Discursive shifts and the normalisation of racism: imaginaries of immigration, moral panics and the discourse of contemporary right-wing populism

ABSTRACT Looking at mediated, political and wider public discourses on immigration in Poland since 2015 and exploring these in the context of the country’s right-wing populist politics, the paper develops a multi-step normalisation model which allows analysing how radical or often blatantly racist discourse can not only be strategically introduced into the public domain but also evolve into an acceptable and legitimate perspective in perceptions of immigrants and refugees. The paper highlights the strategic as well as opportunistic introduction of anti-immigration rhetoric in/by the political mainstream in Poland in recent years, often on the back of the so-called post-2014 European “Refugee Crisis”. It explores normalisation as part and parcel of a wider multistep process of strategically orchestrated discursive shifts wherein discourses characterised by extreme positions have been enacted, gradated/perpetuated and eventually normalised as an integral part of pronounced right-wing populist agenda. The paperfurthers a view that normalisation entails the creation and sustainment of a peculiar borderline discourse wherein unmitigated radical statements are often married with seemingly civil and apparently politically correct language and argumentation. The latter are used to pre-/legitimise uncivil or even outright radical positions and ideologies by rationalising them and making them into acceptable elements of public discourse.


Introduction
"More than half of Italians in poll say racist acts are justifiable" (The Guardian, 12 November 2019) and "Under Matteo Salvini, 45 per cent of Italians say racism is justifiedfar-right extremism has been normalised" (The Independent 18 November 2019) 1 were among just a few headlines of UK and international newspapers and online news in November 2019. According to these, recent polls across Europe have shown that not only are racially-motivated attacks and abuse currently on the rise, but also that the wider "regular" public now increasingly accepts racism as both justified and acceptable social attitude.
article looks specifically at the normalization of radical, racist and exclusionary discourse in the Polish public sphere and argues that not only the introduction of that discourse but also its subsequent, skilful and strategic as well as purposeful recontextualisation supported the fact that, though once considered radical and extreme, racist and xenophobic views evolved into an acceptable, "rational" and largely legitimate perspective in the Polish public's perceptions of migrants and minorities.
Looking into various patterns and stages of the eventual normalisation of such extreme views in Polish public discourse of late, this article points to the salience of various strategies of both discursive and generic normalisation (see below) which supported wider patterns of factual legitimising or imaginary "pre-legitimising" (Krzyżanowski 2014) of the newly-introducedyet eventually perpetuated and normalised (see below)discriminatory views. Indeed, given the fact that this new and strongly intensified anti-immigration discourse in Poland has developed despite very limited numbers of migrants and/or refugees arriving in the country in recent years (Triandafyllidou 2018)especially if compared to many of its EU counterparts, such as Sweden or Germany (Krzyżanowski 2018b)it seems particularly relevant to foreground how strategies of creating various (untrue) imaginaries of migration and the wider construction of immigration-related "moral panic" have been deployed in Poland and eventually dictated the wider trajectory of, in most cases, newly developed immigration-related discourses.
Recognising the complexity of context-specific trajectories of public discourse and various patterns of its strategic "engineering" (Fairclough 1992) by power-holding actors such as politicians or the media, this study further develops a multi-step model of "discursive shifts" (Krzyżanowski 2013c(Krzyżanowski , 2018a. It shows that, further to their initial enactment and introduction, many public discourses are nowadays strategically gradated/ perpetuatedor spread and settledin the public domain before eventually becoming normalised, thus causing not only a change in language but also in wider norms and patterns of perception and discursive representation/constructionof social issues and groups (see Krzyżanowski 2020 in this Special Issue). As the article shows, following such a multi-step trajectory, many publicand in particular media and politicaldiscourses effectively normalise these often outright radical and unacceptable patterns of public expression as a constituent part of pronounced right-wing populist political strategies. This is often achieved via the creation of a new form of public borderline discourse (Krzyżanowski and Ledin 2017 and above), wherein unmitigated radical statements are often married with civil, quasiacademic and politically correct language. These are eventually used to pre-/legitimise (Krzyżanowski 2014(Krzyżanowski , 2016 the effectively uncivil, radical and extremist positions and ideologies to make them look like rational and acceptable elements of increasingly exclusionary and nativist "common sense".

Imaginaries, crises and "Moral Panics"
In their traditional and somewhat "innocent" understanding, social imaginaries may be seen as the multiple ways in which "people imagine their social existence" (Taylor 2004, 23) and do so on the back of envisioning sameness with and difference from "others". Social imaginaries, hence, help to imagine "how we stand to each other (…) how we relate to other groups" (Taylor 2004, 25), thus allowing for the articulation of social identities, in both the present and in the future projection of social order. Indeed, imaginaries almost always entail a combination of the recognisable, or real, "present" and an imagined "future". Hence, "while a strong element of fantasy may figure in the imaginary, there is likely to be a significant "real", verifiable element as well" (Wolin 2008, 18;original emphasis). Or, put differently, even if one considers imaginaries to be largely utopian, their character should be seen as "not just a dream to be enjoyed but a vision to be pursued" (Levitas 2011, 1).
