The Populist Radical Right as Memory Entrepreneur? The Prominence, Sentiment, and Interpretations of History in the German Parliament

Abstract

Populist radical right (PRR) parties' attacks against prevailing historical interpretations have received much public attention because they question the foundations of countries' political orders. Yet, how prominent are such attacks and what characterizes their sentiment and content? This article proposes an integrated mixed-methods approach to investigate the prominence, sentiment, and interpretations of history in PRR politicians' parliamentary speeches. Studying the case of Germany, we conducted a quantitative analysis of national parliamentary speeches (2017–2021), combined with a qualitative analysis of all speeches made by Alternative for Germany (AfD) in 2017–2018. The AfD does not use historical markers more prominently but is distinctly less negative when speaking about history compared to its general political language. The collocation and qualitative analyses reveal the nuanced ways in which the AfD affirms and disavows various mnemonic traditions, underlining the PRR's complex engagement with established norms.

Introduction

The rising populist radical right (PRR) has challenged countries' interpretations of history around the world. Whether it is the Alternative for Germany's (Alternative für Deutschland, AfD) 2021 campaign to restore a ‘normal’ Germany (Deutsche Welle 2021), Donald Trump's ‘Make America Great Again’, or Jair Bolsonaro's promise to ‘return Brazil to [the] past’ (Elliot and Phillips 2018), evoking a sense of nostalgia by referring to history has been a key part of the PRR's mobilization of voters (Elçi 2022, 710; Gest, Reny, and Mayer 2018, 1,695; Norris and Inglehart 2019, 16; Steenvoorden and Harteveld 2018, 29).

Historical narratives matter because they impact the acceptance of democratic institutions and the functioning of party systems and civil society (Boyd 2008; Kubik and Bernhard 2014, 29–30; Müller 2002). To evoke the longing for an often abstract ‘paradise lost’, the PRR regularly refers to events that form a country's collective memory. Some of these statements have caused public outcry, like former AfD leader Alexander Gauland's claim that the twelve years of Nazi rule were ‘mere bird shit in over 1,000 years of successful German history’ (Deutsche Welle 2018). In more subtle ways, in 2019, the AfD demanded a ‘Wende 2.0’, the term for East Germany's peaceful revolution in 1989, implying that Germany has not sufficiently broken with its Communist past (Kallgren and Huet 2019). While interpreting key historical events is usually an ongoing and collective effort, the PRR claims that its view on history is the only accurate one (Müller 2017, 27–35; Rosenfeld 2023, 823; Taggart 2004, 274, 279).

This article examines the PRR's use of history compared to that of other parties in Germany. We use an integrated mixed-methods approach to examine how prominent references to the past are and what sentiments and historical interpretations characterize that language. Focusing on Germany, where the rise of a far-right challenger occurred relatively late compared to other European countries, we conducted a quantitative analysis of all national parliamentary debates between 2017 and 2021. We complement this analysis with an in-depth qualitative content analysis of PRR parliamentary speeches in 2017 and 2018 when the latter's discursive practice took shape in parliament. We find, against common expectations, that the PRR does not use historical language more prominently than other parties, but only engages more frequently with the memory of the Second World War. The sentiments that characterize its discussion of history markedly differ from its overall language. While its overall language is more negative compared to other parties, echoing previous findings (Valentim and Widmann 2023), its language is significantly less negative when talking about history and neutral around terms associated with the memory of the Second World War compared to other parties. A collocation analysis across parties and a qualitative analysis of PRR speeches show that the PRR's interpretation, in terms of accepting, minimizing, or rejecting mainstream historical discourse, varies within and across historical topics. Overall, the PRR employs a multifaceted narrative of historical continuity and distance that challenges common interpretations of history in a complex manner and at the same time locates it within Germany's historical imagination beyond the memory of the Second World War.

These findings have important substantive and normative implications. Recent work has highlighted that the use of language considered a social taboo has pushed normative boundaries and facilitated the spread of nativist and populist language (Habersack and Werner 2023; Valentim 2021). We contribute to this scholarship by specifying the extent to which and how the PRR challenges normative boundaries by politicizing history. Substantively, our analyses illustrate that the PRR, by drawing on various historical themes, rejects only some of the pillars of collective memory while accepting or minimizing others. This indicates a complex and often subtle strategy to redefine discursive rules. Methodologically, we propose closely integrating quantitative and qualitative text analysis. Quantitative text analysis allows revealing patterns of thematic and affective salience across a theoretically guided set of search terms, which are then contextualized through material from a collocation and qualitative content analysis to deepen our understanding of the PRR's argumentative structure.

The next section develops our theoretical expectations. We then explain the rationale for focusing on Germany and outline our methodological approach, followed by a discussion of our findings. The last section concludes and elaborates on the findings' comparative implications.

The Populist Radical Right's Language About History

Historical narratives are the result of negotiations over the present-day meaning attributed to the past, often called collective memory (for example Berger 2002; Feindt et al. 2014; Wüstenberg 2017). This inspires ‘cultural and institutional practices that are designed to publicly commemorate and/or remember a single event, a relatively clearly delineated and interrelated set of events, or a distinguishable past process’, which Kubik and Bernhard (2014, 15–6) have coined ‘memory regime’. The ways in which political actors, such as the representatives of political parties, speak about the past shape memory regimes because such statements travel into public debates (Art 2006) and public policy (Boyd 2008; Müller 2002). The rise and fall of political actors are thus often important crossroads for the development of a memory regime (Kubik and Bernhard 2014, 16).

The rise of the populist radical right constitutes one of the most important changes in contemporary democracies. These parties have manipulated the issue agenda by emphasizing, ignoring, or combining particular positions (de Lange and Mügge 2015; De Vries and Hobolt 2020; Habersack and Werner 2023). We consider populism a key ideological feature of these parties (Manucci 2020; Mudde 2007) because populism, among other things, turns these parties into inherent mnemonic actors. While sometimes associated with a utopian vision of democracy (Tarragoni 2024), populists' emphasis on virtuous people who have been betrayed or robbed by a corrupt elite implies the existence of a better time in the past (Albertazzi and McDonnell 2008, 3; Mudde 2007, 23; Tarragoni 2024, 48). Paul Taggart (2004, 274) has captured this in the notion of the ‘heartland’ – an ideal world that, ‘unlike utopian conceptions’, ‘is constructed from the past’. It emphasizes a conflict-free and culturally homogeneous society preceding the rise of post-materialism (see also Norris and Inglehart 2019). This world contrasts with the present, invoking a sense of nostalgia for what has been lost (Betz and Johnson 2004, 311; Taggart 2004, 274). Populist parties present themselves as the only bearers of the historical truth, in contrast to other parties' false versions of history (Taggart 2004, 274, 279; see also Müller 2017, 27–35). It is by making others accept their vision of the past that populist parties seek to establish cultural control and the hegemony of their interpretation of events (Rosenfeld 2023, 821). Communicating what has been ‘lost’ requires the PRR to speak more prominently about history than other parties.

