intolerance is a negative value; striving to excuse themselves, they shed all responsibility on the domestic political elite. Invincible "objective" and "inherited" obstacles -like culture, religion and the concomitant threshold of tolerance -were called in 32% of mitigation strategies. The latter is especially frequent in the publications coded as ones favouring intolerance (Figure 8). Therefore, the authors assert that intolerance is a strong characteristic independent on subjective will. When the authors disprove intolerance, they explain its persistence with lack of knowledge about others implying that individuals can change their attitudes improving their cognitions about the diversity. Publications expressing neutral attitude justify intolerance by lack of resources mostly. Nevertheless, the political elite's incompetence is the dominating mitigation of all attitudes to intolerance.
Usage of mitigation strategies suggests that the negative attitude to refugees cannot be explained in terms of 'civilization conflict' or lack of resources only. The issue can be instrumentalized seeking to attain other communicative goals. In this case, it staged the contestation of power among political actors who are represented as ones striving to remove or not the burden imposed on Latvia by the European Union bureaucracy and successful welfare states like Germany. At the same time, public communicators exploit shifting the blame on the political elite as a justification of their intolerance.
Cross tabulation of the content analysis, data proves heterogeneity of the discourse about the refugee crisis across the media. Use of topoi and mitigation strategies demonstrates that the topic was mobilized to frame the issues of contestation of power -both in Riga among domestic political actors and between ones in Riga and Brussels -rather than to discuss the economic, cultural and humanitarian consequences of the refugee crisis.

Image of the migration: how crisis took over the migration
In contemporary media, the image is becoming more and more an essential framing element both as the illustration of the text and the independent channel of information. It is important to understand the focus of the visual representation on migration issues that as this study shows is rather alarming because is more concerned with depiction of crisis, less being about migration and migrants themselves. For the research, 492 images were selected: Apollo 20, In the publications on refugees and migration processes on Latvian media, in 37% of the cases only the locals have been displayed, and in 50% of the cases the connection with the local society is not established. These publications often include photos of politicians and officials for an illustrative purpose.
Avoiding true and deep visual stories about the migration process and refugees could signify both resource saving and purposeful construction of visual framing, which excludes third world country citizens. This framing encourages refugee and migration process interpretation in the society as disconnected from everyday life in Latvia and establishes the refugee status as "others". A few exceptions are observed on the media, e.g., NRA -(30%), LA -(46%), LTV -(34%) images show refugees and those involved in the migration process being together with the local society. This indirectly points to the connection and relationship formation, as well as non-hierarchical local and third world inhabitant display on the media.
Only 21% of the cases identify the represented individuals in the publication or the image description, and they are theoretically identifiable in 31% of the cases. Mostly, the depicted people are identified on Tvnet (68%) and LTV (62%). In 28% of the cases, the depicted persons are unidentifiable, as they have been photographed from a distance or in a crowd. Most cases of unidentifiable persons were found on the Russian language newspaper Vesti Segodnja (92%).
Paternalistic display -objectification -of third world inhabitants is one of the situations how the depicted persons are categorized as being different, without identity, and the crowd, where all persons have similar traits and characteristics, which are determined by verbal discourse and recipient framing. Third world inhabitant framing as a group or crowd frames them as an amorphous mass, devoid of personality, and it is one of the most typical ways to attribute threat and danger connotation. It is historically accepted that violent behaviour is more characteristic to men; in a patriarchal society, women play a passive role, hence, a display of a young men's crowd creates a greater sense of threat than a women's crowd. Latvian media depicts mostly men (42%), whereas women (14%) and children (9%) are represented less. The display of men is dominant on the Internetnews portals Delfi (55%), Apollo (55%), Tvnet (54%) are dominating here.
