Alternative Facts as Communicative Practice

The last years have seen a surge in academic and journalistic investigations into what has been called the “post-truth condition.” Journalists and experts diagnose a societywide fear of living in an age “in which objective facts are less influential in shaping public opinion than appeals to emotion and personal belief,” as the Oxford Dictionary defines it in its rationale for why it chose post-truth to be the word of the year 2016 (Oxford Dictionary 2016). This discursive conjuncture of questions revolving around truth has triggered academic interest in various communicative phenomena assumed to contribute to the crisis of truth: fake news, conspiracy theories, and alternative facts. Running through the debate is a shared conviction that what matters about these phenomena is their creation of misleading knowledge about the world we inhabit, how these misguided beliefs spread and why people believe them.

Without disregarding the importance of these issues, the study (Trautmann & Kumkar 2021) this article is based on sought to shed light on a hitherto under-investigated question, namely how alternative facts function in social media interactions. This necessitates a shift in perspective: It takes as its starting point not the supposedly misguided beliefs, but the very interactive processes through which alternative facts are articulated and replied to in discussion. Here I set aside for the moment the question of whether or not these beliefs are misguided, and, more importantly, if they are actually believed by anyone. My questions in designing the study were thus: How do people articulating alternative facts in Facebook discussions do an alternative fact, and how is this doing processed in interaction?

In answering these questions, I draw on sociological approaches and the documentary method, as it was developed by Ralf Bohnsack (1989), building both on Karl Mannheim’s sociology of knowledge (1997) and Harold Garfinkel’s ethnomethodology (1959; Garfinkel & Sacks 1970). The documentary method, which can be both understood as a theory of a sociology of knowledge and as a method for interpreting qualitative data, calls us to focus on shared, implicit knowledge orienting and structuring communication. It thereby allows the researcher to reconstruct the implicit rules by which communication of alternative facts in social media interactions is practiced.

The study “Alternative Fakten im Gespräch” (alternative facts in conversation) was conducted by Hannah Trautmann and me in 2020 and 2021, and by discussing its methodological perspective and interpreting some of its core results,1 this article proposes that the documentary method can contribute to the growing toolbox of digital anthropology. Digital sociology and anthropology are different fields, but they share an interest in practices related to digital technologies. Considering that digital anthropology relies primarily on ethnography, and that digital ethnography itself is understood as a processual endeavor that develops its approaches adaptively (Pink et al. 2016: 33). Analyzing the communicative function of alternative facts enriches our sociological and anthropological understanding of digital truth-making by pointing out how the vernacular practice of sharing and talking about alternative facts is itself embedded in a culture of interaction. This culture of interaction is both supported by the affordances of Facebook communication and by the habits of interaction possessed by those involved in these interactions.

To highlight the innovative potential of such a shift in perspective, the article opens with a short summary of the main threads of scholarly discussions on alternative facts and an outline of our study’s methodological approach. The analysis is summarized along three basic findings: (1) doing alternative facts in social media conversations must be understood as structured by an identity performance of being one of those who know; (2) the identity performance is embedded in a conversational dynamic of collaborative mistrust; (3) the conversational dynamic is immunized against criticism by constructing and policing a distinction between those contributing to the collaborative effort of unmasking alleged lies in mainstream discourse and those obstructing this effort. As a result, the conversational dynamic is marked by an indifference toward the facticity of the content of alternative facts. The article closes by sketching out possible venues for connecting these findings to the existing research on alternative facts.

Of Bots, Bubbles, and Fools: The Discourse on Alternative Facts

It comes as no surprise that the epistemological debate on alternative facts has formed along the lines of pre-existing conflicts around social constructivism simmering since the science wars.2 What is more surprising, given the intensity of this debate, is how little impact it has had on the framing of empirical studies on alternative facts: Most empirical studies from media studies and psychology have assumed that alternative facts can best be characterized (or are problematic primarily) as empirically wrong or socially heretic beliefs. Quantitative studies in social psychology have highlighted how diverse cognitive biases contribute to the susceptibility to believing alternative facts (e.g. Bakir & McStay 2018; Nocun & Lamberty 2020; Pennycook & Rand 2020; Rutjens et al. 2021). Explorations in media studies have focused on the question if certain traits of social media platforms could facilitate the spread of and susceptibility to alternative facts. The most prominent explanations are (1) the rabbit-hole argument, according to which the algorithms of social media platforms tend to direct users to more extreme content to keep them engaged, fostering political radicalization and high informational selectivity (e.g. O’Callaghan et al. 2015); (2) the echo-chamber/filter-bubble argument, according to which the self-selectivity of personalized networks minimizes exposure to divergent information, thereby heightening the probability of confirmation bias (e.g. Bakshy, Messing & Adamic 2015; Spohr 2017); and (3) the missing gatekeeper/bots argument, according to which the lack of enforceable quality standards for news on social media platforms makes it easier for malevolent actors to place and circulate misleading information.

Together, these approaches offer a theoretically plausible narrative for explaining the spread of alternative facts on social media platforms: Malevolent actors (bots) can easily place information online, which circulates in self-validating media environments (bubbles), thereby tapping into psychological predispositions of users who lack the media competency to discriminate between trustworthy and untrustworthy information (fools). The problem with this neat explanation is that the empirical evidence for every single element of this narrative is mixed at best. Ledwich and Zaitsev (2019) provide an analysis of YouTube recommendations with regard to radicalizing content, and argue that the rabbit hole argument does not stand up to scrutiny, because the recommendations actually steer viewers toward less inflammatory, more mainstream content. Guess and colleagues (2023) recently found that changing Meta’s algorithm for what is and is not displayed on peoples’ timelines does not seem to influence their behavior or attitudes, further weakening the case for social media algorithms’ role in influencing peoples’ perceptiveness for extreme political content and alternative facts. Flaxman, Goel and Rao (2016) investigated the impact of social networks and search engines on news consumption and found that the use of social networks actually increased the exposure to a diversity of news sources from across the political spectrum. In particular, the research on the role of social media bots has highlighted the importance of human agents in actively spreading alternative facts (which seem to spread faster and further than “true” information; Vosoughi, Roy & Aral 2018).