H1: We expect populist radical right parties to refer more prominently to history than other parties.

The emphasis on a ‘heartland’ or ‘paradise lost’ has implications for the sentiment populist radical right parties convey when speaking about history. Negative emotional appeals generally characterize their language (Salmela and von Scheve 2018; Widmann 2021), in line with the PRR's proclaimed opposition to the political establishment, accusations of the establishment's political failure, and demands for far-reaching reforms (Akkerman, Mudde, and Zaslove 2014). Consequently, PRR parties use more negative language overall than other parties (Valentim and Widmann 2023; Widmann 2021). However, when comparing the language that the PRR uses when talking about the past with that used when talking about the present, we would expect a more nuanced picture. The notion of the ‘heartland’ or ‘paradise lost’ contrasts a negative present with an idealized past (Elçi 2022, 700). We would consequently expect more positive sentiments when the PRR speaks about the past than when speaking about the present.

Sentiments invoked in political language may also inspire feelings of association or dissociation (Widmann 2021, 164–5), and the memory of twentieth-century fascism has been particularly relevant for populist radical right parties (for example, Art 2006; Caramani and Manucci 2019; Griffini 2023).¹ Although populist radical right and fascist parties are not the same, the former's populist element presents certain continuities to fascism, just as populism includes many of the illiberal ‘core elements of fascism at odds with a liberal and constitutional interpretation of democracy’ (Manucci 2020, 47; see also Müller 2017, 27–8). Ignazi (1992), who famously argued that the ‘new’ radical right differed from the ‘old’ extreme right by avoiding an open nostalgia for twentieth-century fascism, did not discuss these parties' populist features, while these features make us expect PRR parties to express more positive sentiments when talking about the memory of fascism compared to other parties.

H2: We expect the populist radical right's language to be a) less negative when speaking about history compared to its overall language and to be b) more positive than other parties when discussing the history of fascism.

Finally, memory politics requires actors to decide on the interpretation of historical experiences. The strategic selection, or ‘selective memory’ (Griffini 2023, 6), of which past episodes to discuss helps build a narrative that embeds the populist radical right within a country's historical imagination. To group these experiences, Couperus and Tortola (2019) distinguish between a country's ‘dark past’, involving events and characters remembered as negative or shameful, and a ‘noble past’, typically including national victories, liberation, or unification. For example, although countries differ in the extent to which they have internalized the responsibility for fascism (Caramani and Manucci 2019; Manucci 2020, 51–2), the time of the fascist rule has generally been part of Europe's ‘dark past’ (Couperus and Tortola 2019, 107–8).

Accepting a country's interpretation of its ‘dark past’ can help the PRR push its positions and language into the mainstream and appeal to more moderate voters (Couperus and Tortola 2019, 107). For instance, PRR parties have upheld the remembrance of the Holocaust, which can also serve to advance their own agenda (Rosenfeld 2023, 825). The latter is often part of a minimizing strategy, whereby mainstream interpretations of history are downplayed in their significance. For instance, embedding the Holocaust within a wider set of atrocities serves to undermine its uniqueness (Rosenfeld 2023, 825). Similarly, pointing toward the victims of Communist crimes as deserving as much public attention as those of Nazism helps portray the latter as simply another form of dictatorship (Wüstenberg 2019). Condemning other countries' ‘dark past’ or equating events under fascist rule with current events in democratic settings may be alternative strategies for PRR parties to minimize fascism's seriousness, without openly defying the anti-fascist consensus (Griffini 2022, 10–11). In turn, rejecting a country's interpretation of its ‘dark past’, as AfD politician Björn Höcke did when demanding a ‘U-turn in remembrance politics’ (Kröning 2017), risks entailing social sanctions (Valentim 2021, 2,476). We might therefore see an ‘attack and retreat’ strategy. PRR parties reject a country's memory regime, for instance by denying guilt, dismissing or demonizing recognized victims, or embracing notorious perpetrators (Rosenfeld 2023, 824–26), before distancing themselves from such statements (for example Couperus and Tortola 2019, 111; Wodak 2015, 11–18).

These three strategies of accepting, minimizing, and rejecting can also be applied to the PRR's interpretation of episodes of countries' ‘noble past.’ Accepting stories and characters from a country's ‘noble past’ is particularly conducive to PRR parties' narrative of the heartland as the ‘good life before the corruption and distortions of the present’ (Taggart 2004, 274). The heartland is often deliberately ‘shrouded in imprecision’ to not threaten their claim to stand for a homogeneous people (Taggart 2004, 274). Yet, PRR parties' emphasis on ethnic homogeneity and order make events and actors of national struggle, resilience, and perceived grandeur before the arrival of mass immigration a likely historical background to underpin the sentiment that things were better in the past (Elçi 2022, 697–8). Positive references to other European countries' national past and Europe as a historical space of sovereign nations with a common Judeo-Christian heritage can further contribute to this sense of lost national identity (Cerrone 2023; De Cesari, Kaya, and Piacentini 2020, 29). By contrast, minimizing ‘noble past’ episodes like post-war European integration by, for instance, emphasizing problems associated with specific EU institutions or rejecting their interpretation as ‘noble past’ episodes altogether can help PRR parties define their notion of the heartland (Cerrone 2023, 954). It is therefore important to analyze how the PRR interprets events across historical topics.

H3: We expect the populist radical right to make ambiguous statements about the Second World War, which include accepting, minimizing, and rejecting fascism's crimes, and employ references to a de-contextualized ‘dark’ and ‘noble’ past that incorporate diverse themes.

Why Germany

Germany is a particularly pertinent case to test our hypotheses. On the one hand, it can be seen as a hard case for PPR memory entrepreneurship. Despite a mixed record of ‘de-Nazification’ (see, for example Capoccia and Pop-Eleches 2020, 108–13), a narrative of ‘culpabilization’ emerged in the 1950s and subsequently solidified, emphasizing the collective responsibility for Nazi atrocities (Manucci 2020, 51, 123). This narrative has imposed discursive barriers, which, combined with the legal restrictions imposed by Germany's ‘militant democracy’ (Capoccia 2013, 211), have made Germany unique in terms of the political and legal costs for those who challenge the collective memory of the Second World War (Art 2006, 49; Manucci 2020, 123). Germany consequently provides the PRR with a difficult mnemonic environment. We might, therefore, expect the PRR to discuss history less or at least not more frequently than other parties and especially refrain from discussing the history of the Second World War.

