Introduction

The rise of the far right is threatening the antifascist consensus that helped rebuild Europe and the world following World War II. However, the electoral threat presented by far-right parties is only one manifestation of a deeper social phenomenon. The Western world has never been free of either open or latent racism even in times when there was little parliamentary representation of openly racist parties. The visibility of far-right parties and their threat to established politics are bringing the topic of racism to the public agenda. At the same time, a rising, highly educated and politically conscious group of members of racialized minorities is raising its voice in the public sphere.

Two of these voices in the public debate are those of Özlem Cekic and Reni Eddo-Lodge. Both women are European citizens who belong to minorities and try to counter racism. Both used social media to spur public debate and wrote a book to explain their approach. Despite their similarities, these individuals are situated at two different points in the debate on how to best overcome racism.

Özlem Cikec was the first Muslim MP in the Danish Parliament. Born in Ankara with Kurdish roots, Cikec came of age and became politically active in Denmark. After entering Parliament in 2007, she became accustomed to receiving racist hate mail. Her decision to visit the senders of these mails at their homes and have coffee together to talk about politics brought her international visibility. Her TED talk and hashtag #dialoguecoffee garnered broader attention. Her experiences are detailed in the book “Overcoming Hate through Dialogue. Confronting Prejudice, Racism, and Bigotry with Conversation—and Coffee” (Cekic, 2020).

In 2017, which was the same year as the first publication of this book, the Black British journalist Reni Eddo-Lodge published “Why I’m no longer talking to White People about Race” (expanded version: 2018). Of course, the author is fully aware of the paradox in the title. In the Facebook post that gave rise to the book, Eddo-Lodge knew that she was speaking, perhaps even mostly, to White people. Her refusal of dialogue stems from her awareness of the underlying power structures—here referring mainly to those of structural racism—that exist in dialogical situations.

Whether to talk with or to racists is the question that this research essay attempts to answer. Starting from the assumption that the role of social reactions, the democratic majority and the mainstream media in trying to counter right-wing discourses is widely underexplored, the aim of this article is to understand the diverse material and symbolic effects of certain practices of political contestation. It aims to help elaborate counterstrategies against the threat of the far right and to present communicative strategies against hate.

Discourse studies, with their attention not only to language but also to power relations, normative frameworks, and the combination of symbolic and material reality, seem especially promising in understanding what exactly happens when talking with racists. With the help of ideas from diverse authors such as Michel Foucault, Erving Goffman, and Jürgen Habermas, we will show how democratic positions seem to be in an ideological dilemma in which speech acts that try to counter far-right discourses very often produce the opposite effect. We will use the books by Reni Eddo-Lodge and Özlem Cekic as guiding threads to exemplify these arguments and to connect them from a theoretical level to the practice of antiracist activists.

At the same time, relating theoretical reflections to specific practices of racialized speakers will prevent us from prematurely drawing generalisations about communicative strategies. As the speakers themselves are also part of the complex context of discourse, the particular situation of discourse participants must be taken into account. Being female, racialized, and well educated, as in the case of Cekic and Eddo-Lodge, has important impacts on the possibility of pronouncing effective antiracist discourses.

The reflections presented here should help overcome the pitfalls and performative contradictions of some discursive practices, especially in public communications.

Discourse studies

Discourse studies have done much to further the understanding of the successes and fallacies of the discourses of far-right movements and have provided the means by which to contest right-wing communicative strategies. Classical studies, especially from Critical Discourse Analysis such as van Dijk (1993, 2009) or from the Discourse Historical Approach (Reisigl & Wodak, 2001, 2016; Wodak & Richardson, 2013), help us understand the inner logic of racist discourses and the manifestation of this inner logic in everyday racism. These studies have shown the existence of a racist discursive structure that only seldom appears as open racism and often appears as “racism without race” (Balibar & Wallerstein, 1991), using codes and metaphors that camouflage the racist message.

As a kind of second-order hermeneutics (Diaz-Bone, 2005), discourse analysis does not identify the fully conscious speaker but a social and discursive structure as the origin of racist speech acts. For example, it has been shown that the elementary impersonal semantic structure of racist discourse can be summarised in brief in four basic principles (see Herzog, 2009). First, there is a clear differentiation between “them” and “us”, independent of whether the groups are described in cultural, ethnic, religious, or racial terms. Second, “they” are described as inherently problematic. This means “they” can be labelled uncivilised, dirty, or criminal as well as needy and dependent on social aid due to an essential feature of their being (and not due to a hierarchically racialized and unequal society). The third basic principle of racist semantics is that there already exists an excessive quantity of “them”. In racist discourse, there is always too much of “them” in “our” space, or at least the threat thereof. The fourth principle refers to the understanding of society as a limited space, i.e., “container thinking” (Charteris-Black, 2006).