All of this suggests that the model of a manipulative environment and its naïve inhabitants is missing a central element: the spread of alternative facts as a communicative and therefore contingent social phenomenon. It is, in other words, something that (lots of) people have to do to make a difference in communication. This doing cannot be understood by framing it negatively as failure (of judgement), instead it requires competence and practical judgement: one has to be able to communicate alternative facts in a way that resonates with others, so that the alternative facts can make a difference in communication. This question of what it actually means to successfully communicate an alternative fact (in the sense of them being taken up and making a difference in communication, not necessarily in the sense of convincing anyone) has however remained under-theorized and under-researched in the literature on alternative facts. The very few qualitative studies on alternative facts rely on interviews and therefore on reflections on doing alternative facts rather than on a reconstruction of the practice itself (Duffy, Tandoc & Ling 2020; for conspiracy theories: Buchmayr 2019).

Reconstructing the Doing of Alternative Facts

Social media interactions are a promising point of entry for investigating the practice of doing alternative facts. First, there is a widespread assumption in both journalistic and academic discourse that the rise of social media has been crucial for understanding the articulation, circulation, and consumption of alternative facts. The plethora of empirical media and communication studies research on the topic bears witness to this assumption, as do theoretical arguments that put forward the communicative structures of digital publics as the central element for understanding the so-called crisis of truth (Habermas 2021; Staab & Thiel 2021). A better understanding of the functioning of alternative facts in social media will help to further illuminate the nexus of social media and spread of alternative facts, regardless of if one agrees with the underlying hypothesis of the central causal role of social media in the rising prominence of alternative facts (or even with the idea that their importance has risen at all [Carlson 2020]).

Methodologically, furthermore, social media conversations are a promising entry point for qualitative reconstruction because of the peculiar form of data they constitute: Digitalized interactions are self-documenting interactions (Schwarz 2021) and, in the case of Facebook conversations, quasi-public self-documenting interactions. This allows for de-localized and de-temporalized observation. Since the conversations do not follow the classical understanding of interaction as constituted by bodily co-presence of the participants (Baraldi 2015; Goffman 1983) and are potentially never-ending, the researcher can scroll through them, read, and interpret them from a perspective that is not categorically different from the way an everyday participant would perceive these conversations: they too would scroll through, read, and potentially comment on the conversation. Of course, this factors out a great deal of complexity, including the embeddedness of digital practices in the everyday lives of their participants and the material, embodied quality of doing social media communication (e.g. Bareither 2020; Couldry 2004). Yet compared to the time and effort needed for collecting material for classical conversation analysis, which required the in-situ recording of conversations, this accessibility is a huge advantage. Where once one had to travel to meetings of protest movements or to recruit focus groups (e.g. Kumkar 2018) that were then recorded and transcribed in order to be interpreted, the exchanges on social media platforms offer themselves as data that is easy to access and already existing in written form.


To approach alternative facting, the doing of alternative facts, as a communicative practice in concrete interactions, our study relied on the documentary method approach to conversation analysis (Przyborski 2004; Bohnsack 2004; Bohnsack, Przyborski & Schäffer 2006). The documentary method’s focus is to reconstruct implicit orienting knowledge that allows for competently partaking in communicative practices. By comparing how different contributions to an interaction relate to each other and by comparing modes of relating across different interactions, this method aims at identifying the implicit rules structuring these interactions.

The documentary method is well suited to reconstructing the practice of doing alternative facts from social media conversations for two reasons: First, by conceptualizing the conversations as autopoietic, self-referential systems, it allows to turn the necessity of not knowing the subjective perspective of the participants in the conversation (and, in fact, not even the participants themselves) into the virtue of focusing on the immanent dynamics and rules of the conversations. Second, beyond the focus of classical conversation analysis on the formal traits of conversations, the documentary method calls for including the semantic content of conversations in the interpretation. More precisely, through reconstructing the dramaturgy of discourse, it aims for grasping the knowledge shared by participants about the implicit rules orienting the discourse (cf. Bohnsack 2010: 123ff.). This dual focus of documentary conversation analysis is a fruitful tool for the digital ethnography of alternative facting because it allows the researcher to first consider the conversational dynamics on the Facebook pages of the “Alternative für Deutschland” (AfD), an extreme right-wing party in Germany, as a culturally specific way of engaging with the affordances of social media, and second, to reconstruct the specific role of communicative forms within this conversational dynamic.

I assume – in line with the methodological literature on conversation analysis for online interaction (Reeves & Brown 2016; Farina 2018; Meredith 2019) – that the basic formal traits of conversations on social media can be understood as being homologous to those in face-to-face interactions: In both types of conversations, communication (1) unfolds as (quasi) real-time-interaction and (2) delineates itself from its communicative environment by each communicator referring to the others’ utterances as belonging to the same conversation. Participants therefore must somehow (3) process the problem of turn-taking, (4) develop techniques for dealing with (mis)understanding and repair breaches of communicative scripts, all of which must be accomplished in (quasi) real time. This necessitates reciprocal interpretation of those answering each other and thus collaborative sensemaking, and turns utterances into reactions that have a relatively fixed place within a broader communicative context. The extent of the homology between face-to-face interaction and social media conversation should still be regarded an open question (cf. Schwarz 2021), but interactions on Facebook share at least enough features with spoken interaction for conversation analysis to shed light on differences between the two. I therefore do not suggest that this kind of interaction is necessarily perceived as conversation by those partaking in it, but the decision to speak of “social media conversations” merely highlights the methodology of approaching them.