On the other hand, Germany could also be seen as a most likely case for memory entrepreneurship. While the narrative of culpabilization remained predominant at the elite level (Art 2006, 77–85), it has come under pressure at the societal level. Encapsulated in Helmut Kohl's statement on ‘the mercy of the late birth’, a counternarrative emerged that sought to leave behind the mnemonic emphasis on the Nazi past and ‘normalize’ the country (Kansteiner 2006: 126; Manucci 2020: 123–24). The 2006 World Cup, 1999 NATO campaign in Kosovo, and, in particular, national reunification with the Communist East, which had externalized the responsibility for Nazism, fostered this narrative as an important part of the mnemonic discourse (Manucci 2020: 124). It consequently entailed a ‘supply gap’ with regard to alternative interpretations of history,² including positions rejecting the ‘uniqueness’ of Nazi crimes compared to those committed by Communists and colonialists (Urban 2022; Wüstenberg 2019). Such contrasting views about memory politics in Germany make the case particularly intriguing to test our hypotheses.

A Quantitative and Qualitative Text Analysis of Parliamentary Speeches

To test our hypotheses, we draw on speeches given in Germany's national Lower House (Bundestag) between October 2017 and February 2021, taken from Rauh and Schwalbach (2020). Founded in 2013 as a Eurosceptic party in opposition to the Merkel government's EU bailout policies, the AfD rose to prominence during the 2015/16 so-called refugee crisis, when it was a loud voice in national discourse and successful in subsequent subnational and the 2014 European elections (Arzheimer 2015; Dilling 2018). Its emergence challenged the language of mainstream parties who had to respond to the new rival already before the AfD entered the national parliament (Heinze 2022; Valentim and Widmann 2023).³

It is vital to study the AfD's language following its entry into the national parliament. Valentim (2021) has shown the PRR's national parliamentary entry to be important for a country's discursive development by providing these parties with a platform widely covered by the media and legitimizing PRR attitudes among voters. While mainstream parties already had to react to the AfD in subnational parliaments, and subnational party elites advised their national-level parties on how to engage with the AfD before the 2017 election, subnational party responses had varied notably (Heinze 2022; Heinze 2023, 197–200). National-level parties' various formal and policy responses after the AfD's Bundestag entry underlined that there was no ‘magic formula’ to follow (Heinze 2023, 200). Moreover, even if subnational responses had guided national-level mainstream parties' strategies, it was not predetermined that their initial response to the AfD would persist. In fact, parties' responses at the subnational level had ‘changed over time’ (Heinze 2022, 340), and the AfD itself has continued to change its behaviour (Heinze 2023, 200). All this stresses the AfD's first term in the national parliament as an important challenge for the party system (Wiliarty 2021, 167-9). Finally, the time between 2017 and 2021 is well-suited for a comparative analysis. In terms of its ideological outlook and membership base, the AfD was very different in 2017 from the party that had emerged in 2013. The early AfD was primarily driven by Euroscepticism and scholars debated the extent to which it qualified as a right-wing populist party (Arzheimer 2015; Berbuir, Lewandowsky, and Siri 2014; Dilling 2018). Beginning with the 2014 East German state elections, the AfD's transformation into a populist radical right party was marked by the party's split in 2015 and again in 2017 just days after the Bundestag election when groups around Bernd Lucke and Frauke Petry respectively quit (Dilling 2018, 98–99; Heinze and Weisskircher 2021, 265). The AfD in 2017 thus more closely resembles the AfD in the years that followed, and as we know it today, than the party in its earlier stages of development.

To assess the prominence at which different parties speak about history, we established three vectors of terms. The first vector represents explicit references to history and includes the terms Geschichte (‘history’), Erinnerung (‘memory’ or ‘remembrance’), erinnern (‘to remember’), gedenken (‘commemoration’), and Vergangenheit (‘past’). The second vector captures more subtle temporal references to the past and includes the terms früher (‘previously’), damals (‘then’ or ‘at that time’), and ehemals (‘once’ or ‘formerly’), which are indicative of a language centred around the past and often include comparisons between the past and the present. The third vector includes right-wing root terms, which revolve around the collective memory of Nazism and the Second World War. For this vector, we drew on Art (2006) and expanded his list with terms that emerged around the memory of Nazism and the Second World War after the completion of Art's study (for example Halle and Hanau referring to far-right shootings in 2019 and 2020 respectively; see Supplementary Material S1 for the full list and S2 for the raw frequencies).

We chose to focus on these vectors because we are interested in the PRR's politicization of history in general. Historical references carry a complex history of meaning, and the terms that politicians use to refer to historical events are often indicative of political positions (for example using ‘liberation’ or ‘surrender’ for the end of the Second World War). For a quantitative analysis, search terms need to occur relatively frequently across units of analysis and should not be related to particular political ideologies themselves. We expect our three vectors to meet these requirements. By stemming the search terms, vectors 1 and 2 capture a variety of common expressions to refer to history and memory specifically and the past more generally. To capture the highly salient topic of the Second World War references, Art's (2006) – to our knowledge unique – list offers search terms that capture various topics used across political groups to refer to the memory of Nazism and that are the result of an extensive qualitative content analysis and historical research. Our qualitative analysis helped further examine the validity of the search terms and ensured that we did not capture an overly high share of false positives (for example using terms like ‘to remember’ without referring to history). We also assume that such a non-historical use of vector terms is similarly distributed across parties as it reflects general language practices rather than features related to political positions.

We then tested for cross-party differences in terms of the relative frequency with which speakers of Germany's parliamentary groups use the respective terms. We employed a Poisson regression analysis with the share of sentences using historical, temporal, and right-wing root terms as the dependent variables, respectively. At this initial stage, we used party and the total number of words spoken as independent variables.

To investigate the sentiment expressed by AfD speakers compared to those of other parties, we leveraged Rauh's (2018) extensive dictionary to classify positive and negative terms. This dictionary has the advantage that it takes the semantic embedding of words – that is, to say a word's meaning in its specific context – to some extent into account. In particular, it allows us to deal with negations that invert a term's meaning through the use of additional words like ‘not good’.

To test whether the AfD's language overall and around historical markers significantly differs from each other and other parties, we undertook several analyses. First, we determined the de-meaned sentiment per party in the context of each of the three vectors, setting a window of 10 terms around the root terms, as well as the baseline. To do so, we first calculated the relative sentiment per 10-term segment as follows:

$$rel\;sentiment\;_{i, j\;} = \displaystyle{ \sum \nolimits{ positive\;sentiment_{i, j}} \over {\sum \nolimits {words\;in\;segment_{i, j}}}}-\;\displaystyle{{\sum \nolimits negative\;sentiments_{i, j}} \over { \sum \nolimits words\;in\;segment_{i, j}}}$$

where i is the speaker's party membership and j is the root term from the three vectors respectively. We then determined the relative sentiment per root term group, thus not distinguishing between different parties. This group mean is subsequently subtracted from the relative sentiment per party to determine the de-meaned sentiment per party per term group. The de-meaned sentiment expresses whether a party is more positive or negative in its language around the respective root-term group compared to the group mean. In other words, if we picked a random part of speech in the corpus of memory root terms, the variable captures how much the sentiment of the language from one party deviates for this root term group from the average language in parliament in this context. It is therefore not possible to compare the scores directly across root term groups but only within each group.