Discourse analysis has shown how even in the centre of society, these principles are communicated constantly without being seen as problematic and without being perceived as elements of a racist discourse (Herzog, 2009). Furthermore, much research has been performed in discourse analysis on the use of specific metaphors regarding migrants and ethnic minorities. For example, metaphors from the realm of natural disasters are not only exaggerating but also naturalising social conflicts (Charteris-Black, 2006).

However, one of the main contributions of discourse studies is the relation of the textual (or symbolic) level of analysis to other elements of social analysis, such as materialities or power structures (Beetz & Schwab, 2017). Although there are very different understandings and disciplines involved in the development of the postdisciplinary field of discourse studies, one of its core elements is to understand discourses as speech acts. This means that we “do things with words” (Austin, 1962). Beyond the words, there are realities created, things done, and power positions conquered, defended, or questioned. Discourse analysis is always more than a mere text analysis.

The triangle of discourse analysis can be described as the combined analysis of texts, contexts, and practices (see also Angermuller et al., 2014). Here, texts refer to written or oral texts but can also be the textual translation of symbols and images. It has even been argued that all meaningful structured elements can be read as text (Herzog, 2016; Ruiz Ruiz, 2009). Context is a very broad concept and can mean broader sociopolitical and historical backgrounds, as well as concrete speech situations, i.e., the context of interaction, including the speakers with their social positions. In addition, context also often refers to the discursive context in which speech acts are embedded, i.e., to previous and parallel discourses. Regarding practices, these can be caused, induced or shaped by discourses. For example, hate speech can be an incitement to racist practices of discrimination. Furthermore, practice also refers to typical practices of discourse production; writing academic texts, presenting news, or informal chats with the neighbour are all practices of the (re)production of discourses.

Racism is a complex phenomenon with ongoing discussions about its features and main characteristics. Debates exist, e.g., about the ontological status of race, about whether racism is mainly a cognition, an affect or an attitude, or what role individuals and institutions play in the reproduction of racism (e.g., Lepold & Martinez, 2021). In discourse studies, racism has been described as a specific form of discursive exclusion (Herzog, 2009). Following the triangle of text, context and practice, migrants and racial minorities are constructed in text and speech in a specific, negative way different from other members of society. Migrants are often excluded from the practice of hegemonic discourse production. They do not have the same access to the arenas of the public sphere, such as parliaments or mass media. Finally, hegemonic discourses often produce specific social contexts that materially exclude minorities from mainstream society, e.g., through hierarchical citizens’ rights.

In discourse research, these three elements and their relations can be interpreted in very different manners. At the same time, the analysis of the elements, i.e., texts, contexts, and practices, only describes the objects with which we are working. The analysis, however, is usually not an aim in itself. Discourse analysis often aims at another triangle (Angermuller et al., 2014), such as the triangle of knowledge, power, and subjectivation. Discourse analysts are very interested in how knowledge is constituted, validated, or challenged in society. The analysis of power relations can help to understand the circulation of this knowledge. Furthermore, accepted knowledge also helps to ground and stabilise power relations. Therefore, discourse analysis is interested in how power relations are constituted, maintained, or challenged through discourses. Finally, discourse analysts are often interested in the production of diverse subject positions in society, their identity, and their self-awareness. This analysis often goes together with the analysis of “knowledge” about others, i.e., about an alterity from which one’s own identity makes sense.

Regarding racism, we can understand racism as specific knowledge about “the other” that includes categorisations of humans, a specific description of group characteristics and a (hierarchical) evaluation of these characteristics (Holz, 2001). This knowledge contributes to the creation of specific social places or identities for groups, i.e., specific subject positions offered for those identified as belonging to a particular set of human beings. However, knowledge production requires a certain power to exist as well as to exert itself. Modern racism is very much related to the state with its power through educational institutions, citizenship laws, borders and policing practices (see also Schwab, 2017).

To comply with these research goals, discourse analysis draws from three different sources (see Angermuller et al., 2014). Hermeneutically influenced discourse analysis aims at meaning. This meaning is seldom understood as stemming from an original author but more in the sense of a “second-order hermeneutics” (Diaz-Bone, 2005) that situates meaning in the supraindividual space of the social. With regard to the possibility of countering racism, this means that racist discourses do not necessarily express an individual’s conscious and intended meaning but reproduce a socially established way of talking. Through discourse analysis, these kinds of unconsciously transported meanings can be made conscious.