However, more often than not, imaginaries are strongly ideological constructs and visions/conceptions of society enacted in line with pre-defined political goals aiming to reproduce and sustain power. Imaginaries, hence, often carry visions of social, political or economic order by arguing for its alleged "crisis" (see e.g. Krzyżanowska and Krzyżanowski 2018), whose specific function is to bring an alarmist tone to descriptions of the "present" social reality and its apparently critical state. Crisis imaginaries hence often cease to serve as actual descriptions of the "real" and become to large extent ideas invented strictly for "political" motives (Sum and Jessop 2013, 396). Their aim is to formulate visions of the future which, as such, may never come to fruition yet effectively yield a significant degree of legitimacy to current actions of (powerful) actors essentially geared towards controlling the present. They thus allow the construction of often outright utopian ideas which serve as a "means of legitimising the particular social arrangements prescribed" (Levitas 2011, 214), and they do so whilst strategically not making "explicit the values involved in particular constructions of individuals and societies" (ibid.). For example, the recent case of "Brexit" in the UK is certainly a case of such an ongoing constructedand reconstructedcrisis. Its aim is not to strive for the common good or a better collective futureas these remain particularly uncertain and vaguebut, by using utopian future projections, to grant legitimacy to current political elites and their present actions (Krzyżanowski 2019;Zappettini and Krzyżanowski 2019).
Enacting and constructing crises is, however, nothing new in publicand especially political and mediadiscourse, which traditionally carries heavily ideologized visions of the collective present and future. But crises are also enacted discursively in the public domain by targeting specific social groups, in a process of "scapegoating" (Wodak 2015) or constructing a social "stigma" around them (Tyler and Slater 2020). Such agent-centred construction of imaginary has, since the 1970s, often been referred to as the "moral panic" (Cohen 2011;Thompson 1998), i.e. a process of creating a web of (often untrue) information around specific social actors and groups which, by means of their recontextualisation between media and political discourse, becomes widely spread as negative opinions and as more or less core, albeit stereotypical, elements of common sense (Hall 1978).
In his widely acclaimed "Folk Devils and Moral Panics", published in 1972, Stan Cohen described the process of enacting moral panic as strategic and intentional, as well as political ideological, and to some extent also cyclical in nature. As he argued, whenever moral panic is enacted A condition, episode, person or a group of persons emerges to become defined as a threat to societal values and interests; its nature is presented in a stylised and stereotypical fashion by the mass media; the moral barricades are manned by editors, bishops, politicians and other right-thinking people. (Cohen 2011, 1) However, while such a construction of moral panic can be seen as short-term, Cohen very rightly argues that it most commonly is not. Moral panic hence has the immense propensity to outlive the immediate spatial and temporal contexts in which it is constructedwhich to a large extent makes it similar to the logic of many "crises" (Koselleck 1996;Krzyżanowski 2009Krzyżanowski , 2019 and thus contributes to the long-term stigmatisation of targeted individuals and social groups. Or, as Cohen himself puts it, moral panic can have "more serious and long-lasting repercussions and might produce such changes as those in legal and social policy or even in the way society conceives itself" (Cohen 2011, 1).
Such longevity of moral panicas well as its universality and applicability in various contexts further to those of its original enactment or inceptionis possible due to its largely discursive character, as well its reliance on re-mediation. This allows moral panic to be based on a peculiar "spiral effect" (Thompson 1998, 26), which turns moral panic into a cumulative and a growing aggregation of negative opinions and attitudes about a certain phenomenon or group. This effect allows the expression of outright "hostility toward the group or category regarded as [a] threat" (Thompson 1998, 9). It also demonstrates the normalisation and acceptance of often purely imagined (negative) attitudes, the enactment and perpetuation of whichfrequently in a purely imaginary wayis often among the key ideological goals of those behind moral panic (Hall 1978).

Immigration and the normalisation of racism: discursive shifts
Probably more often than any other social issue, immigration has certainly been the most common object of many social and political imaginaries and the target of many nationally specific as well as cross-national moral panics. Already, the classic critical-analytic work on media and political discourse on immigration has shown us explicitly how various types of "immigration scares" are constructed and spread on the back of both explicit and mitigated racialised opinions and views (e.g. Krzyżanowski and Wodak 2009;Matouschek, Wodak, and Januschek 1995;van Dijk 1991van Dijk , 1993Reisigl and Wodak 2001;Richardson 2004;Santa Ana 2002;Wodak and van Dijk 2000). The same is true of more recent studies, which clearly show that immigration remains the key object of contemporary populist "politics of fear" (Wodak 2015), i.e. a form of over-emotionalised "anxious politics" (Albertson and Gadarian 2015) wherein creating anxieties around immigration and constructing it as a threat remains central to right-wing populist and other imaginaries of society, politics and the economy (Wodak and Krzyżanowski 2017).
By the same token, many analyses, including recent ones, show that the traditionally negative path-dependencies in speaking about immigration in the public domain remain livelier than ever (Benson 2013;Haynes, Merolla, and Karthick Ramakrishan 2016). The usual frames of securitisation of immigration (i.e. showing it as a danger/ threat) or its economisation (representing it as a burden/ cost) are still among the key ones in public discourse and practice (Buonfino 2004). In a similar vein, patterns of politicisation and mediatisation of immigrationnow often strategically used in combination (Krzyżanowski 2018a(Krzyżanowski , 2018bKrzyżanowski, Triandafyllidou, and Wodak 2018)put fears, threats and wider moral panic around immigration on top of many of the ongoing, politicised doctrinal conflicts (almost irrespective of their actual reasons and foci). They also become conservative and nativist media organisations' whip against liberal and pro-immigration activists and politicians. In this way, the new, strong anti-immigration voices embodied by, for example, the illiberal online uncivil society increasingly draw on strong xenophobic and nationalist framesor even their radical, for example Islamophobic, and other variantsoften blatantly borrowed across national boundaries and contexts (Krzyżanowski and Ledin 2017).