The abundance of online data makes a careful sampling all the more important – especially when the method of interpretation allows for only a relatively small amount of data to be meaningfully interpreted. From all the possible examples for alternative facting in social media, the study chose conversations on the Facebook pages of the AfD because the right-wing populist AfD as a party positioned itself firmly in fundamental opposition to the federal government’s measures to control the Covid-19 pandemic, and its members frequently participated in protests against these measures and even voiced alternative facts about the virus in parliamentary debates. Facebook was chosen because it is still the most broadly used social media platform in Germany and because its users are, compared to other social media platforms, often from the very same demographics that are said to be receptive toward alternative facts – less media-savvy, older persons who tend to have lower educational degrees than users of other social media platforms (Beisch & Schäfer 2020).

As a first step, Hannah Trautmann scrolled through and read approximately 350 conversations which occurred between February 2020 and June 2021 (focusing especially on September to November 2020) on the Facebook pages of the AfD on the national and state level, as well as from selected members of the Bundestag, the German national parliament. From this dataset, around 90 conversations were selected and archived. They included alternative facts elaborate enough to allow for meaningful interpretation. Out of these, 13 conversations were selected for detailed reconstruction.

After the analysis of the first conversations, additional cases were chosen, considering if the conversational dynamic seemed to provide instructive horizons for comparison – for example if the articulation of alternative facts was met with criticism or if this criticism replied to alternative facts in different forms. However, throughout all the conversations, I identified a relatively uniform pattern when it came to their conversational dynamic and the role of alternative facts in this dynamic. The examples provided in the following were chosen because they turned out to be illustrative in the sense that the short quotes already allow for discussing the basic structural elements of the conversation dynamic.

To protect the privacy of those partaking in the conversations, all names have been replaced by pseudonyms, whereas I tried to reproduce the core identity markers in the original profile name. For the same reason, the following quotes will not be referenced with links to the original conversations, which were however saved and archived after the project’s completion.

Collaborative Mistrust: Un-truthing the “Sheeple” Discourse

All conversations containing alternative facts display one overarching pattern: (1) the dominance of identity performance as a structuring principle of interaction, (2) a recurring thematic dramaturgy of the overall conversation which can be defined as collaborative mistrust marked by the shared understanding of perceived inconsistencies in mainstream media, pointing to a deeper truth to be uncovered and the collaborative collection of clues supporting this suspicion, (3) a self-immunization against mainstream criticism in which the facticity of alternative facts is rendered less important by emphasizing a sharp “us” versus “them” distinction. I understand these mutually constitutive aspects as three perspectives on the same object – a collaborative, communicative “un-truthing” of what is perceived as mainstream discourse. In what follows, I present these three aspects with illustrative examples from the conversations.

Identity Performance: Intimacy with Strangers

The first thing that struck me during my first, “long preliminary soak” in the material (Hall 1975: 15) when I was scrolling through the timelines of the different Facebook pages, was the emotional intensity of the interactions. I understand the emotionality of the interactions as the affective dimension of a structuring principle of communication on these Facebook pages, a principle I call intimacy with strangers.

There is no sign of commentators personally knowing or even recognizing each other. Yet they address each other in a highly personal way – they share their emotional states (“I am frustrated,” “I feel the same way”), comfort each other over declared stress (“you are not alone”), or attack each other as “stupid.” It seems as if the participants in the discussion share my first impulse when reading these comments – the urge to find out who the persons behind them are – and act upon the perceived identity performance at least as much as upon the propositional content of the respective comments. The conversations are therefore never purely about the explicit political topic, but most of the time chiefly discussions about those participating in them. This becomes especially palpable with the policing of the “us” versus “them” distinction which occurs whenever the identity performance of participants is problematized (see below).

Identity performance is understood here as a universal aspect of communicative practice. As Schimank (2016: 142–165) elaborated with recourse to the tradition of symbolic interactionism, identity can be understood as the actor’s self-image (s)he projects in social interaction and which (s)he seeks to validate as her/his communicative address. Identity performance is therefore an ongoing task of every communicative process – but it is an especially challenging task in the conversations under scrutiny here for at least two reasons. First, other than in interaction with friends, relatives, and acquaintances, the communicator cannot rely on a pre-established consensus on his/her identity, sedimented from earlier interactions. By commenting on a Facebook conversation composed of comments made by strangers, one must build one’s identity performance from scratch, one cannot rely on pre-established social roles which usually lighten the burden of identity performance in interactions with strangers. The “collapsed context” of quasi-public communication on social media forces the communicator to reckon with the fact that one has very little control over who perceives one’s communicative utterances and in what context they interpret them (Hogan 2010). Second, the very mode of communication on Facebook, which centers on curated profiles as the communicative addresses of communication (ibid.), introduces an always lingering suspicion as to whether and to what degree these avatars’ identity cues (name, profile picture, etc.) can be taken at face value.

Theoretically, this implies two strategies for identity work in social media communication: either to frame one’s interventions in an as-personal-as-possible way, or to be as stereotypical and therefore unambiguous as possible in one’s identity performance, for example by posting as a proud German, a loyal supporter of the AfD, or a concerned mother. As will become evident in the following, the conversations on the Facebook page of the AfD clearly gravitate toward the latter. Thus, while focus on the Facebook interactions excludes the material-situational setting of those doing the commenting and therefore prohibits me from inferring anything about the importance or function of the practice of commenting for their personal identity work, I am nevertheless able to observe commenting as a profile-identity performance. At least for the interactions under scrutiny, social media interactions in which users are present mainly as avatars, do not seem to reduce, but rather to enhance the importance of identity performance for understanding the interactional dynamic.