Pragmatics, as the second influential theoretical tradition of discourse studies, is interested especially in what is done, i.e., (re)produced, created, and how this action takes place. Pragmatics understands communication as not only depending on words but also on the (symbolic and material) context of interactions. In any speech act, participants draw on preexisting knowledge. For countering racism, this means that this context and preexisting knowledge must be taken into account even if addressing a specific (racist) situation.

The third theoretical tradition that has informed discourse studies is that of (post)structuralism. The creation of order, patterns, regularities, and structures as well as moments of rupture and subversion are at the core of research questions influenced by (post)structuralist discourse analysis. From this perspective, racism is always linked to a stable and regular interwoven symbolic and material order. At the same time, this perspective often shows how this order is precarious and can be challenged and subverted, as internal racist logics are never able to fully grasp reality.

Meaning, the production of meaning, its relation to the social order, and practical effects in reality are not separated but constitutive interwoven and dependent elements. Regarding disciplinary boundaries and theoretical traditions, discourse studies cannot be thought of only from one perspective but must always be thought about in relation to other traditions and disciplines.

One of the main challenges for the analysis of racist discourses and antiracist contestations is that not all the elements of the analysis follow the same line of logic. Racist “knowledge” does not necessarily lead to racist action. The practical translation from one element of analysis to the other depends on a plurality of conditions. In the same sense, it has also been said that the reactions of the democratic majority and the mainstream media have contributed—mainly involuntarily—to the success of right-wing politics. The media maker, through the inner logics of discourses and of the “discursive infrastructure”, such as the economy of media attention, can produce outcomes that contradict the intention of the individual participant. Therefore, even antiracist speech acts can often have opposite material effects. From “performative contradictions” (Butler, 1997) to “ideological speech acts” (Herzog, 2021), what is said can sometimes be in contradiction to what is done through the speech act. As racism is such a complex phenomenon, antiracist contestation has to take into account the aforementioned different levels and elements of racism and cannot be limited to an easy, well-meant definition.

Power, subject positions and materialities

Armed with these intellectual tools, we can now re-examine our question of “overcoming hate through dialogue” (Cekic, 2020) or “no longer talking to White people about race” (Eddo-Lodge, 2018).

From the preceding summary, we can understand that it is not an abstract but a theoretical question whether directly affronting racist speech is an action. Speech acts do not exist “as such”. Text and talk are always embedded in contexts, structures, and power relations; speech acts are performed by and to concrete agents, draw on preknowledge and other symbolic and material resources and have important effects. The question of whether to enter a communicative interaction must include the questions of who, when, and how to enter which specific communicative situation.

In her book, Cekic compared the hatred towards Muslims in Denmark with her former (and other Muslims’) hatred towards non-Muslim Danes. In both cases, hateful stereotypes, prejudices, and generalisations create a situation of social distance preventing identification with the other. Although this is correct from a formal point of view, it totally omits the social context and the power relations at stake. Racist stereotypes, and not anti-Danish stereotypes, affect the lives of ethnic minorities in Denmark from their job prospects and health conditions to even their life expectancies. Anti-Danish prejudice lacks this power. Racism is not just prejudice or hate. People have prejudices towards all kinds of groups. They can also have prejudices against white, middleclass men. One could hate supporters of the Chelsea football club or Real Madrid for many reasons.

The formal analysis of hate omits the social dimension, which is at the core of discourse studies. Racism is not only hate. Racism is hate plus power. The term “hate speech” is therefore misleading in regard to racism. Racism includes hate discourse, i.e., speech acts that have the power to create reality, subject positions with corresponding social hierarchies, material effects, etc.

Of course, racism is more than individual hate. I can hate my neighbour or my ex-boyfriend independent of whether these persons belong to a minority group. In contrast, racist speech acts draw from a socially available stock of (racist) knowledge. Specifically, they draw from powerful possibilities to speak from a social system that legitimises these kinds of discourses.

Therefore, neither the racist discourse nor the racialized power structure originates in the participants of linguistic interactions. Of course, structures need to be reproduced by social agents to function as structures (see e.g., Bourdieu, 1991). This means that a change in these practices (including speech acts) can change the racializing structure of society and discourses. However, it is not up to the individual actor in the specific situation to change this structure. Although several individual contestations and alternative discourses can change the awareness of the hegemony of particular discourses, every single speech act, in every specific situation, is still embedded in a structurally racist society.