But while xenophobic, anti-immigration attitudes have indeed become widespreadin particular in several countries with a history of, and debate about, immigrationrecent years have seen a very significant new dynamics. On the one hand, in contexts where discursive traditions of talking aboutand imaginingmigration have a long tradition, we have seen a rather obvious radicalisation of immigration-related voices and views and a perpetuation as well as eventual normalization of often discriminatory views in the public domain, often on the back of recurrent moral panics that keep "alive" various facets of immigration scares (e.g. UK and anti-immigration voices in the context of Brexit). On the other hand, in countries with a (previously) limited intensity of debate on immigration and related issuessuch as the Polish case analysed belowwe have not only seen the introduction of immigration as a new topic in the public domain, but also the rather immediate and evident radicalisation of immigration-related views with strong anti-immigration rhetoric and attitudes becoming part of accepted patterns of speaking and thinking about migrants, refugees or asylum-seekers.
In order to capture many of these dynamics, this article proposes a multi-step model of discursive shifts (Krzyżanowski 2013a(Krzyżanowski , 2018a. The latter helps exploring how various macro-level, transnational frames of imagining society (e.g. populism, neoliberalism, radicalisation, securitisation etc.), otherwise known as "discursive change" (Fairclough 1992), are recontextualised in actor-and context-specific discourses of various public, and in particular political and media actors. However, as seen here, discursive shifts are considered not only from the point of view of their "vertical" relationship to macro-level discursive change and various patterns of its meso/micro level accommodation. They are also seen here, namely, in a "horizontal" order which recognises various stages/steps of how such shifts are gradually introduced, spread and pivotal for not only transformations in public discourse as such but also in wider array of norms of social, political and economic conduct (e.g. in relation to immigration, see below).
As proposed here, discursive shifts are viewed as consisting of three stages whichthough necessarily overlapping and not drawn with rough edgesrecognise the complexity of how discourse is introduced into the public domain (in most cases via the recontextualisation of existing local/ global discourses), how it then further spreads and settles across the public domain and various social fields along with their key spaces and genres, and how it finally contributesand serves as evidence ofto deeper change in norms of public expression. More specifically, those stages are (see also Figure 1): . Enactmentthe first phase of a discursive shift wherein discourse is strategically introducedor enactedby means of recontextualisation into the public domain with "new" discursive elements, including themes and arguments, as well as various linguistic means which support their realisation. This is the stage where the key link is established between the "global" level of discursive change (and transnational frames such as insecurity/threat, radicalism etc.; see above) and the context-and actor-specific discourse at hand, and where a promulgation of wider ideas and concepts with context-and actor-specific meanings occurs. . Gradation/Perpetuationin this second stage, not only does a combination of discursive elements take place to form a "new" discourse, but so does its further diffusion and reinforcement across various social fields, spaces and genres. In line with a hierarchical view of recontextualisation (Bernstein 1990;Krzyżanowski 2016Krzyżanowski , 2018a, the ordering of discourse is undertaken here in the course of the recontextualisation process running between "primary" context/s of discourse (where the discourse was enacted, e.g. in the political field), its recontextualising or "carrier" discourses and contexts, and its secondary i.e. "target" contexts. This multistep process helps diffusing discourse into new social fields by gradating iti.e. augmenting and reinforcing key arguments and features -and by hence allowing it to settle in the new context/s where it is effectively perpeatuated. . Normalisationthis third and final stage serves as evidence that not only has the discourse as suchalong with its key elements like themes, arguments and supporting meanschanged in the course of a discursive shift, but also that the latter has contributed to a deeper change of norms of public expression. The existence of widespread "borderline discourse" (Krzyżanowski and Ledin 2017) links elements of previous (e.g. civil) and new (e.g. uncivil) discourse, as well as emphasising the development of new norms wherein radical and unacceptable views are widely used yet while being strategically "clad" in quasi politically correct and acceptable themes, arguments and statements. Strategies of, for example, rationalisation of radical ideas and views (e.g. racism, xenophobia etc.) or pre-legitimizing discriminatory actions against specific social groups (e.g. immigrants) by means of future projections and imaginaries are also typical evidence of the radicalisation of views as well as their evident normalisation. Whereas the first and second stages of discursive shifts described above have been elaborated in the earlier conceptions of that model (see, in particular, Krzyżanowski 2018a), the third stage of discursive shift, i.e. normalisation, requires further explanation. As evidenced by its definition above, normalisation is conceptualised here alongside the key theoretical and empirical "discursive" approaches proposed to date (see Krzyżanowski 2020 in this Special Issue). Hence, on the one hand, the view of normalisation deployed here draws extensively on Foucauldian conceptions, and in particular on the view that normalisation is rooted in a wider, social reordering of discourse and is a part of regulating the introduction and legitimisation of new norms, indeed often in stark contradiction to existent and acceptable social values. In a similar vein, Foucault's devotion to imaginaries as part of normalisation "models", along with the centrality of the distinction/classification of social actors and groups in the process, is just as crucial here to emphasise how normalisation is formed discursively and how it relies on imaginary distinctions and difference-building processes. Such discursive displays of normalisation are also among the key aspects of Vaughan's (1996) conception followed here. Hence, just like Vaughan, this paper argues that normalisation is enabled as well as eventually evidenced by the development of a peculiar language of normalisation which skilfully introduces newand often obviously deviantsocial norms. Similarly, normalisation is viewed here as an inherently gradual process too: ranging from the introduction of new norms and the (gradual) eradication of old ones, through the legitimisation of a new normative order, down to acceptance of what was previously deemed "unacceptable" or "deviant" as acceptable patterns and ways of doing and thinking.