The interactions are, in other words, not so much exchanges on the marketplaces of ideas as Facebook would like to portray it (Maddox & Malson 2020), but rather arenas for individual users’ identity performances. This is an important finding and raises the questions: What kind(s) of identity are performed and validated in interactions around alternative facts? What does it take to be a competent participant in them?

Conspiratorial Communication: Collaborative Mistrust

To be a competent participant in interactions means to develop a feeling for the rules these interactions are played by (cf. Bourdieu 1977). One could add that this also means to develop a feeling for the problem processed in them. As it turns out, someone who treats the interaction on the Facebook pages of the AfD as a conversation about the corona virus (or immigration, or government policies, for instance) runs the risk of failing to contribute to processing its actual problem. A short example from a conversation from November 2020 about the German federal government’s measures against the Covid-19 pandemic suffices to illustrate this. This opens with a post by an administrator of the Facebook page:

It is unfair that restaurants have to close, even though they have invested so much in infection control measures. There is no scientific evidence for the risk of catching an infection there being higher than in public transportation, which can be used still. […] We therefore demand that restaurants can be open in December, given the high importance of the Christmas season for their business.3

This opening post is a classical political comment from an opposition party: A measure taken by the government is accused of being inconsistent and thereby unfair, and the post calls for the measure’s revocation. Instructive for understanding the peculiar dynamic of doing alternative facts are the comments reacting to this post; user Sabrina Sabine remarks: “I am frustrated about the Corona policies of the federal government. The lockdown is pointless and the only way to fight the virus is herd-immunity. We do not know the true reasons for the lockdown. But the truth will soon be revealed…” The inconsistency diagnosed by the original post is taken up in a very peculiar way – it is processed as a sign to be interpreted, something that points to “true reasons” not yet known but known to be there. This structure is replicated throughout the comments that respond to Sabrina Sabine. Kati Kramer for example replies: “I feel the same way. I feel duped by scientists raising and lowering the number of infections as they please…” Another user explains: “You are right. It is interesting that we didn’t hear of influenza this year.” And yet another user concludes:

Corona is a harmless sniffle; we don’t even need herd immunity. Unfortunately, too many sheeple [Schlafschafe4] believe in the government and play along. It is 19:33 o’clock again [A reference to the year the Nazis took power in Germany]. You are wrong about what the Corona virus is actually about and all head in the wrong direction.

None of these comments picks up the policy inconsistency of the original post. Instead, all of them validate Sabrina Sabine’s proposition to treat this inconsistency mainly as a sign of something lurking underneath. That there is something lurking underneath is undisputed – it is the unifying orienting scheme that allows all these comments to validate each other’s propositions, albeit them being at odds otherwise. After all: if the corona virus is characterized as too infectious to be contained by a lockdown, a harmless sniffle, the annual influenza or totally made up, as Kati Kramer seems to imply, can hardly be discarded as minor differences. Yet they are all acceptable in the framework of a supposedly deeper understanding. Only if there is implicit agreement to treat the observable discrepancy between official versions and alternative versions as a mere sign of a deeper, hidden truth, does the material incompatibility become marginal. Of course, one could argue that some of these inconsistencies are the result of hastiness: Maybe some commentators simply do not bother to read the comments before replying to them. Complementing the analysis with other methods of observation such as interviews could provide better insight into interpretation of the conversation itself. However, since some of the comments explicitly refer to each other by referring to their propositional content and/or by tagging commentators, hastiness is obviously not an exhaustive explanation for the apparent contradictions. And something else is striking: Even though it is only hinted at, all commentators seem to imply that they know what this deeper truth is – a sinister plan by the government to use the Covid-19 pandemic to implement dictatorial measures. This is most explicit in the last user’s metaphor of it being “19:33 o’clock” again – a claim that nobody contradicts throughout the conversation – and which obviously has a double function: given the broadly shared consensus on condemning Nazis in the German public, which is at least performatively shared even by many of those one would otherwise count as part of the extreme right, it puts one on the righteous side of history, and, via the use of the metaphor of the “clock” it furthermore induces a sense of high urgency: time is almost up.

The dynamic of the interaction – and really all other interactions around alternative facts interpreted in the study conducted with Hannah Trautmann – can thus be described as that of a collective investigation, of collaborative mistrust. This collective investigation has three basic building blocks: Its starting point is the framing of an event, a scandal, or an observable inconsistency as a riddle. This riddle is then processed by interpreting it as a clue pointing toward a deeper truth, which is treated as being implicitly known so that the actual work of investigating consists mainly of a process of collective plausibilization by collecting (rather than connecting) further clues pointing toward a conspiracy explaining the initial riddle.

This dynamic has important implications. The overarching identity performance that carries it is that of those who know versus the herd of “sheeple.” There even seems to be some kind of competition around who has the most radical take on the deeper truth. This becomes more evident when we include another strand of this very same Facebook conversation into the interpretation: that of the potential to act upon the proclaimed misdeeds of the government. A first participant announces that she will soon throw her mask away, a second one follows, proclaiming that she already threw it away in March, and the third one explains that she never wore one. There is a peculiar contrast between the small gesture of throwing away one’s mask (which is the only political action discussed in this regard) and the apocalyptic imaginary of a looming dictatorship investigated via the articulation of alternative facts. This contrast might furthermore be read as indicating the centrality of the collaborative enactment of mistrust and the corresponding identity performance of being one of those who know. Alternative facts have a twofold communicative function in this context: they provide building blocks for collaborative research, while serving as proof of being an especially daring investigator. Remarkably, due to the primary function of identity performance, the facticity of alternative facts is largely sidelined. The comments validate each other in the sense of an unstated purpose conceived of as collective and inclusive. They do this without explicitly challenging the alternative facts of the other comments, even when they materially contradict each other.