When contesting racist discourses, one must be aware that all speakers are embedded in this structural situation of inequality. Often for participants negatively affected by racism, the discourse is not about abstract problems but about them. Eddo-Lodge states it directly, “If you are an immigrant—even if you’re second or third generation—this is personal. You are multiculturalism. People who are scared of multiculturalism are scared of you” (Eddo-Lodge, 2018: p. 19). White communication partners have the privilege of not depending on the outcome of the conversation. They can afford to be uninterested. Even Cekic, who insists on the need to talk to racists, writes about several encounters generating important psychological stress. In a structurally racist society, the racialized individual is in a weaker position.

At the same time, the lack of minorities in discourses about minorities is in itself problematic. It has been widely researched how the discourse about minorities is mostly this: a discourse about minorities and not a discourse created by or with minorities (see e.g., Herzog, 2009). This underrepresentation of minorities in discourses that negotiate their public identity creates biased discourses, i.e., a structurally subordinated identity for minorities. At the same time, these unequal practices of discourse production are constantly reproducing the unequal power structure in which specific social groups have privileged access to shaping public opinion.

These structural inequalities, the power effects of discourses, and the different possibilities for access to public visibility and attention cannot be ignored when analysing the possibilities of countering racism. Although it is nobody’s fault or merit in being born with a specific skin colour or sex, one can critically relate to the social consequences stemming from this situation. This could mean that instead of contesting hate directly, White participants could also choose to step aside and let others co-construct the discourse. This practice not only creates alternative speech but also creates alternative power relations where minorities are not the object but the subject of the production of discourses and social structures.

Of course, structural inequality does not simply vanish with the presence of minority speakers. It is still there and can be felt as a powerful oppressive structure for those speakers. However, again, structural inequalities can be made aware by speakers contesting right-wing discourses. With Habermas (1984), we can understand that there exists the possibility of a meta-discourse. The discursive situation itself can be put at the centre of the debate. In other words, instead of engaging in a debate about whatever topic, one can aim at centring the debate about the distribution of the power to speak, to be heard, and the unequal material and emotional consequences of discourse.

In her attempt to counter racist discourses, Cekic does not follow this strategy. Perhaps she is following David Graeber’s advice, who recommended “the defiant insistence on acting as if one is already free” (here: Graeber, 2015; see also Graeber, 2013). From this viewpoint, deliberately ignoring racist structures can also be a strategy to counter racist structures. However, here, we could fall into the trap of colour-blind racism, i.e., the ideology that the best way to end discrimination is not taking into account the ethnic or racial background of our interaction partners. This may seem to be a reasonable approach to achieving equality. However, in our societies, ethnic and racial backgrounds matter. Pretending to be blind does not make these structural inequalities producing discrimination go away.

Pat Parker impressively captured the dialectic of colour-blindness in the first two lines of one of her poems:

For the white person who wants to know how to be my friend

The first thing you do is to forget that I’m Black.

Second, you must never forget that I’m Black.

However, David Graeber does not say that inequalities do not exist and that we have to ignore them. His mention of the “defiant insistence” makes clear that one can be fully aware of the discriminatory structure, but one does not have to submit to these structures. Instead of trying to counter oppressive structures using meta-discourse about their presence in the specific communicative situation, this approach would mean a practical resistance by not submitting to discriminatory structures, deliberately ignoring them and, thus, not taking part in their reproduction.

Norms

With Habermas (1984), we can say that in communicative processes, there is a certain normative basis of the communicative process itself that is presupposed and renegotiated. He mentions the four universal validity claims: (1) comprehensibility, (2) truth, (3) rightness, and (4) truthfulness. In every moment, we can theoretically call into question the comprehensibility of a statement, contest its factual truth, criticise the normative rightness of the relations expressed through the communicative situation, or call into question the subjective truthfulness of the participants. For example, when questioning the truth about a racist statement or criticising the normative rightness of racist language, one is interrupting the normal way of holding a conversation by engaging in a kind of meta-discourse, or a second-order discourse. However, at the same time, by entering this meta-discourse, one implicitly accepts certain norms of a third order.

By debating claims, one is implicitly accepting that these statements are worth debating and that they can rightfully be debated in this specific situation. Now imagine a speech act negating the existence of the Holocaust. Holocaust negation in some countries is even considered illegal. By providing arguments against Holocaust denial, these affirmations that question the very existence of this genocide are elevated to the selected group of speech acts that can be legitimately made in a debate. In other words, even by contesting a speech act, one legitimates its possibility.