We ran two sets of models for the general parliamentary language and the three vectors of terms capturing historical language. We specified negative binomial generalized linear regression models after we confirmed an overdispersion of the dependent variable. Our dependent variable was the number of positive and negative terms respectively per speech segment by an individual speaker. We created separate variables capturing the count of positive and negative terms for each segment and speaker rather than a single variable that captures the difference between positive and negative terms. This avoided reducing very emotional speakers, who use a high number of positive and negative terms, to the same value of zero as speakers who refrain from emotional language. Party was our main independent variable. In addition to the total number of words, we controlled for gender, age, foreign born, and region of birth.⁴ These variables have traditionally reflected important compositional differences across Germany's parliamentary groups (for example Geese 2020; Kintz 2011; Kintz 2023), and they have been associated with differences in political communication (for example Adams et al. 2023; Coffé, Helimäki, and von Schoultz 2023; Rittmann 2024). A first set of models regressed the total frequency of positive and negative terms, respectively, within the general parliamentary language on the above variables and we then ran the same model taking as a dependent variable the sentiments measured within ten words before and after each of the entries in our three vectors.

To better understand the AfD's interpretation of historical events, we computed the collocations, which express an association between two terms that is statistically more frequent than could be expected by chance (Evert 2004). We computed these collocations for the three vectors and aggregated them on the party level to understand the prevailing meanings across them.⁵ The analysis of collocations and right-wing root terms already points toward the importance of considering different interpretations in the context of specific historical topics. However, trying to analyze the AfD's engagement with different historical topics exclusively through a quantitative analysis faces serious challenges. Scholars would need to (a) agree on a dictionary of terms that (b) capture the diversity of ways that could be used to refer to particular topics and (c) occur sufficiently frequently in parliamentary speeches for a meaningful analysis. It would therefore miss rare but potentially powerful expressions that accept, minimize, or reject a country's memory regime, and risk missing unexpected topics.

We therefore embedded a qualitative approach in the research design, using the terms included in the historical and temporal vectors to select AfD speeches from the 2017–2018 Bundestag minutes.⁶ An in-depth reading of all AfD speeches and interventions during that time allowed for expanding on this set of speeches. From the resulting speeches, we kept those entries that made a mnemonic reference. This includes, for example, references to the memory of the Second World War, German partition, and the GDR. We included AfD entries making references to the collective memory of other countries, following Schreier's (2012, 83) recommendation to adopt an inclusive approach when selecting material because such references often had implications for the interpretation of German history. We then adopted a thematic criterion for segmenting the material. We coded all sentences that made a mnemonic reference and extracted, when necessary, additional sentences for clarification. We then marked the parts of the extract that discussed the same theme as belonging to the same unit of coding. Units of coding were numbered consecutively within each entry, while contextual units were marked with a ‘c’. Thematic categories were formulated by combining a concept- and data-driven approach. Based on the theoretical reasoning outlined in the previous section, we expected the memory of the Second World War, German partition, colonialism, and European civilizationism to figure prominently. We refined and expanded on these categories through subsumption (Schreier 2012, 115–7).

This approach generated a database including 763 coded units from 317 entries among the 1,134 AfD Bundestag speeches, parliamentary questions, and interventions in 2017 and 2018. We began this part of the analysis in 2017 because we expected the start of the parliamentary term to be particularly relevant for the AfD in positioning itself in Germany's mnemonic tradition. We stopped in 2018 when new thematic categories stopped emerging. Extending beyond 2018 would have thus meant only more entries on the same categories and consequently taught us little new – at least until external conditions or changes in issue salience would start bringing up new categories, which, however, is not the focus of our paper. Our coding frame can be found in Supplementary Material S5, which has been generated and applied to 25 per cent of the material as part of an initial pilot phase. The coding of each unit was double-checked by the same coder three months after the initial coding with consistency being above 90 percent.

The Radical Right's Politicization of History

Prominence: Not an unequivocal mnemonic entrepreneur

To our surprise, and against H1, the AfD is not the expected unequivocal memory entrepreneur. Figure 1 displays the share of speeches that use one of the terms in the historical and temporal vectors divided by the overall number of speeches given by that party. We can see that the AfD's use of these historical markers largely reflects the average across parties (Fig. 1a). Looking at only the raw frequency, the Social Democrats and the Christian Democrats use this language slightly more frequently overall. Yet, this difference relates to the higher number of sentences they speak in Parliament, which the multivariate analysis in Supplementary Material S3 confirms. The Greens and the liberal FDP are the least likely to develop their discourse with historical markers, which is statistically significant for the Greens. Looking at the relative frequency of the vector of temporal markers (Fig. 1b) further underlines that the AfD does not employ such terminology notably more frequently than most other parties.

Figure 1. Relative frequency of historical language, national parliament. (a) Historical markers. (b) Temporal markers.

By contrast, terms typically associated with the memory of fascism are used substantively more frequently by AfD parliamentarians (Fig. 2). Once again the Greens refrain most clearly from using this kind of historical language. The other parties show greater similarity but are all significantly less likely to use these terms than the AfD (see Supplementary Material S3 for the regression results).

Figure 2. Relative frequency of right-wing root terms, national parliament.

We thus find only partial support for H1. Germany's populist radical right is not the unequivocal mnemonic entrepreneur we expected, engaging with history overall as frequently as other parties. However, it does mobilize history with a distinct emphasis on the memory of fascism. To further understand these historical languages, we investigate cross-party differences and similarities in sentiment when speaking about history.

Sentiments: The bleak present and the lost heartland

Historical language is likely to resonate among its audience due to the sentiments that it conveys. We expect radical right parties to use more positive language around historical markers than in their political language overall (H2a) and to be more positive than other parties when referencing the memory of the Second World War (H2b). The descriptive statistics presented in Fig. 3 already provide some insights in this regard.

Figure 3. Mean sentiment score, national parliament.

First of all, we see what basic insights from party politics would make us expect: the two governing party groups (that is CDU/CSU and SPD) use notably more positive language than the four opposition parties (that is the FDP, Greens, Die Linke, and AfD), expressed in a more positive sentiment score overall as well as across our three vectors. Second, comparing our baseline scores across parties supports Valentim and Widmann's (2023) finding that radical right parties use more negative language compared to other parties. The AfD's overall language is the most negative on average. It is important to underline though that its baseline sentiment score resembles that of the other populist party in Germany's Parliament – Die Linke. Third, when looking at the sentiment scores for memory and right-wing root terms, the descriptive results provide some preliminary support for H2a and H2b. The AfD's language is slightly less negative when speaking about history, but substantively less negative when speaking about the memory of fascism. In fact, it is the only party that displays such a pattern. Sentiments around terms associated with the memory of the Second World War are similar to the parliamentary average, unlike in other contexts.