Throughout all the conversations I read, alternative facts were barely taken up in discussion other than as material on which to build. Even though their content quite often was incompatible with that of other alternative facts articulated in the same discussion, these contradictions were rarely made explicit. This is foremost a descriptive find, yet the very few instances in which such contradictions are pointed out provide some hunches as to why drawing attention to these contradictions might happen so rarely.

Policing Collaborative Mistrust: When Facticity Becomes Topical

Throughout the conversations reconstructed, the very few instances in which the factuality of alternative facts was called into question can be sorted into two types that differ both in the way the factuality is problematized and how this problem is processed in the conversation itself. The first way is to problematize their factual basis: one points out that the statement made (the alternative fact) is factually incorrect, because it is incompatible with what one oneself holds to be true. As we have seen, ignored, not-commented-on contradictions are the rule rather than an exception – the question of factual correctness of alternative facts might in this case pose a problem for observers, but it is not communicatively processed as a problem in the interaction itself.

If someone explicates the contradiction by challenging the factual correctness of an alternative fact, this is, similarly, treated as a minor disturbance and repaired rather quickly. In a conversation on left-wing-violence, for example, one user explicitly contradicts the claim that “the Antifa”5 receives funding from the governing parties in Germany. She posits that the Antifa does in fact not exist as an organizational entity, but is an umbrella term for numerous forms of engagement by various actors, making problematic the argument that the government supports violence by supporting the Antifa. This contradiction is responded to only by one user (the one that articulated the original claim) – who does not address its content, but instead doubles down on the problem of violence: “Antifas beat up people, set cars on fire, and put policemen in danger. So what’s the difference to Nazis?” With this renewed emphasis on what the commentator now posits as the actual problem of the conversation (the danger of left-wing violence), the contradiction is effectively defused – no one follows up on it, the conversation continues as if the contradiction was never raised.

This supports the finding that the content of alternative facts is of secondary importance, whereas their primary function is being vehicles of an identity performance: as someone partaking in the collective investigation against the government. Even if someone takes issue with the facticity of a claim (if it for example touches upon a subject which is important to them), this does not really pose a relevant problem for the overall conversation – it is simply ignored by the others.

The second way in which contradictions are made topical is different and involves problematizing the way in which alternative facts are articulated. What is problematized in these cases is not the propositional content of the alternative fact but its legitimacy as information to be included in the collective investigation – for example by doubting its source, or by denying that the information is of relevance for the problem discussed. An example of this occurs in a conversation about educational policies. The original post by the AfD comments on US educational policies and suggests that discussions on anti-racist curricula for math education in the state of Oregon were motivated by left-wing radicals trying to hide the weaker math performance of persons of color. The ensuing discussion quickly coalesces into a collective investigation into how the general demise of education in Germany should be understood as part of a plan to create an uneducated public in a de-industrialized country as a prerequisite for installing a dictatorship. One user objects to this as jumping to conclusions and suggests: “[…] It would be great to have a source for the US news story, so one could check it.” Other users quickly retaliate: “I think this is a trick question”; “This is no proposal, this is madness”; “And two times two is ten?” – and even when the original commentator tries to calm the discussion by proclaiming that he does not in fact support the proposed reform in Oregon but simply suggests to check the source first, the reply is a simple “if you ask stupid questions, you get stupid answers.” This pattern repeats itself in every instance of formal problematizations of alternative facts observed: The other participants react by either straight-out insulting the commentator (“stupid questions”), by ridiculing their objection (“two times two is ten?”) or by marking it as an attempt to de-rail the discussion (“trick question”). In other instances, the users who raise objections about facticity or relevance are called upon to “contribute constructively” and to find the missing links and examples themselves – or are simply accused of being hostile.

The two types of processing critique both bear witness to how conversations are immunized against the question of the facticity of alternative facts. This is made possible by the very dynamic of the collective investigation and the corresponding identity performance. Substantive contradictions, even if pointed out, can be discarded easily because the material content of the proposition is secondary to their function in the collective investigation. Whether Antifa really is an organizational actor receiving funding is secondary to the proposition agreed upon that the radical left is violent and dangerous and somehow in collusion with the government. This is more complicated in the case of criticisms with regards to the way in which alternative facts are brought up in discussions, because these criticisms problematize the mode of investigation itself. This becomes a threat to the identity of the user whose contribution to the conversation is thereby called into question, and to the collective identity of investigators, because it problematizes their collaborative performance of being one of those who know. This threat can be neutralized, however, by reinterpreting it through the differentiation between us and them – either the one voicing the criticism must prove her/his loyalty by repairing the damage, or (s)he is marked as an outsider whose intervention is considered hostile and therefore misleading.

It is highly plausible that these immunization strategies are supported by a central difference between face-to-face interactions and Facebook interactions – namely the latter’s mediation through avatars – in two regards. As mentioned at the beginning, the interactions show no indication of participants knowing each other personally or knowing each other from earlier online interactions. If we furthermore consider that many profile names do not appear to be actual, legal names, this indicates that those sharing alternative facts on the Facebook pages of the AfD do not have to personally fear sanctions from outside the immediate communicative context.6 This lowers the stakes of sharing information that might otherwise disqualify the commenter as a trustworthy partner in follow-up interaction. At the same time, the lingering insecurity over who is really behind the profile likely heightens the importance of profile identity performance within the immediate communicative context, allowing commenters to re-interpret criticism along the us–them distinction in order to shield oneself from criticism.