In his inaugural lecture “The order of discourse”, Foucault (1981) described other ways of responding to this kind of communication offer. Instead of contesting them, respondents could exclude them, treat them as “noise”, insane, or as standing outside of the rules of truth production. These forms of treatment are also contestations, but they could be described as practical contestations that do not give legitimacy to the communicative offer in the statement of the other.

It is important to acknowledge that the debate of contesting by communication or by exclusion is not merely theoretical, as one must consider the specific situation in which the speech occurred. It is not the same for a ten-year-old to be confronted with the negationist slogans on the internet as for an academic to counter the negationists who try to discuss this thesis in an academic setting. There are different social spaces where different types of discourses can be made.

At the same time, one must be aware of the available social power to exclude. Extreme positions cannot always be easily excluded, effectively marked as insane, or as standing outside of the rules of truth production. In society where far-right parties have entered important positions in politics and the media, these positions often have effectively entered many social spaces from which it is now difficult to exclude them.

Nevertheless, the de facto inclusion of certain positions into social spaces does not mean that one must confirm this inclusion. Again, the positions of Cekic and Eddo-Lodge can be considered contrary in this regard. By sitting down and talking with racists, Cekic is acknowledging the legitimate interest of these people with regard to, for example, a safe environment, economic wellbeing, and worries about threats to their identities. However, it must be said that even for Cekic, there are red lines that justify not talking to racists. Criminal comments and threats made against her were not acknowledged as legitimate speech acts but handed to the police.

Eddo-Lodge, on the contrary, does not want to talk to people who deny the existence of structural racismFootnote 1. There are people from the White majority who deny being in a structural advantage or who think that White people in Western society are constantly being threatened by ethnic minorities, thus creating the idea of reverse racism. By agreeing to talk to these people, Eddo-Lodge would have accepted an unequal speech situation in which she would be forced into a situation where she had to justify not only her arguments but her very existence in this society.

Here, the issue about whether to talk to racists must face one important question that has yet to be answered—why should we talk to racists? What do we expect from this situation? Do we want to change some of the basic assumptions of the other? Do we want to change the ideas of bystanders or the public? Do we expect to learn something from racists, i.e., do we accept the idea that racists can make us change our minds? Or do we want to understand racism from a scientific point of view? We use the notion of understanding here in the Weberian sense of Understanding Sociology, i.e., a way of trying to comprehend the inner reasons of the acts and thoughts of the other. At first glance, understanding is not simply agreeing but aims at getting to know the subjective, interior sense of the other. This analytical approach to understanding can then later be used for all kinds of reasons, e.g., to develop counternarratives to racism. In summary, the question could be understood as whether we are talking to or with racists, whether we want racists to talk to us, or, as in the case of bystanders, we truly are talking to a broader public.

The question about who the addressee is of the conversation has rarely been touched by discourse studies. Discourse as an impersonal structure seems to spread all over society. However, it has different effects on those who speak, are spoken to, or listen as bystanders. Bystanders, or the “Third” as Simmel (2009; see also Fischer, 2013), names it can have important effects on the legitimation of communication. The sheer presence of the Third, can change the communication situation and the social implications. For example, by not intervening, the Third is confirming the rightfulness of the communication situation.

One of the approaches to the different effects on different individuals is the adoption of subject positions (Angermuller, 2014). Speaking can create different identities and alterities through narratives, appellations, labelling, and so on. Depending on the counternarrative, the result can be the creation of two identities: (1) the good, nonracist identity and (2) the bad, racist identity. The counternarrative could now (a) reinforce one’s own identity, i.e., the certainty of moral superiority; (b) try to persuade bystanders to come (or stay) on the side of this positive identity; (c) try to convince the other of this moral inferiority of racism, inviting the other to change sides; or (d) label the other as racist, thus producing an exclusion of the other and its discourse.

However, counternarratives are not compelled to create opposing identities. It can be thought of as narratives that do not divide the world into black and white or good and bad. In this counternarrative, racism would then be seen as something that is reproduced by (almost) all of us to different degrees (see Herzog, 2019). Here, it seems that we are not facing two different subject positions that are categorically different. Rather, we are facing the very same subject position of the racism-reproducing subject. The differences between the subjects would then be only gradual. Nonetheless, here too we could think of two categorial different identities: (1) those who face their racism, thus trying to behave in a more ethical way, and (2) those who deny or even justify being part of the reproduction of racism. Again, the discourse can have the four different effects described above: (a) reinforcing one’s own position, (b) persuading bystanders to confront their own racism, (c) trying to convince the other of their implicit racism as a first step to overcome this racism, or (d) excluding the other who is not facing his or her racism and the related discourse.