Indeed, the combination of the key conceptions above is crucial in highlighting how legitimizations, pre-legitimizations and imaginaries all figure both in the construction of "moral panics" and in their subsequent normalisation. The former and the latter obviously build on distinctions and on visions of the post-normalisation order (Foucault 2007) and are rooted in duality of the normalisation language (Vaughan 1996). They, similarly, allow legitimising views and opinions as well as normalising radical or deviant positions along traditional discursive patterns of legitimization by, for example, making references toauthority, path dependency, moral values etc. (van Leeuwen 2007;van Leeuwen and Wodak 1999). But they also, subsequently, pre-legitimise (Krzyżanowski 2014) normalisation by creating imagined (and often untrue) future-related visions of the state of affairs whilst drawing on allegedly present-oriented and fact-based claims. All of these, importantly, become core elements of the "borderline discourse" (Krzyżanowski and Ledin 2017) of normalisation, wherein imaginaries and pre-legitimationstied to moral panics, visions of "crisis", etc.create ambiguity as to the facts. By the same token, wider discourses of historical, religious or other values (see below) are also deployed here to serve as pre-legitimators of the ensuing normalisationand hence effectively also radicalisationof views.

Discursive shifts and normalisation of racism in Poland: analysis
The analysis below aims to illustrate how the key stages of the discursive shifts elaborated above have been prominent in discourses of the Polish post-2015 public sphere. The main idea here is to depict how in the course of 3-4 years starting with the period of the Polish pre-election campaign of 2015, one has witnessed not only the enactment of wider immigration-related discourses in Poland -where these have been very significantly missing (Krzyżanowski 2014)but how since their strategic inception and through subsequent perpetuation/gradation, the discourse on immigration in Polish politics and media acquired a largely negative, anti-immigration meaning which was progressively normalised, often in radical ways.
The focus of the analysis is, specifically, on a number of discursive strategies and arguments (see below for details) which have, in an obvious way, distorted the facts and presented a hugely imaginary and untrue, negative image of immigration, embodied in most cases by unspecified "refugees". The analysis aims to show how these were deployed as part of a wider strategy to create false imaginaries and "moral panic" around immigration, and how this has been achieved in a context where, on the one hand, no specific ideas about migrants existed before and, on the other, there were, and still are, very limited numbers of immigrants.
Analytically, the focus here is not on radical sites of production and reception of public discourse but on the wider Polish public sphere and especially its right-wing or conservative areas, in most cases those at least officially deemed right of centre or mainstream. As such, the analysis departs from political discoursespecifically of the Polish right-wing populist government party, PiSand its various formats expressed and recontextualised in late 2015 in such spaces and genres as: parliamentary speeches and interventions, selfmediated online discourse of, for example, YouTube videos and posts, or speeches and addresses at pre-election rallies. Whilst the political discourse dominates the first two phases of enactment and gradation/ perpetuationconsidered jointly belowit is also in that phase that a sample of parallel/ subsequent media (daily tabloid and conservative weekly) discourse is analysed in order to show the recontextualisation and diffusion of key themes and arguments from political discourse. The focus on media discourse and genres also continues in the following section devoted to the third, i.e. normalisation, phase. The aim is to observe through which key arenas the discourse originally introduced and disseminated in the enactment and gradation/perpetuation phases was effectively normalised. However, the second phase also includes other spaces/genres as sites of normalisation of discriminatory and racist discourse, such as the online political campaign discourse (specifically, video spots posted by PiS on Twitter), but also quasi-intellectual discourse which proved immensely efficient in causing radical anti-immigration views to proliferate.
The analysis below is conducted in line with the analytical categories of critical discourse studies of right-wing populism and anti-immigration rhetoric (Krzyżanowski 2012(Krzyżanowski , 2013bWodak 2015). It follows a multilevel pattern (Krzyżanowski 2010) that distinguishes between entry-level (thematic, content-oriented) and in-depth (strategyoriented, especially argumentative) analysis. Within the latter, in-depth level, the focus is in particular on key patterns of argumentation and discursive strategies. The key category used here is, hence, that of topoi (or argumentation schemes/headings; Krzyżanowski 2010), especially forms known from public and political anti-immigration discourse (in particular topoi of danger/threat, burden/costs, cultural/religious differences etc.; see Reisigl and Wodak 2001), whilst also looking at such discursive strategies as those related to legitimation, including via authority, morality, path-dependency or alike (see van Leeuwen 2007), or various patterns of representingtextually and visuallysocial actors by means of, for example, their collectivisation or aggregation, their foregrounding or backgrounding etc. (van Leeuwen 2008). For the reasons indicated above, recontextualisation (Bernstein 1990;Krzyżanowski 2016) also appears as the key overarching category of analysis and as a key tool for understanding the connection established in the course of the analysed discursive shifts between topics, topoi/arguments and discursive strategies deployed in various spaces and genres of the analysed public discourse.

Phase 1: Enactment
The analysis departs from the focus on the discourse of PiS, i.e. Poland's current government right-wing populist party which in both the 2015 and 2019 national parliamentary elections secured a majority of seats in the Polish parliamentwith 37.6 and 43.6 per cent of national votes, respectively, and in each case formed a majority government. As a party that has very frequently changed its political ideology which shifted immensely over almost two decades of the existence of PiS and its antecedents (Gwiazda 2008;Krzyżanowski 2011), since 2015 PiS has clearly embarked on a project of re-inventing itself as an ethno-nationalist party (Krzyżanowski 2018a;Markowski 2016Markowski , 2019. As a consequence of that, PiS, in a rather evidently opportunistic way, also chose to invent itself as a party that is not only "sceptical" about immigrationabout which the general public's knowledge was indeed limitedbut also, in fact, holding outright anti-immigration views (for more details of the political and policy context, see Krzyżanowski 2018a).