Reconstructing the doing of alternative facts as a practical, communicative accomplishment in concrete interactions sheds light on some aspects of alternative facts on social media platforms that have hitherto gone unnoticed. Perhaps most striking is that, while sidelining the question if those articulating or validating alternative facts do in fact believe in them was initially a methodological decision, the reconstruction has shown this sidelining to be a feature of the conversations itself. It is not the propositional content of alternative facts that makes the difference in conversation, but its performance style – the shared suspicion that there is more to the story is far more important than the actual content. The primary function of alternative facts is to support the identity performance of the commenter as someone who is one of those who knows and to contribute to the collective effort of investigating into the suspected government conspiracy behind the mainstream discourse. The actual content of alternative facts is treated almost with indifference by those reacting to them.

In the Facebook conversations analyzed, this practical indifference toward the facticity of alternative facts is documented in the stark difference between reactions toward two kinds of opposition. On the rare occasions when someone doubts the content of an alternative fact, this opposition is always ignored almost completely. If, however, the legitimacy of alternative facts as parts of the collective investigation is called into question, the one who raised this issue is quickly confronted with the decision to either withdraw her/his objection or to be cast as a hostile outsider. This can be understood as a way of processing the problem of facticity by connecting it to the problem of the always lingering insecurity from not knowing who the persons behind the avatars partaking in the discussion are. To demonstrate one’s trustworthiness is to expressively contribute to collaborative investigation into the government conspiracies. Calling into question the trustworthiness of contributions to these investigations thus is interpreted as a hostile act by an outsider. By doubling down on the collaborative identity performance of the daring investigator, one is thereby able to shield one’s own comments against criticism.

In other words, the heated philosophical debate over whether alternative facts are a construction of truth or flat-out lies (or errors) is, in practice, rendered less relevant by a communicative practice which limits itself to un-truthing “original facts,” while effectively de-problematizing the question of the facticity of the alternative facts proper. This has an interesting effect: contrary to what the filter-bubble argument would lead us to expect, alternative facts do not flourish because the original facts are unknown in those conversations or because people were not exposed to divergent sources of information. Quite the contrary: Since what makes a proposition in these conversations plausible is not the coherence of their informational content with other propositions in the conversation, but rather its emotive coherence with the shared assumption that the mainstream discourse is (intentionally) misleading, a competent participation in the collaborative effort to “debunk” the discourse of the “sheeple” requires primarily acknowledgement of this discourse. In the same manner, epistemological uncertainty is not the futile breeding ground of alternative facts that it seems to be – quite the opposite. Since the foundation of the collective identity performance is the shared expression of fundamental opposition to mainstream discourse, alternative facting flourishes when mainstream social knowledge can be assumed to be obvious and widely known.

The sample from which I drew my examples was highly selective. Further comparative research would need to investigate whether the patterns found in these conversations can be further generalized to alternative facting on other social media platforms such as Twitter/X or Telegram, or even the comment sections on the webpages of classical news organizations or other political groups. However, a cursory reading of discussions on Trump-supporting Facebook groups indicates that neither the topics discussed nor the way they are discussed are limited to the case of German right-wing social media discourse. For now, even the question of whether the reconstructed patterns are specific to digital interaction or if homologous patterns can also be identified in offline interaction on alternative facts must remain open.

Studying the function of alternative facts in other contexts and especially how they travel between those contexts could also connect the findings of this article to those of anthropological studies on legends, fake news and conspiracy theories. While it turns out that in vernacular political interactions on right-wing Facebook pages, alternative facts are mostly used as building blocks in the collaborative enactment of mistrust against the perceived mainstream discourse, they are undoubtedly actively constructed as theories in other communicative arenas.


  1. The study was funded by the Otto Brenner Foundation and conducted between November 2020 and June 2021. For German-speaking readers, the working paper of this project (Trautmann & Kumkar 2021) offers a more comprehensive insight into the material, which can only be cited selectively in this article. Also based on this study is a monography (Kumkar 2022), which offers a more theoretical discussion of the function of alternative facts in social media, mass media, and political communication. [^]
  2. The term science wars refers to a series of heated debates amongst academics (mainly in the US) in the 1990s on the authority of science in making truth-claims. One very prominent example is the discussion around the so-called Sokal hoax, in which the physicist Alan Sokal had handed in a nonsense-paper named “Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity” (Sokal 1994) to the cultural studies journal Social Text – and got it published (cf. Lingua Franca 2000). [^]
  3. All quotes are the author’s translations of the original German conversations which were paraphrased for privacy concerns in the research project. [^]
  4. Like the English “sheeple”, the German neologism “Schlafschafe” is firmly rooted in conspiracy theory discourse, referring to the herd of people mindlessly trotting along with the others. While the English term is a combination of sheep and people (Cambridge Dictionary n.d.), the German Schlafschafe is a combination of Schlaf (sleep) and Schaf (sheep). [^]
  5. “Antifa” is an umbrella term for diverse groups, organizations and initiatives from the political left that fight against radical right-wing parties and ideologies. While the term has its historical roots in the “Antifaschistische Aktion” of the German Communist Party of the interwar period, today it is taken up in diverse political contexts around the globe. Comparable to the US, where then-President Donald Trump infamously called for designating it as a terrorist organization, ignoring both that the US does not have a domestic-terrorism law and that Antifa is not, in fact, an organization in any meaningful sense of the concept (Bogel-Burroughs & Garcia 2020), the term Antifa is often used by the German political right to refer to an organization imagined to be both powerful and violent. [^]
  6. This might be different on other social media platforms such as Twitter/X, where the common practice to re-tweet a comment and thereby present it to a broader public, makes communicative sanctions for the profile much more probable. [^]