To probe the descriptive differences seen in Fig. 3 more rigorously, Table 1 presents the results of a negative binomial generalized linear regression model. The dependent variable in all cases is the number of positive and negative terms, either per speech (model 1) or around the respective terms in each of the three vectors (models 2, 3, and 4). For a one-unit change in the predictor variable, the difference in the logs of the expected counts of the response variable is expected to change by the respective regression coefficient, holding the other predictors constant.

Table 1. Results of the sentiment analysis: negative binomial generalized linear model

Note: *p < 0.1, **p < 0.05, ***p < 0.001.

Sentiment, baseline language (Model 1): All parties use significantly more positive terms than the AfD in their speeches overall. These differences are significant and especially substantive for the SPD, FDP, and CDU/CSU. The difference in the logs of the estimated counts of positive terms is 0.224, 0.197, and 0.145 units higher for these parties' MPs, respectively, compared to AfD MPs. All parties, except for Die Linke, also use significantly fewer negative terms in their overall language than the AfD. These findings echo Valentim and Widmann (2023), while using a different operationalization of the dependent variable, which justifies confidence in our model.

Looking at the results of Model 1, it is also noteworthy that female parliamentarians are more likely to use more positive language in their speeches than male parliamentarians, adding to recent findings on gender effects in political communication (Adams et al. 2023; Coffé, Helimäki, and von Schoultz 2023; Haselmayer, Dingler, and Jenny 2022; Rittmann 2024). In turn, older parliamentarians use less positive language, holding the other variables constant. As we would expect, speakers tend to use more positive and negative terms the more words are spoken. Finally, parliamentarians from the part of Germany that was under Communist rule during the Cold War (that is the East) tend to use both more negative and positive terms than their counterparts from the country's North-West.

Sentiment, memory root terms (Model 2): When restricting the analysis to the frequency of positive and negative terms within a ten-word distance from our memory root terms (that is our first vector of terms), we find support for H2a. While the other parties (except Die Linke) tend to use significantly more positive terms than the AfD around our memory vector, we find fewer and smaller significant differences between the AfD and the other parties in their use of negative language around the memory root terms. Only the differences between the AfD and Germany's at that time the governing Christian Democrats and the Social Democrats continue to be statistically significant. In contrast to the findings for the baseline model, Die Linke is rather similar to the AfD when it comes to the sentiments around memory terms, which is particularly noteworthy since both parties are also quite similar in how prominently they use historical markers (see Fig. 1a and 1b above). The cross-party differences regarding the use of negative terms also disappear when comparing the AfD with the Greens and the FDP respectively, although it should be noted that both the Greens and FDP also use significantly fewer memory terms than the AfD, which might explain why the coefficients stop being significant. These findings indicate that among its overall negative language, history is the AfD's more positive terrain, as H2a expected.

Some other findings are also noteworthy. Female parliamentarians do not seem to discuss history in more positive or negative terms than male parliamentarians, while older parliamentarians were less likely to use negative terms around historical terms. Moreover, when focusing on their use of positive and negative language around the historical markers, parliamentarians from Germany's North West are no longer significantly different from those from the East, whereas parliamentarians from the South use less positive and more negative language. It might suggest that Southern parliamentarians are less nostalgic than those from the East, potentially as a result of a different mnemonic trajectory and/or better present-day living conditions. Probing different explanations for these differences goes beyond this paper's scope.

Sentiment, temporal root terms (Model 3): The sentiment expressed around the temporal markers provides an alternative operationalization to test H2a. Around this second vector, the cross-party differences are less pronounced. The CDU/CSU and SPD are the only party groups whose speakers use significantly more positive and less negative words around temporal markers compared to AfD speakers. The Greens tend to use fewer negative words than the AfD around this vector. Other cross-party differences are not significantly different from zero. However, we would expect party affiliation to matter less for this vector. It includes terms like früher (‘previously’), damals (‘then’ or ‘at that time’), and ehemals (‘once’ or ‘formerly’), which are more general temporal markers used to refer to events commemorated in Germany's collective memory as well as to more mundane events connected to everyday politics. In turn, and similar to Model 1, female parliamentarians are again more likely to use more positive language than male parliamentarians.

Sentiment, right-wing root terms (Model 4): To test H2b, we investigate cross-party differences in the sentiment expressed around terms associated with the memory of the Second World War. Around such terms, all party groups except Die Linke use significantly more positive language. The CDU/CSU and the SPD also use significantly less negative language than AfD speakers. While we unpack this finding below, it goes against H2b by showing that AfD speakers do not express more positive sentiments around the memory of fascism compared to speakers from other parties. The AfD does not seem to openly challenge Germany's memory regime around fascism and the Second World War, at least not in terms of the sentiments its MPs express around the terms commonly associated with the remembrance of this period. It might suggest that Germany's memory regime continues to restrain the political language of the populist radical right at least to some extent, irrespective of those occasional statements that attract significant public attention.

In terms of the other variables, older speakers tend to use less negative language around right-wing root terms when holding other factors constant. Regionally, only speakers from Southern Germany express significantly different sentiments, using less positive and more negative language around these vector terms than East German speakers. Differences between speakers from East and Northwest Germany are not significantly different from zero. This is noteworthy since such differences would be expected given East Germany's different memory regime after the Second World War (Kansteiner 2006, 110, 120). In turn, the general direction of the effects of the total number of words echoes the previous models.

The results presented thus far support H2a but not H1 or H2b. AfD parliamentarians are not the unequivocal mnemonic entrepreneurs we expected in terms of how prominent historical references are in their speeches compared to other parties (cf. H1). Their more negative overall language is also reflected in their more negative language when speaking about history compared to other parties. Yet, as we move from the baseline language model to the memory root terms, the changing significance levels of the coefficients for partisanship suggest that history is in fact the AfD's positive terrain (H2a). In turn, parties do not meaningfully differ in their use of more indirect temporal references to the past. Finally, AfD speakers seem to adopt a more neutral tone in terms of using less positive and generally similarly negative language around terms associated with the memory of fascism compared to those of other parties (cf. H2b).