The gaze of a person in a photo symbolically represents the social interaction and the power relationship between the observer and the displayed person. A look at the camera symbolises awareness of the presence of the observer/photographer. Later, publishing the image, it attributes equal power relation to both person and recipient involved in the observation act. However, a look away from the camera symbolises a state of social dominance or control: "we" see "them", without "them" being aware of "our" observations. It creates a state of power relation where "we" can avoid "their" gaze, but "they" cannot escape "our" gaze. Migration related publications on the Latvian media, in 60% cases; the displayed persons' gaze is directed away from the camera and symbolically places the recipient in a dominant power relation to the displayed person. Most frequently, the look is turned away on Russian language media PBK (92%) and Vesti Segodnja (90%).
Picture descriptions must explain what is displayed in the picture, identifying the people, as well as provide information that allows figuring out the connection between the picture and the verbal context of the story. However, in 29% of the cases the used photograph was not linked to the text. Illustrative photos, coming from image stocks, photo agencies and social media, are most frequently found on PBK (68%), Apollo (50%).
In a situation, where politicians and journalists provide most information to the media, unsigned photos represent third world country inhabitant exclusion from media agenda. Exclusive media framing provides the recipient with a deceiving view at the third world inhabitant display and encourages stereotypical visual framing, which is more based on imagination and previous visual experience.
Upon beginning of the research, it was assumed that the technical solutions of image post-processing, which are popular in photojournalism, would also be encountered in the Latvian media. Coding the photographs, dramatization effects were sought for, as they would indicate the desire to use colour and tone scenography characteristic to mass consumption film aesthetics in a third world country inhabitant representations. Drama-tization effects do not contradict the ethics of photojournalism; however, they signify the construction of the image and encourages the recipient to frame them according to his/ her previous aesthetical experience.
Dramatization effects are used only in 12% of the images -mostly on NRA (36%), LA (29%), Diena (25%), TV3 (17%) stories. In 85% of the cases, no use of dramatization effects was found. Considering the large number of illustrative images (29%), images of unidentifiable source (7%), image content where locals are displayed (37%) and situations with no connection to the locals (50%), the absence of dramatization effects can be explained by the local visual journalism separation from Western journalism traditions. Firstly, the photo is not perceived as a serious informative media, but as an attention grabber and imagination stimulus for the recipient. Secondly, it reflects the local media management priorities, namely, purposeful resource economy, choosing free images for the stories. Thirdly, it points out a mindful avoidance of provocative and scandalous high-level photojournalist coverage, protecting themselves from recipient counter reactions.

Conclusions
The conclusions were made according to the research questions asked: RQ1: Latvian national media follow the political process development in Latvia and the European Union when it comes to migration representation. The choice of topics is limited; it includes issues of political decisions. Officials of various levels and politicians, who define both topics and attitude towards migration, create news content. Only a few media have found a way to speak to the refugees or "actors" -the organisations that took on the actual task of asylum seeker integration.
Latvian media speak of migrants as a group. Exclusive strategies are also used in the visual message choices, as migrants are often depicted as unidentifiable, dangerous, as a part of a large group. Media republish politician remarks, which are illustrated with politician portraits, whereas seldom have the media had the chance to interview migrants, display individual life stories and situations in context of the migrants' represented countries. Therefore, a disproportion of information source and message diversity has occurred, rendering the reflection of migration process in a general way, thus lessening the possibility for inhabitants of Latvia to acquire knowledge and experience of asylum seekers and migration circumstances.
RQ2: Representing migration, the Latvian media have mostly relied on rational arguments. Probability argumentation is the second most popular argumentation type, which is based on a lack of knowledge, the tendency to reflect events populistically, using argumentation against a person, including potential consequences and threats. A relatively large number of arguments have been created also on the basis of the analogy to other historical events or culture and religious aspects.