Bakir, Vian & Andrew McStay 2018: Fake News and the Economy of Emotions. Digital Journalism 6(2): 154–175. DOI:

Bakshy, Eytan, Solomon Messing & Lada A. Adamic 2015: Exposure to Ideologically Diverse News and Opinion on Facebook. Science 348(6239): 1130–1132. DOI:

Baraldi, Claudio 2015: Interaktion [Interaction]. In: Claudio Baraldi, Giancarlo Corsi & Elena Esposito (eds.), GLU: Glossar zu Niklas Luhmanns Theorie sozialer Systeme [Glossary for Niklas Luhmann’s theory of social systems]. Frankfurt a.M.: Suhrkamp, 82–85.

Bareither, Christoph 2020: Media of the Everyday: The Contribution of European Ethnology to the Field of Media and Digital Anthropology. Journal for European Ethnology and Cultural Analysis (JEECA) 4(1): 5–27.

Beisch, Natalie & Carmen Schäfer 2020: Ergebnisse der ARD/ZDF-Onlinestudie: Internetnutzung mit großer Dynamik: Medien, Kommunikation, Social Media [Results of the ARD/ZDF-online survey: internet use displaying strong dynamics]. Media Perspektiven 9/2020,

Bogel-Burroughs, Nicholas & Sandra E. Garcia 2020: What Is Antifa, the Movement Trump Wants to Declare a Terror Group? The New York Times, September 28, 2020, (accessed April 26, 2022).

Bohnsack, Ralf 1989: Generation, Milieu und Geschlecht [Generation, milieu, gender]. Opladen: Leske + Budrich. DOI:

Bohnsack, Ralf 2004: Group Discussion and Focus Group. In: Uwe Flick, Ernst von Kardoff & Ines Steinke (eds.), A Companion to Qualitative Research. London, Thousand Oaks & New Delhi: Sage, 214–221.

Bohnsack, Ralf 2010: Rekonstruktive Sozialforschung [Reconstructive social research]. Opladen & Farmington Hills: Barbara Budrich.

Bohnsack, Ralf, Aglaja Przyborski & Burkhard Schäffer (eds.) 2006: Das Gruppendiskussionsverfahren in der Forschungspraxis [The method of group discussions in practical research]. Opladen & Toronto: Barbara Budrich.

Bourdieu, Pierre 1977: Outline of a Theory of Practice. Cambridge: Cambridge University Press. DOI:

Buchmayr, Florian 2019: Im Feld der Verschwörungstheorien – Interaktionsregeln und kollektive Identitäten einer verschwörungstheoretischen Bewegung [In the field of conspiracy theories – rules of interaction and collective identities in a conspiracist movement]. Österreichische Zeitschrift für Soziologie [Austrian journal of sociology] 44(4): 369–386. DOI:

Cambridge Dictionary n.d.: Sheeple. In: Cambridge Advanced Learner’s Dictionary & Thesaurus, (accessed April 11, 2022).

Carlson, Matt 2020: Fake News as an Informational Moral Panic: The Symbolic Deviancy of Social Media during the 2016 US Presidential Election. Information, Communication & Society 23(3): 374–388. DOI:

Couldry, Nick 2004: Theorising Media as Practice. Social Semiotics 14(2): 115–132. DOI:

Duffy, Andrew, Edson Tandoc & Rich Ling 2020: Too Good to Be True, Too Good Not to Share: The Social Utility of Fake News. Information, Communication & Society 23(13): 1965‒1979. DOI:

Farina, Matteo 2018: Facebook and Conversation Analysis: The Structure and Organization of Comment Threads. London: Bloomsbury.

Flaxman, Seth, Sharad Goel & Justin M. Rao 2016: Filter Bubbles, Echo Chambers, and Online News Consumption. Public Opinion Quarterly 80(1): 298–320. DOI:

Garfinkel, Harold 1959: Common Sense Knowledge of Social Structures: A paper distributed at the session on the Sociology of Knowledge, Fourth World Congress of Sociology, Stresa, Italy, September 12, 1959. DOI:

Garfinkel, Harold & Harvey Sacks 1970: On Formal Structures of Practical Action. In: John C. McKinney & Edward A Tiryakian (eds.), Theoretical Sociology: Perspectives and Developments. New York: Appleton-Century-Crafts, 338‒366.

Goffman, Erving 1983: The Interaction Order: American Sociological Association, 1982 Presidential Address. American Sociological Review 48(1): 1–17. DOI:

Guess, Andrew, Neil Malhotra, Jennifer Pan, Pablo Barbera, Hunt Allcott, Taylor Brown, Adrian Crespo Tenorio, Drew Dimmery, Deen Freelon, Matthew Gentzkow, Sandra Gonzales-Bailon, Edward Kennedy, Young Mie Kim, David Lazer, Devra Moehler, Brendan Nyhan, Carlos Rivera, Jaime Settle, Daniel Robert Thomas (…) & Joshua A. Tucker 2023: How Do Social Media Feed Algorithms Affect Attitudes and Behavior in an Election Campaign? Science (381): 398‒404. DOI:

Habermas, Jürgen 2021: Überlegungen und Hypothesen zu einem erneuten Strukturwandel der politischen Öffentlichkeit [Thoughts and hypotheses on a new structural transformation of the public sphere]. In: Martin Seeliger & Sebastian Sevignani (eds.), Ein neuer Strukturwandel der Öffentlichkeit? [A new structural transformation of the public sphere?] Wiesbaden: Nomos, 470–500. DOI:

Hall, Stuart 1975: Introduction. In: Anthony Charles H. Smith (ed.), Paper Voices: The Popular Press and Social Change 1935‒1965. London: Chatton & Windus, 11‒24.