However, while PiS visibly aimed to not only introduce but also champion an anti-immigration agenda in Poland, the party still needed to face the fact that, as such, the discourse about immigration was until 2015 relatively marginal within the wider Polish public sphere. Despite having some extensive experience as a country of emigration as well as a long-standing and clearly troubled history of approaching multiculturalism and ethnic/ national/ religious minoritiesviz. Polish still rampant and widespread anti-Semitism often extended to other minorities (Michlic 2007;Ostolski 2007; Jasińska-Kania and Łodziński 2009) -Poland and specifically the Polish public sphere did not really "possess" many viable imaginaries of immigration. These, therefore, first had to be createdor enactedbefore strategically changing the discourse to have a clearly exclusionary direction/tone, as was eventually achieved by PiS.
Such enactment occurred most visibly in the week commencing 14 September 2015 via, in particular, PiS' own political discourse, as well as being augmented by PiSsympathetic conservative media. The party's own discourse in this phasepreviously analysed at length in Krzyżanowski (2018a)was characterised by, as usual for the party, the omnipresence of the party chairman, J. Kaczyński, who, skilfully using his "traditional" rhetorical strategies (esp. rhetorical questions), largely enacted not only the party's discourse on immigration as such but also showed that the bulk of it was characterised by an antiimmigration topos of danger/threat. Thus Kaczynski spoke extensively of cultural, religious and other dangers that allegedly stemmed from the increase in immigration which, although often not referred to or specified (viz. a strategy of "indetermination"), by logic of implicatures mainly consisted of imagined "refugees" supposedly arriving in the course of the then "Refugee Crisis". However, using his classic format of topos of example, Kaczyński used manyde facto untrueexamples of the plight of several (Western) European countries that allegedly "suffered" from immigration, mainly associated with Islam/Muslims. To support his views, in his parliamentary intervention of 16 September 2015, 2 Kaczyński mentioned the following "examples" (see Example 1): Example 1: If somebody says all of this is not true then have a look around Europe, let's take Sweden. There are 45 zones there governed by Sharia law, there is no control by the state (…) There are fears over hanging the Swedish flag in schools, as is the custom there, because there is a cross on the flag (…) Or what is going on in Italy? Churches have been taken over and are often treated as toilets. What is going on in France? Non-stop arguments, Sharia introduced, even patrols which check if Sharia is being observed.
While Kaczyński's and PiS' discourse textually introduced a quasi-logical and quasi "example/fact-based" anti-immigration logic and specifically Islamophobic discourse, some Polish conservative newspapers and magazines followed suit in exactly the same period. They eagerly added various arguments, as well as significant visual representations, to the danger/threat argumentation legitimised by PiS' broader right-wing populist politics.
For example, at the end of the week preceding the one in which Kaczyński delivered his parliamentary address above, Polish tabloid Super Express devoted a significant part of its front pages to the "Refugee Crisis" by, on the one hand, claiming "Alarm: 90 thousand Arabs will flood Poland" ("Alarm: 90 tysięcy Arabów zaleje Polskę", 10 September 2015, see Figure 2) or that "Terrorists might be among the refugees" ("Terroryści mogą być wśród uchodźców", 11 September 2015, see Figure 3). Here, both collectivisation and aggregation were deployed as strategies supporting the above argumentation, with visual representations that supported the above headlines representing large numbers and queues of individualsallegedly refugeesor juxtaposing such "refugees" with "terrorists" in order to support danger-related claims as well as draw on the closely related topos of criminality.
Interestingly, the topos of numbers was also very prominent as the source of constructed arguments with the hugely exaggerated total number of "90,000" claimed in one headline and supported by further number-related statements claiming that: "The European Union is forcing us to accept 13,000 refugeesand each of them will have the right for their familyon average 7 further peopleto join them" ("Unia Europejska każe nam przyjąć 13 tys. uchodźcówkażdy będzie miał prawo sprowadzić rodzinęśrednio 7 osób", 10 September 2015).
While the tabloid dailies initiated anti-immigration arguments in the media domain, these claims were certainly reinforced within just a few days by the conservative weekliessuch as, very notably, W Sieci and Do Rzeczy (see Figures 4 and 5) which not only, yet again, alludednow visuallyto the large number of supposedly "incoming" refugees, but also clearly spoke about the fact that the refugees were part of an alleged "invasion", as in the frontpage headline: "They are invaders, not refugees" ("To najeźdźcy, nie uchodźcy", Do Rzeczy 14-20 September 2015, Figure 4). Within the visual representations, image collages were also used, as on the front page of W Sieci (Figure 4), with a compilation of stereotypically (judging mainly by the attire etc.) Middle Eastern looking men, including some holding guns, who were allegedly breaching Polish border barriers and disrespecting the Polish national emblem. Interestingly, the latter, collated image wasin its figurative formalso a recontextualisation of a widely known image of the Nazi-German invasion of Poland in 1939, 3 which has proliferated in Polish media and history books, and hence is very suggestive by way of its strategically implied parallels to the current "refugee invasion".