The results presented so far also indicate the limitations of a sentiment analysis on its own and the value of an integrated mixed-methods design. Especially, the more positive language of all parties (except Die Linke) when discussing the memory of the Second World War seems surprising. These vector terms include Antisemitismus (‘antisemitism’), Bewältigungspolitik (‘policy of overcoming’), Holocaust, Mahnmal (‘monument’), or Auschwitz (see Supplementary Material S1 for the full list). By capturing the sentiment within a ten-term window around the root terms, it is possible that these parties' positive sentiment score is driven by statements such as, ‘It is great that Germany, after Nazism, turned democratic’. As the CDU/CSU, the SPD, the FDP, and (to a lesser extent) the Greens were governing Germany in one constellation or another since the formation of the Federal Republic in 1949,⁷ we might expect such statements to be more frequent in their speeches than in those by opposition parties like Die Linke and the AfD. Similarly, the Christian and Social Democrats' background as Germany's traditional governing parties might help explain the positive sentiment they express around historical and temporal markers.

Thus, it is important to complement the sentiment analysis with a collocation analysis, which we will do in the next section. It allows capturing patterns in the words around which the vector terms are used.

Interpretations: Continuities and ruptures in views on history

To study the stance and argumentative structure parties adopt when speaking about the past, we analyzed the terms collocating with the entries in our three vectors. To our surprise, the AfD does not notably differ from other parties in terms of its collocations around Geschichte (‘history’), which is noteworthy given that this includes combinations like unsere Geschichte (‘our history’) and deutsche[n] Geschichte (‘German history’, ‘of German history’) (Figure S4a in Supplementary Material S4).

Instead, we find the expected emphasis on collocations that connect the past to the present and use the past as a reference point for today. This includes combining Vergangenheit (‘the past’) with terms such as schon and bereits (‘already;’ as in ‘already in the past’), Zukunft (‘future’), and jüngeren (‘recent’; as in ‘the recent past’; Fig. S4a). It is also striking that the AfD, and to some extent Die Linke, tend to significantly overuse terms that express a moral imperative to keep the memory of the past alive (for example, Geschichte eingehen, lehren Geschichte, erinnern wichtig, moechte erinnern, gut erinnern; in English: ‘engage with history’, ‘lessons from history’, ‘importance of remembering’, ‘would like to remember’, and ‘remembering well’). From among the temporal markers, expressions like schon damals (‘already at that time’) are typical of the AfD's as well as Die Linke's discourse (Figure S4b). In a similar vein, another distinct marker of the AfD is the terminology of früher schon (‘already back then’) as well as früher gab (‘back then there was’). The collocation analysis suggests that the AfD's limited mnemonic entrepreneurship, as seen through the memory and temporal root terms, translates into some differences in emphasis.

The collocation analysis shows that there are strong differences in the ways in which the AfD speaks about the history of Nazism (Figure S4c). The phrase Verbrechen [des] Nationalsozialismus (‘crimes of National Socialism’) figures less prominently in AfD speeches compared to those by other parties, notably Die Linke. By contrast, the phrase Widerstand Nationalsozialismus (‘resistance National Socialism’) is overly prominent among AfD parliamentarians. Moreover, terms such as Volk (‘people’), which have a very strong historical and emotional connotation with National Socialism, are rarely used by parties other than the AfD and the way it is used by the AfD is unique. Its members combine Volk with unser (‘our’) but also deutsch (‘German’) or eigen (‘own’), which most other parties avoid altogether. Only the CDU/CSU and the FDP carefully use this vocabulary and include terms like verpflichtet (‘obligated’) or gewählt (‘elected’) in the context of Volk.

The qualitative content analysis unpacks the AfD's interpretations of history in greater depth. The theory- and data-driven approach generated four overall categories, which capture the AfD's main historical topics. The subcategories within each topic relate to the specific stance and argumentative structure the speakers adopt.

Topic 1 – Nazism and the Second World War (around 25 per cent of coded units):⁸ Around a quarter of the coded units makes reference to the memory of Nazism and the Second World War, indicating that the topic figures prominently within AfD speeches but is by far not the only historical topic AfD speakers engage with.⁹ In line with H3, they express different and at times contradictory positions in Germany's collective memory of the Second World War. A small share of statements, approximately 5 per cent of all coded units, accepts Germany's culpabilization, but it is often used to frame the AfD's present agenda. For example, accepting Germany's historical responsibility for the Holocaust is connected to statements rejecting antisemitism and justifying opposition to the immigration of Muslims into Germany (for example, statements 24.4, 111.2-3, 229.1, 240.1). Similarly, statements defending sections of Germany's Basic Law that responded to the Weimar Democracy's failure to stop Nazism, such as article 18 on militant democracy and article 21.2 on the banning of anti-democratic parties, are used to underpin the AfD's narrative of a present-day broken system (117.3, 206.1, 228.3, 234.2). They are also used to justify the AfD's calls for more aggressive policies toward the radical left and Islamist groups (41.2, 111.4, 228.5). Finally, acknowledging the crimes of Nazism is a way for AfD politicians to express their outrage about political opponents comparing the AfD with Hitler's Nazi Party (for example, 7.1, 62.2, 207.2, 214.1, 230.1-2).

More often, however, the AfD's language about the memory of Nazism is more nuanced and can be classified as minimization – or ‘whataboutism’ (approx. 13 per cent of units). These statements acknowledge crimes committed in Germany's name but relativize them by comparing them to other authoritarian (for example 24.5-6, 37.1-2, 92.1, 216.2, 258.6, 311.3) and democratic regimes (for example 3.2, 35.4, 126.1). Other parties are accused of following in the Nazi's footsteps (for example 62.3-4; 117.1) and embracing racism against Germans (56.1, 62.1, 121.1). Their policy proposals are compared to the centralization under ‘both […] the Nazi dictatorship and […] the SED dictatorship’ (11.2) or are described in terms of Allied post-war policies (for example, ‘re-education program’, 46.3, 215.1). To criticize contemporary policies, AfD speakers adjust notorious phrases like Goebbels's Total War speech when speaking of the ‘total EU/Europe/Euro’ (for example 21.2, 85.2, 107.1, 125.3) or the slogan ‘work sets free’, which appeared above the entrance of the Auschwitz concentration camps (184.4). They also use terms associated with Nazism, such as ‘living space’ (Lebensraum) or ‘forcing into line’ (gleichschalten, Gleichschaltung) (for example, 45.1, 172.3, 211.2-3). The Treaty of Versailles is compared with the Treaties of Rome with regard to the repercussions for Germany, and the EU is criticized for being responsible for ‘financial damage at world-war level’ (26.1-2; also 202.2). A democratic vote on delegating more national sovereignty to the European level is called a ‘self-enabling clause’ (38.2), and the Second World War is considered just another war in a series of conflicts (for example, 31.4, 69.1, 81.5, 109.2, 110.2, 285.3). Finally, Nazism is regularly referred to as ‘national Socialism’ (nationaler Sozialismus) (30.2, 32.2, and 63.5) or ‘another socialism’ (221.2), and the GDR is referred to as Nazism's ‘brother in disguise’ (93.1).