Hogan, Bernie 2010: The Presentation of Self in the Age of Social Media: Distinguishing Performances and Exhibitions Online. Bulletin of Science, Technology & Society 30(6): 377–386. DOI:

Kumkar, Nils C. 2018: The Tea Party, Occupy Wall Street, and the Great Recession. Basingstoke & New York: Palgrave Macmillan. DOI:

Kumkar, Nils C. 2022: Alternative Fakten: Zur Praxis der kommunikativen Erkenntnisverweigerung [Alternative facts: the practice of communicative refusal to know]. Berlin: Suhrkamp.

Ledwich, Mark & Anna Zaitsev 2019: Algorithmic Extremism: Examining YouTube’s Rabbit Hole of Radicalization. ArXiv:1912.11211 [Cs], (accessed April 26, 2022).

Lingua Franca 2000: The Sokal Hoax: The Sham that Shook the Academy. Lincoln: University of Nebraska Press.

Maddox, Jessica & Jennifer Malson 2020: Guidelines without Lines, Communities without Borders: The Marketplace of Ideas and Digital Manifest Destiny in Social Media Platform Policies. Social Media + Society 6(2): 1–10. DOI:

Mannheim, Karl 1997: A Sociological Theory of Culture and its Knowability (Conjunctive and Communicative Thinking). In: Structures of Thinking: Mannheim’s Collected Works Vol. X. New York: Routledge, 141‒288.

Meredith, Joanne 2019: Conversation Analysis and Online Interaction. Research on Language and Social Interaction 52(3): 241–256. DOI:

Nocun, Katharina & Pia Lamberty 2020: Fake Facts: Wie Verschwörungstheorien unser Denken bestimmen [Fake facts: how conspiracy theories shape our thinking]. Köln: Quadriga.

O’Callaghan, Derek, Derek Greene, Maura Conway, Joe Carthy & Pádraig Cunningham 2015: Down the (White) Rabbit Hole: The Extreme Right and Online Recommender Systems. Social Science Computer Review 33(4): 459–478. DOI:

Oxford Dictionary 2016: Word of the Year 2016, (accessed June 5, 2021).

Pennycook, Gordon & David G. Rand 2020: Who Falls for Fake News? The Roles of Bullshit Receptivity, Overclaiming, Familiarity, and Analytic Thinking. Journal of Personality 88(2): 185–200. DOI:

Pink, Sarah, Heather Horst, John Postill, Larissa Hjorth, Tanja Lewis & Jo Tacchi 2016: Digital Ethnography: Principles and Practice. Los Angeles: Sage.

Przyborski, Aglaja 2004: Gesprächsanalyse und dokumentarische Methode: Qualitative Auswertung von Gesprächen, Gruppendiskussionen und anderen Diskursen [Conversation analysis and documentary method: qualitative interpretation of conversations, group discussions, and other discourses]. Wiesbaden: Springer VS.

Reeves, Stuart & Barry Brown 2016: Embeddedness and Sequentiality in Social Media. In: Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work & Social Computing. New York: Association for Computing Machinery, 1052–1064. DOI:

Rutjens, Bastiaan T., Sander van der Linden, Romy van der Lee & Natalia Zarzeczna 2021: A Group Processes Approach to Antiscience Beliefs and Endorsement of “Alternative Facts”. Group Processes & Intergroup Relations 24(4): 513–517. DOI:

Schimank, Uwe 2016: Handeln und Strukturen: Einführung in die akteurtheoretische Soziologie [Action and structure: introduction to actor-centered sociology]. Weinheim, Basel: Beltz Juventa.

Schwarz, Ori 2021: Sociological Theory for Digital Society: The Codes that Bind Us Together. Cambridge & Medford: Polity Press.

Sokal, Alan 1994: Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity, (originally published in: Social Text 46‒47: 217‒252, accessed November 27, 2023). DOI:

Spohr, Dominic 2017: Fake News and Ideological Polarization: Filter Bubbles and Selective Exposure on Social Media. Business Information Review 34(3): 150–160. DOI:

Staab, Philipp & Thorsten Thiel 2021: Privatisierung ohne Privatismus: Soziale Medien im Digitalen Strukturwandel der Öffentlichkeit [Privatization without privacy: social media in the digital structural transformation of the public sphere]. In: Martin Seeliger & Sebastian Sevignani (eds.), Ein neuer Strukturwandel der Öffentlichkeit? [A new structural transformation of the public sphere?] Wiesbaden: Nomos, 275–297. DOI:

Trautmann, Hannah & Nils C. Kumkar 2021: Alternative Fakten im Gespräch: AfD-Diskussionen auf Facebook [Alternative facts in dialogue: AfD-discussions on Facebook]. OBS Working Paper, (accessed September 25, 2023).

Vosoughi, Soroush, Deb Roy & Sinan Aral 2018: The Spread of True and False News Online. Science 359(6380): 1146–1151. DOI:

Nils C. Kumkar, Dr. Phil., is a junior researcher (post-doc) sociologist at the SOCIUM, Research Center on Inequality and Social Policy at the University of Bremen. His research interests are qualitative methods, sociological theory, sociology of critique and protest, and social inequality. He recently (2021) co-authored a chapter on qualitative methods in Betina Hollstein et al. (eds.): Soziologie – Sociology in the German-Speaking World (Berlin: De Gruyter).