Phase 2: Gradation/Perpetuation
The second phase of the development of Polish anti-immigration discourse after 2015 consists, generally speaking, of a widespread recontextualisation of the aforementioned textual and visual arguments, strategies and representations. This recontextualisation has, however, multiple meanings. It, on the one hand, servedin the classic understanding of recontextualisation (Bernstein 1990;Krzyżanowski 2016)the diffusion of discursive elements or their hierarchical subordinationby way of following patterns introduced in the enactment phase above. But, on the other hand, the analysed discourse further augmented and fortifiedand effectively discursively "intensified" (Reisigl and Wodak 2001)many of the previously introduced arguments and claims, including making them ever-more radical and ever-less mitigated. It also, in addition to the above, very visibly distorted the facts as well as in classic post-truth mannerused completely untrue examples, information and "fake news" as elements of description and interpretation of "real" processes (specifically, the "Refugee Crisis").
One key example from this phase comes from PiS' own political discourse that was selfmediated online by the party via an online video post from Antoni Macierewicz, a deputy chairman of PiS and formerly Minster of Defence in a PiS government. His video was posted on 24 September 2015 in the series of Macierewicz's regular videos "The Voice from Poland" ("Głos z Polski"). The latter were at that time regularly disseminated via the Youtube channel of a Catholic-conservative media conglomerate and long-term supporter of PiS, i.e. Radio Maryja / TV Trwam (for an extensive thematic and in-depth analysis of the post, see Krzyżanowski 2018a). In the post, Macierewicz alludes, just like Kaczyński before (see above), to a number of dangers posed by refugees who will allegedly soon be arriving in Poland in great numbers, whilst he also attempts to build a paradoxical argument wherein a claim is made that refugees should be helped "back home" rather than in the European countries to which they flee.
Most importantly, however, Macierewicz sustains the aforementioned topos of danger/ threat by continuingand indeed particularisingthe idea that refugees or migrants in general were coming to Europe as part of a skilfully-planned invasion. This argument was even visually "supported" in his post by presenting an essentially fake military-looking invasion map (with no source given; see Figure 6), which was allegedly distributed in the refugee camps. The map clearly shows that Poland was the main target of a supposed "refugee invasion", as well as serving as the key transition country in a wider alleged European "attack" by refugees and migrants. The salience of this example is such that, while the discourse not only retains its abstract characterwith the alleged invasion being directed against European culture, values or Christianityit is also additionally endowed with a high degree of materiality. The refugee "invasion" is, then, seen here as something additional happening in the physical spaceof countries, borders, territoriesand hence as a physical as well as an abstract or symbolic "attack".
The use of "fake news" as aboveand of not only distorted but also outright absurd, quasi-factswas also present in the discourse of PiS which carried the above ideas and was expressed and recontextualised in further spaces and genres. Most notably, it was present in a very notorious election rally speech by Kaczyński delivered in October 2015, i.e. towards the end of the then Polish national-parliamentary election campaign (and further recontextualised via, for example, the party's official Youtube channel). During his speech, Kaczyński, as always, asked a number of rhetorical questionsincluding evident exaggerations, e.g. whether "the news about getting 100,000 Muslims into Poland was true" (NB: the Polish refugee quota was set to a little over 7,000; see Krzyżanowski 2018a). But he insinuated that in the course of the refugee crisis there were outbreaks of many diseaseshence implying those diseases were brought to Europe by migrants/refugees (see Example 2).
Two very vital issues transpire from the above example and its use of discursive strategies. On the one hand, it shows that Kaczyński deploys a very peculiar set of elements of well-known Polish historical discriminatory discourse by mentioning specific nominal phrases, e.g. "parasites" or "protozoans", characteristic of easily recognisable anti-Semitic discourses. However, Kaczyński recontextualises these into the current and essentially Islamophobic statement/argument. He thus creates a peculiar conflation of racism, wherein elements of an "old" and recognisable language of discriminationalso referred to as biologizing metaphors (Reisigl and Wodak 2001) is now recontextualised into a new discourse directed against new, stigmatised social actors (current refugees, and in particular Muslims). 4 By the same token, he endows the current discourse with a very high degree of materiality. This not only relates to specific places ("Vienna", "Greek islands") that remain passivizedand shown as if under attackbut also to the actual bodies of refugees, which are shown as material yet active "carriers" of diseases and biological dangers.

Example 2:
There are already symptoms of very dangerous diseases not seen in Europe for a long time: cholera on the Greek islands, dysentery in Vienna, some are saying there are also other, even more serious, diseases. And there are also geographically-based differencesthere are various types of parasites, and protozoans, which are not dangerous to those people's organisms, but can be dangerous here. It is not about discriminating against anybody, but all this needs to be checked. It needs to be checked and hence the Minister of Health should clarify how this looks at the moment, because from what we know, what our activists know (…) it all looks very bad. 5

Phase 3: Normalisation
Finally, the third normalisation phase of the highlighted discursive shift is characterised by not only the continuation of the discursive processes enacted and subsequently perpetuated abovethus proving the ongoing, normalising viability of strategic recontextualisation processesbut also by the presence of a peculiar "borderline discourse" which serves as evidence of a modified value system. Such discourse, as shown below, points to the change in attitudes around once deviant positionsin our case very obvious racismwhich are now being normalised by effectively being "clad" in acceptable, civil discourses helping to, inter alia, rationalise racism or even create arguments about its apparent moral virtues.