At the same time, in less than one-tenth of coded units, AfD speakers reject Germany's culpabilization. Some statements openly reject Germany's memory regime as shameful and self-denying, opposing the alleged ‘guilt complex that renders the country defenceless […] against all insult, rape, and overrunning’ (62.5), Germany's ‘vacation from history’, and its ‘self-interest-denying flight into pseudo internationalism’ (81.6). They criticize the current regime for allegedly using the ‘Nazi’ or ‘racism club’ (Nazi-Keule, Rassismuskeule) to silence opponents (for example, 47.1, 74.3, 78.1, 102.1, 189.2, 227.1), instrumentalizing the war-time memory to justify a fight against ‘everything that is not on the political left’ (216.3; see also, 63.4, 130.2, 138.2, and 211.1), and for being responsible for Germany's demographic change (59.2-3) and erosion of national identity (49.6, 248.2). Statements demand major revisions to the collective memory of Nazism and the Second World War (for example, 288.3). They reinterpret key historical events such as the country's ‘liberation’ as a ‘capitulation’ (35.1), refer to territories by their name used during the NS period (20.1, 57.5), and blame the Versailles Treaty for the Second World War (285.1). They criticize the Allies and emphasize the suffering caused by their war campaign (for example, 61.1-2, 145.1, 239.1, 263.1), praise Germany's efforts to rebuild after the War (296.1), and oppose the narrative of European integration as a post-war peace project (for example, 81.2-4, 158.1, 202.1).

Topic 2 – Communism (approx. 19 per cent): The narrative of Nazism being merely a different form of Socialism is reinforced by the AfD's reference to the memory of Communist rule. A small share of coded units focuses on the period of the Communist regime and emphasizes Communism's crimes and repression. However, most of the time (around 16 per cent of units), Communist repression is evoked as a point of comparison to claim that present-day Germany has insufficiently broken from its Communist past. This includes statements criticizing Germany's reunification process (for example 70.2) and claiming that Communist elites and ideas continue to exercise a strong influence in Die Linke (for example, 30.1, 41.1, 64.1, 104.1, 177.1, 189.5, 246.5, 300.1), the Greens (for example, 151.2, 161.1, 174.3, 254.1, 281.1, 298.2), and Germany's other parties and institutions (for example, 15.1, 34.1, 39.1, 42.1, 48.1, 53.1, 75.2, 105.3, 118.1, 219.5, 316.2). Present-day politics is equated with Communism (for example, 17.1, 29.1, 60.5, 125.6, 144.1, 188.1, 205.2, 319.1), and the government is criticized for seeking to create a ‘GDR 2.0 extended to West Germany’ (223.1). Terms and phrases associated with Communist rule such as ‘expropriation’, ‘real existing Socialism’, ‘decomposition’ (Zersetzung, that is an anti-opposition strategy employed by the State Security), or Walter Ulbricht's infamous statement ‘No one has any intention of building a wall’ are used to describe contemporary policies (for example, 16.2, 27.2, 51.4, 86.2). In turn, the AfD presents itself as the only viable opposition in an environment where all other forces are far-left (for example, 63.7, 107.3, 161.5). Overall, its engagement with the memory of the Second World War and Communism indicates a strategic politicization of past events to promote the AfD's worldview without unequivocally breaking with the dominant narratives around these topics.

Topic 3 – Germany in Europe and the world (19 per cent): AfD speakers use their greater leeway with regard to topics that have been less ‘policed’ in German mnemonic discourse. In contrast to the Second World War, a critical engagement with Germany's and other countries' imperial and nationalist past started much later (Urban 2022, 84–87). This topic plays an important role in how AfD speakers convey their notion of the ‘lost heartland’. Approximately 5 per cent of the coded units portray pre-1918 monarchical Germany as a positive point of reference, praising its foreign relations (for example, 8.1, 82.3), rule of law (for example 147.1), military strength and values (61.3, 101.2), and education system (for example 139.1, 179.1, 219.4). Former Reich chancellor Otto von Bismarck, in particular, is portrayed as a role model (14.1, 43.2, 50.1, 69.2, 77.4, 95.2, 101.3, 165.2, 275.1, 285.5).¹⁰ By contrast, AfD speakers are notably silent about the fact that Germany was not a democracy before 1918 and that social groups like women, Social Democrats, Catholics, and Jews were the target of discrimination and repression. A negligible share of statements criticize Germany's imperial past, and very rarely do statements downplay and minimize Germany's pre-1918 authoritarianism, often by equating contemporary policies with Prussia's anti-Catholic ‘culture struggle’ (for example, 78.2, 122.1, 211.5, 228.1) and imperial agenda (for example, 10.3, 46.1, 80.2, 116.1, 164.1, 311.8).

Only a relatively small share of statements engages with the memory of German and other countries' colonialism. Statements openly rejecting the notion that the West carries a historical guilt for colonial exploitation are very rare. They include the rejection of an alleged ‘postcolonial ideology’ that aims to establish ‘a mindset of superiority and entitlement among the Global South toward the West’ (46.4). The West's alleged ‘guilt complex’ is portrayed as being responsible for inefficient development aid (132.2, 193.1, 195.2), missed economic opportunities (253.1), and a failed immigration policy (133.1). When discussing the responsibility for resolving current international crises, AfD speakers tend to point at the former colonial powers France and the United Kingdom (for example, 9.2, 10.4, 96.1-2, 153.1, 252.1). While small, this subcategory is relevant since highlighting the human rights abuses committed by other countries can serve as a strategy to question Germany's special guilt as a result of Nazism.

The AfD's argumentative structure changes when moving from the memory of politics in the colonies toward the memory of politics in the metropole. Other countries' history of nationalism tends to be praised, with concepts like patriotism (31.2), sovereignty and identity (49.1, 258.5), and self-preservation (247.1, 248.1) presented as role models for present-day Germany. Particularly noteworthy is that while we have found AfD statements re-interpreting the past of the West's relationship with Russia in a way that legitimizes Putin's foreign policy and annexation of Crimea (for example, 52.2, 73.2, 76.1, 152.1), such statements have been surprisingly rare given the PRR's generally strong links to Russia (Shekhovtsov 2018). At the international level, echoing recent work on the radical right's ‘civilizationism’ (Cerrone 2023), European history is predominantly presented as the history of sovereign nations (for example, 13.1, 60.2, 150.1, 229.3, 257.2), whose shared values (for example 153.2, 175.3, 311.4) were historically defended against non-Christian ‘besiegers’ (210.3).