In her book, Eddo-Lodge used the first three strategies. By presenting her own position, she is confronting the reader with her own embeddedness in racist structures while convincing the other that such racist structures exist. Cekic, on the other hand, is also trying to blur the clear line between the good, antiracist identity and the bad, racist identity. However, as she is doing so by seeing racism as just another form of hate and prejudice or a different form of framing one’s legitimate worries about the future, she cannot develop a structural notion of racism as a power structure. In a structurally racist society, individuals have, from the very start, different positions that Cekic is unable (or unwilling) to detect for the sake of creating an unbiased communication atmosphere. However, by doing so, she accepts the structural racist bias of society.

Another important issue is the framing of the conversation (see Goffman, 1974). Frames are culturally, socially, and contextually determined definitions of reality. These frames allow the participants of an interaction to make sense out of objects and events. As in the case of a painting, frames pose certain limits. At the same time, they allow for certain freedom regarding the content. Therefore, frames do not determine the exact content of what is painted (or said), but they are a very effective way to limit what theoretically could be painted (or said) to a very small unit that is almost impossible to cross, at least if one is to leave the frame intact. Therefore, for example, it has been criticised that, in Germany, there was an unusually high proliferation of television debates on various aspects of migration and cultural diversity and that almost all of these debates framed migration as a problem. Once one accepts migration and migrants as a problem, even the most benevolent speech acts, the most progressive interventions, and the best of intentions turn into a blunt knife. Independent of the will of the participants, what is communicated is the validity of the frame, i.e., the validity of the perception of migrants and migration as a problem. Lakoff (2004) showed that identities or metaphors could also work in an analogous way. Once one accepts an identity or metaphor as valid, one is bound to its limits, such as a painting being limited by the frame.

Regarding our issue of talking to racists, these reflections bring us to the question of what should we say to racists? By accepting the topic of communication, we already impose an enormous limitation on our conversation. By taking part in a radio debate on the problem of migration, one is already accepting that migration is indeed a problem. Everything that can be said within this frame implicitly reproduces the very idea that migration is a problem.

The constraints of this situation were clearly lived by Özlem Cekic. Racists do not want to talk about racism. They want to talk about migration, Islam, or threats to “our” way of life. By accepting this frame, Cekic is put into a defensive position. In doing so, she then has to show her loyalty to Danish society or share certain concerns about radical Islam. Moreover, as said above, this kind of debate is not abstract; it is about the very existence of Özlem Cekic as a Muslim in Danish society. She not only has to defend some ideas but also has to defend herself. This is the material power of frames. In some of her descriptions of the communication situation, one can concretely feel the power and satisfaction of the White interlocutor in this situation. Switching from a debate about migration towards a debate on racism is almost impossible.

Conclusions

Talking to or with racists is a complicated task. One of the possibilities to overcome communicative pitfalls would be not talking to racists. This position can give the appearance of a radical ideological purity distancing oneself as clearly as possible from racist positions. Nonetheless, whether this strategy is also the best one to combat racism is a different issue.

On the other hand, always praising the goodness of communication without analysing the conditions of the communicative situation can equally help create a positive self-image as a dialogical, tolerant, and open-minded person. However, as we have seen, the outcome of the dialogue does not depend on the arguments interchanged in this situation but on the power structures, communicative frames, and normative epistemes embedded in the dialogue.

When thinking about entering into communication with racist positions, there is no clear, easy, and once-and-for-all answer. One has to reflect about the addressees, the topic, the framing, the bystanders, the material and normative situation, the structural power involved, and many aspects more before being able to assess the benefits and costs of engaging in dialogue.

Discourse studies, with its expertise on power, knowledge and subject positions (the triangle of aims of discourse studies), with its exposure of text, practices and context (the triangle of discourse analysis), with its rich conceptual tools such as materialities, material and symbolic realities, norms, and frames, etc., and especially with its insights about the interplay of these elements, can help to disentangle how the outcome of a communicative situation depends on more than just the words chosen or on the intentions of (one of) the participants. If racism is more than an individual attitude but a form of social organisation, then the question also must be how engaging in dialogue can help to change the underlying racialized power structure. Structures are reproduced by human beings. However, it might be almost impossible to change racist structures without human beings being conscious of the structural character of racism and the racist character of social structures.