The first continuity aspect is certainly visible in the fact that strategies such as. the aforementioned conflation of racism (see Example 2) are now also recontextualised into further spaces and genres. Figure 7 shows, for example, how such a discursive strategy is used in media discourse, and specifically on the notorious cover of 'W Sieci' conservative weekly, from early 2016, allegedly representing an "Islamic rape of Europe" ("Islamski gwałt na Europie"). The said cover merges as well as conflates ethnically-based racist discourse (note the varied, non-Caucasian ethnicity of the alleged rapists' groping hands) with a sexist argument (the wider allegory of rape and the foregrounded, actual violent physical and sexual attack on the depicted female). It thus creates implicature wherein ethnic diversity equals violence, including of a sexual nature. But this argument is then also linked with a historical 'myth' of the 'Rape of Europa' (note the implicature in the headline and the EU flag as the only attire of the attacked female). Hence, through all of the above, a multidimensional pattern of recontextualisation is put in place. It not only continues but, yet again, further fortifies and augments (gradates) the earlier arguments by adding to these even more new aspects of the danger and threat supposedly posed by migrants and their current, alleged invasion.
But while the constant striving for ever-more complex discriminatory arguments augmented by ever-less mitigated statements is vital, probably the central strategy that becomes symptomatic for the normalisation phase is the existence of the said "borderline discourse" that links uncivil ideologies such as racism with quasi-civil language. The standard element of that discourse is a strategy of rationalization which, rather than calling for discriminatory or racist acts, presents and pre-legitimises quasi-rational arguments in favour of discriminatory ideas, as well as showing how and why these should eventually translate into current and future discriminatory statements and acts.
Such a rationalisation strategy was certainly vital in, for example, some instances of the recent, Polish quasi-intellectual Catholic discourse associated mainly with (strongly) nationalist-conservative viewsand embodied by a number of publications from conservative publishing houses or media including those already mentioned in the analysis above (see e.g. Figure 6). A very typical case among such publications is, inter alia, a 2017 book "Immigrants at the Gates" ("Imigranci u bram") which presents Islamallegedly now carried into Europe by migrants/migration and especially recent refugeesas a direct threat and inherent opposition to Christianity. The book links such unrelated phenomena as attacks on Christians in various parts of the world to alleged Muslim violence (an indeed "classic" element of Islamophobic RWP discourse; see Krzyżanowski 2013b), as well as rationalising the obvious anti-immigration/Islamophobic stance. It constructs a wider, complex argument claiming that Christians should certainly not be supporting Muslims and should, if possible, engage in active acts of "defending against" them.
The caption of the said book (see Example 3) first furthers, on its back-cover, the above argument and effectively rationalises non-support for refugeesor even their discriminationas an essentially Christian act and one in defence of Christian values. Example 3: Every three minutes a Christian dies somewhere in the world (…) The death of a Christianalmost always at the hands of a Muslimdoes not outrage the mass media. It is different with all the misfortunes of believers in Islamthat we should all be outraged by. 6 But while it is initially mitigated, the argument eventually "switches" and becomes outright discriminatory (see Example 4), by claiming that Muslimsnow strategically activated -"destroy churches and households of Christians" as well as "murder, rape, burn, injure, cut heads and lay bombs". The use of sarcasme.g. naming Muslims as "victims"is also strategic here as their victimhood is thus questioned, with Muslimspresented as violent aggressors Example 4: Nobody else but the brothers of Islamic victims still destroy temples and households of Christians, they murder, rape, burn, injure, cut heads, lay bombs. Some of them then go, or more specifically are exported, to Europe. There is no way of knowing who is a refugee, who is a migrant and who is a terrorist.
But such a paradoxical linkagebetween a rationalisation strategy, (purposefully misinterpreted) Christian theology and discriminatory, Islamophobic argumentscould also well be seen in Polish radical-Catholic conservative media discourse. For instance, the daily Nasz Dziennik from mid-2017 shows how an argument about "Hostile Immigration" is built on misinterpreted and distorted quotes from key Christian theologians, e.g. St. Thomas Aquinas (see Example 5) who is argued to have allegedly spoken about "immigration which is harmful to the common good". But at the same time, secular and civic rationalisation is also addedvia references to "the state", the "common good" and "citizens"thus creating the impression that not only civil but also Christian values effectively allow discrimination against migrants and, in particular, Muslims. This argument also, interestingly, further enables the usual "invasion" argumentation (topos of danger and threat, see above), along with implicatures whereby the level of current migration clearly suggests that it is organised (though, as usual, rhetorical questions are used to mitigate the directness of such a statement/implicature).
Example 5: The aim of the state is the defence of the common good of its citizens. There are specific situations in which immigration can be harmful to the common good. Saint Thomas Aquinas called this "hostile immigration". (…). Namely, in a situation where Europe has lost its foundations, by renouncing its roots, there are people at its door who have strong will and strong religious motivation. The question about the level of this immigration alone raises a question: is it an organised phenomenon or not? 7 However, while the strongly conservative-Catholic media drew extensively on the above combination of different rationalizations, instances of PiS' own rhetoric show that using  disinformation in a discourse inciting anti-immigration moral panic was equally efficient and telling for the normalisation process. A vital example was, inter alia, the notorious pre-election video spot "A safe, regional government" ("Bezpieczny samorząd") published by PiS on its official Twitter account in the course of the 2018 regional election campaign in Poland (see Figure 8 and 9). In the spotopenly directed against pro-immigration mayors of major Polish cities such as Warsaw, Poznań, Gdańsk, Kraków or Lublin (i.e. all the major Polish cities notoriously uncharted by PiS and traditionally governed by its liberal opponents)a quasi-news report was shown along with the argument that Polish big-city mayors do not care for public safety as they are protective of migrants/ refugees who are otherwise known to be violent, or even active as terrorists Figure 9.