Topic 4 – The history of pre-reunification West Germany (approx. 25 per cent): As part of embedding their party within the imagination of the noble past, AfD speakers tend to present the AfD as following in the footsteps of non-radical right traditions from West Germany's post-war history. Around 19 per cent of units do this by evoking the memory of former politicians or public intellectuals. Presenting themselves as following in the footsteps of politicians like the former Christian Democratic chancellor and minister of the economy Ludwig Erhard (for example, 21.1, 108.1, 154.1, 245.4), CSU leader Franz-Josef Strauss (36.2, 189.6, 218.4, 298.5), and former liberal foreign secretary Hans-Dietrich Genscher (10.2, 60.1, 199.1) supports the AfD's narrative of simply picking up what Germany's mainstream right has lost. Such political references are rounded off with broader societal connections, with AfD speakers citing or paraphrasing public figures and academics (for example, 24.12, 89.1, 140.1). At the same time, and echoing the party's electoral appeal across political camps (Hansen and Olsen 2019), speakers repeatedly refer to the memory of left-wing politicians and public figures, like Willy Brand or Herbert Wehner, when promoting their party as the defender of working-class interests and a reconciliatory stance toward Russia (for example 43.1, 80.3, 146.1).

References to particular individuals are complemented with statements evoking the memory of other parties' de facto or alleged past positions and policies, such as Germany's social market economy (for example, 72.1, 144.2, 245.3). Accounting for around 5 per cent of the coded units, such statements help underpin what Germany has allegedly lost and present the AfD as the force that will restore this heartland. It is consequently not surprising that we find statements that depict the AfD as the successor to previous mass movements associated with major transformations in German society, like the political protest movement APO in the 1960s and the democratic opposition movement in the GDR (for example, 38.3, 204.3, 224.4).¹¹ Overall, the collocation and qualitative content analysis of the AfD's engagement with history lend strong support for H3.

Discussion and Conclusion

To the best of our knowledge, this article undertakes the first systematic investigation into the extent to which and how the populist radical right (PRR) politicizes history compared to other parties. Bringing together the literature on the PRR and memory politics, we expected PRR parties to refer more prominently to history than other parties (H1), to be less negative when speaking about history compared to its overall language (H2a), and to be more positive than other parties when discussing the memory of fascism (H2b). Moreover, rather than unequivocally defending twentieth-century fascism, we expected the radical right to draw on various de-contextualized historical themes to locate themselves within the country's historical imagination by accepting, minimizing, and rejecting the memory regime in place (H3).

We tested these expectations for the case of Germany by closely integrating quantitative text analysis with approaches linguists use together with a qualitative analysis to gain a deeper understanding of the meaning of language. Against H1, the AfD did not, overall, refer to history more prominently than other parties, but they discussed the history of Nazism more frequently. In turn, we found partial support for the second set of hypotheses. Whereas AfD speakers used significantly less positive and more negative language than other parties' parliamentarians overall, these differences substantively weakened and often disappeared when restricting our sentiment analysis to words around the historical and temporal markers. This suggests that while the AfD's more negative language compared to other parties persists when talking about history, history is in fact the AfD's positive terrain. Yet, in contrast to H2b, the AfD employed significantly less rather than more positive language compared to most other parties around terms relating to the history of fascism.

The collocation analysis revealed notable differences in the choice of words around these markers. AfD speakers stood out in their use of phrases emphasizing national pride and Germany's resistance to Nazism. This impression is further bolstered through our qualitative analysis. While AfD speeches included statements that openly rejected Germany's memory regime, more frequent were those that minimized Nazism or embedded accepting Germany's guilt within a frame that defended the party's anti-Muslim agenda. Such argumentative strategies are best picked up by a qualitative analysis of systematically selected material. Together, they paint the picture of a complex mnemonic entrepreneur, which, in line with H3, uses a range of modes of engagement across and within historical topics.

These findings have important implications for the comparative study of the populist radical right, political language, and the future of liberal democracies. They contribute to the studies demonstrating that the PRR's nativist, populist, and emotional language has challenged existing discursive spaces and helped attract new supporters to their demands for far-reaching revisions to democratic systems around the world. Studies of collective memory have highlighted that a country's mnemonic consensus is foundational for its political identity. Challenging that consensus is one of the most foundational threats to political order. However, the extent to which the populist radical right's discursive fight is also a fight over the (re)interpretation of the past has been underexplored, even though populism's rejection of the present for its betrayal of an idealized past would make us expect the populist radical right to be a textbook memory entrepreneur.

Our findings provide a step toward redressing this gap. They highlight that how rather than how much the radical right engages with history matters. While the AfD was not the expected hyperactive mnemonic entrepreneur, its historical language markedly differed from other parties. Waving different sentiments and historical themes into its parliamentary speeches has allowed the AfD to create a narrative of historical continuity and distance that helps promote its reinterpretation of history without risking immediate ostracization. In particular, shortly after entering the German federal parliament, this strategy allowed the AfD to become part of the political discourse. Rather than unequivocally seeking open defiance, we see more subtle mnemonic attacks. Whereas the high-profile cases that create public uproar are likely to generate pushback against the AfD, the more subtle attacks on historical consensus are less controversial, while having the potential to work their way into parliamentary language and beyond. Rather than by defending Germany's Nazi past, the positive view of the ‘heartland’ is instead constructed by leveraging Germany's pre-1918 history, other countries' national struggles, an allegedly common Judeo-Christian European heritage, and by locating the AfD within the historical continuity of West Germany's post-war political leaders. This strategic (de)selection of historical interpretations and topics embeds the populist radical right within countries' broader historical imagination beyond the memory of the Second World War and thus contributes to its potential normalization.

Our article has proposed a methodological approach to uncover this strategy. Combining quantitative text analysis with a sentiment and collocation analysis around conceptually generated keywords, which are then used to guide an in-depth reading, has been a fruitful approach to test whether the populist radical right differs from other parties in its engagement with history. There is a risk that statements causing a particularly intense public outcry might direct our perception while leaving us blind to the PRR's everyday language. Our approach redressed this by studying systematically both the prominence and sentiment of parties' engagement with history compared to their speakers' overall language. However, investigating whether the PRR differs from other parties in its engagement with history goes beyond investigating the prominence of and sentiments underpinning historical references. Their mode of engagement, meaning their substantive stance and argumentative strategy, are an important part of this investigation. This is where our collocation and qualitative content analysis came in. They helped answer a separate but relevant question for our inferential goal. Studying political language about history thus benefits from what Seawright (2016) coined an ‘integrative’ multi-method design, which our article proposes.

Finally, finding such multi-layered but substantive attempts at re-interpreting the past in Germany, which has had historically a much more rigid memory regime but also a greater supply gap in memory politics than other countries, calls for investigations elsewhere. Scrutinizing whether the populist radical right pursues a blunter and mono-topical approach in countries with a different collective memory of fascism, like Austria and Italy, or with a greater emphasis on individuals' freedom of speech, like the US, is an insightful avenue. Our article has laid the groundwork for such work by providing a rich understanding of the repertoire of sentiments and historical themes the populist radical right draws on in its public (re)interpretation of history.