In all media, when speaking about migration, the burden topos is dominant. It means that if issues caused by refugees place a burden on the country or any institution, this burden must be got rid of. Despite the large proportion of intolerance and exclusive discourse in the research selection, the use of intolerance justification strategies signifies a positive tendency. The authors of the opinions are aware of intolerance not being valued nowadays, and its causes must be explained. Most frequently, the authors in the actions of the political elite see the cause of "our" intolerance. The unsurpassable obstacles, an "objective" and "inherited" habitus (created by culture, religion and difference tolerance threshold), therefore uneasily changeable, have been mentioned in 32% of the justification cases. Especially interesting is the fact that exactly the publications that have been determined as intolerance encouraging, authors use references to the habitus (i.e. "natural tolerance threshold, culture, religion, mentality characteristics) as a justification strategy. Thence, authors of intolerant remarks confirm that intolerance has a deep-rooted cause little depended on human will. This is explainable when taking in account that immigrants and refugees where often mentioned in media in the context of criminal cases and brought in to wider scheme of Muslim terrorist as frame of reference -the other, the scary unknown, the alien -media representation of an immigrant is rooted in discourse of the other as impossible to understand and predict, in a way, a difficulty of enacting it in the media story as an agent. So, the immigrant is still a stranger -even if the attitude towards intolerance is negative, authors justify its existence with lack of knowledge. In publications where, neutral attitude towards intolerance is expressed, it is mostly justified by lack of resources.
Use of justification strategies indicate that the negative attitude towards refugees cannot merely be interpreted in the context of "civilisation conflict". The topic of refugees can serve as a tool to reach other goals. In this case -to stage a political dispute between parties who attempt or do not attempt to get rid of the burden on Latvia.
RQ3: The visual representation of the migration during crisis as it was studied in the material taken from mainly photographs published with articles as illustration or in some rare situations as an independent visual coverage showed that in most cases politicians and representatives of different organisations involved into regulation of the crisis were represented. If the picture depicted migrants themselves -objectification took place with technical and semiotic means of photography and/or men in chaotically filmed crowd were typical for this cluster of pictures. Little number of persons depicted (excluding local politicians) could be identifiable to audience. A relatively big proportion of all those pictures were taken from (foreign) data bases and had, in fact, no links to Latvia. However, dramatization effects were used only rarely.
Thus, our hypothetical assumption that editorial strategies and journalistic work in framing and argumentation during representation of migration crisis in Latvian media was developing an exclusive and intolerant discourse because of ideological and cultural assumptions was only partly approved -even if exclusive discourse based on cultural panics and nationalist ideology played a remarkable role, overall discourse in media was more inclusive than exclusive, economic, financial and emotional factors played bigger role than the cultural and ideological due to the main focus on local and EU political agenda and acting of government authorities. Even if the main formulations came from journalists (along with mistakes in terminology regarding migration), different level officials influenced the mainly rational discourse more than experts or refugees themselves who were depicted and described mostly as aliened objects. It is possible further to make a conclusion that editorial strategies were driven by either journalistic culture of the media outlet or marketing interests and show the lack of knowledge about migration in broader sense. This is why the media followed a political agenda even if the political inability to take a decision was critically viewed.
Mostly tendencies discovered in this study are typical for all media. It is difficult to find significant differences between different types of media or media using different language. However, some differences found can be explained by culture (forms and styles of communication expected and supported by certain audiences) of a media outlet, ideology and by editorial program. Public media and largest news websites provide more professional, rational arguments and balanced information on migration. Largely coverage of migration crisis by commercial media was driven by the alarming tone, little in-depth analysis and formed in reaction to public moods to be attractive on media markets. The quality of representation of migration crisis is connected to an editorial decision about the quality of journalism in particular medium -if the quality journalism is a concern -variety of genres, sources, formulations and attitudes depicted is bigger, analysis and reports from places of events -more presented, even if the ideological stance of an editorial board is clear -like it is with national conservative paper LA. And vice versa, when the editorial program and media culture is openly market driven then media create fear and hate -few particular media (PBK, Vesti Segodnja, NRA) in migration representation clearly showed a desire to express -through selected sources, frames created and claims made -more intolerant and exclusive attitude towards migrants, displaying them as a threat and burden.
Declaration of Conflicting Interests: The authors of publication declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding. Media content research was carried out within the project "Development of a responsible, diverse and high-quality journalism in Latvian national and regional media, encouraging third world country citizen integration in Latvia" financed by the EU Asylum, migration and integration fund.