The "news report", which clearly implied factual information and thus also allowed for a peculiar generic normalisation (via the use of acceptable and recognisable information genres), was essentially a collage ofat firsta negative image of faces of liberal (non-PiS) politicians supporting immigration (Figure 8), followed by a number of short spots (Figure 9) showing alleged migrant/refugee/Muslim violence, including images of physical violence, explosions or police-criminal street fights and shootings. Those imageswhich were crucial elements of a discursive normalisation tied in with the aforementioned, generic one, were then connected to others showing "refugees"visibly only menapparently waiting in large numbers at what was presented as one of Europe's border crossings. The textual captions (news strips) added a necessary meta-narrative: they provided "locations" of the alleged newsthe name of an allegedly peaceful, small west-Polish town ("Nowe Miasteczko") or a "border crossing" ("przejście graniczne")as well as the implied, wider context of a "crisis with refugees" ("kryzys z uchodźcami"). But the Figure 9. Selected shots from @pisorgpl election spot "A safe, regional government" ("Bezpieczny samorząd") from 17 October 2018.
captions also, importantly, recontextualised the usual format of rhetorical questionsso well used in PiS immigration-related discourse otherwise (see above)with an overarching one: "Is this how Poland will look in 2020?" ("Czy tak będzie wyglądała Polska w roku 2020?").

Conclusions
The above critical discourse analysis has highlighted howby introducing moral panic about immigrationthe contemporary public discourse has undergone significant changes and shifts in the context of evolving right-wing populist politics and ideologies. Whilst of huge importance to the wider change in discourse and attitudes across the public sphereespecially with regard to issues such as immigrationthe aforementioned analysis of discursive shifts points in particular to the fact that, as such, those shifts are not arbitrary or accidental in nature. On the contrary, major changes in the public dynamics of discourse of late are, as has been shown, elements of conscious, well-designed, opportunistic political communication strategies which allow for the spread of ideas and ideologiesin our case often outright racist and discriminatory anti-immigration viewsacross the wider public domain.
Looking specifically at the case of Poland, the analysis has shown a very significant, if not utterly radical, discursive shift in public discourse and how it has been reinforced and sustained over time. This has in particular been achieved by the use of very skilful strategies of recontextualisation (Bernstein 1990;Krzyżanowski 2016) which create a very peculiar and indeed path-dependent, new order of discourse. Indeed, having looked at different stagesof how the discourse of anti-immigration has not only been enacted and subsequently spread or perpetuated, but also eventually normalised in its often obviously racist and discriminatory formthe analysis has shown that ongoing processes of recontextualisation are central in creating as well as sustaining some very negative stance on immigration and in "settling" and perpetuating strong, anti-immigration views, in our case in Poland. This, in turn, shows how public discourse can be effectively "engineered" (Fairclough 1992) and its trajectory effectively controlled and changed in line with political motives and ideological aims which, however, eventually also bring a deeper change to intra-societal, intergroup relations and attitudes.
The peculiarity of the highlighted Polish case resides in the fact that, as such, and unlike many other countries in Europe and elsewhere, Poland did not have any coherent or comprehensive immigration-related discourse until ca. 2015 (Krzyżanowski 2014(Krzyżanowski , 2018a. It also, it should be reiterated, developed the recent immigration-related discourse largely on the basis of imaginaries and irrespective of facts, e.g. of very minimal numbers of migrants and/or refugees arriving in Poland. However, looking at Poland has certainly made it possible to showin an in-depth and systematic mannerthe entire multistep process of discursive shifts: starting from enactment, going though gradation/ perpetuation and on to the normalisation of various discursive elements, and wider ideological positions, around immigration. Yet, in these stages of discursive shifts, particularly the last one, i.e. normalisation, remains crucial. It is, perhaps, the one which, as has been shown, not only explicitly points to the actual dynamics or change in language and discourse or in patterns of "talking about" social actors, situations or objects of knowledge (especially encapsulated by the first two phases of the proposed model). It also, or perhaps particularly, points to the fact of how, reading the dynamics of discourse and its evolution, one may point in the direction of much deeper changes to norms of not only public expression but also the perception and "classification" of social actors and groups that has taken place in the course of discursive shift/s. As such, normalisation also shows how, if enacted and perpetuated, the longevity of moral panic does indeed change "the way society conceives itself" (Cohen 2011, 1) and its members, some of whomin our case migrants/ refugees are often used as objects of ideological and politically-opportunistic strategies.
However, what was shown extensively above is also the fact that the said, sustainableand eventually (skilfully) normalisedmoral panic does not rest on any ad hoc ideas or accidental strategies related to immigration. On the contrary, the said immigrationrelated moral panic is effectively made continuous and eventually normalised via a peculiar borderline discourse (Krzyżanowski and Ledin 2017) which verges on civility and uncivility by using quasi-civil language, while, however, seeking to effectively project uncivil or outright radical positions and views (such as, in our case, racism). Characterised by strategies of, in particular, recontextualisation of arguments (danger/threat topoi), discursive elements (rhetorical questions, frame shifts), genres or salient discursive strategies (rationalisation but also mitigation, pre-legitimation), as well as many untrue imaginaries of immigration and refugees, this borderline discourse allows "dressing" up radical ideologies and social views in quasi-rational, modern discourse. It also allows creating untrue arguments which, however, escape being perceived as uncivil, manipulative "fake news" and disinformation due to their civil appearance which effectively normalises them in political and media discourse and in both traditional and online public spheres.