Conspiratorial Hermeneutics via Words & Images

Facilitators: Tommy Shane, Marc Tuters, Tom Willaert

General Introduction

While conspiracy theories have long been understood as irrational narratives produced by extremists in the margins of political and social life, a body of humanities scholarship considers them as critical responses to the complexities and uncertainties of (post)modern life (Knight 2000). While conspiracy theory is often imagined to be a right-wing preoccupation, with the pandemic we have seen the emergence of new "diagonal movements" that cut across traditional left/right distinctions sharing the conviction that all power is conspiracy (Slobodian & Callison 2021).

In this summer school project, we set out to make an empirical contribution to the study of online conspiracism by examining conspiratorial narratives on twitter and instagram. We thereby combined digital methods and close reading techniques, bridging gaps between big data analytics and discourse analysis, in an approach that, following Gerbaudo (2016) could be called ‘data hermeneutics’:

Based on this approach, the project explored and tested two hypotheses central to contemporary theorizing around conspiracism in the humanities. First, this concerns the hypothesis of a ‘diagonalism’ in which conspiracy theories and the concepts they operationalize cut across traditional dichotomies between the political left and right. In order to test this hypothesis, we turn to a large corpus of conspiratorial tweets, and we propose a novel methodology in which we explore the diverging or converging usages of curated lists of keywords from the domains of critical theory, new age, epistemology (words related to truth and knowledge), and social justice terminology. In order to infer and plot the different meanings of these terms, we investigate them on the basis of word trees and plot them on a version of the ‘political compass’.

The second, related hypothesis that we test in the project, is that of the potential ‘convergence’ of conspiracy theories over the course of the coronavirus pandemic. While fringe conspiracy theories have long been an integral feature of participatory web culture, our research found that they have proliferated during the coronavirus pandemic (DMI 2020, Stolper 2020). Popular interpretations of this proliferation have suggested that previously divergent conspiracy theories have converged into a "conspiracy singularity" during the pandemic (Merlan 2020). Having found anecdotal support for this convergence hypothesis in the 2020 summer school and 2021 winter school on Instagram, the second part of this project performed a quali-quantitative analysis of how hashtags, named entities and images in a large Instagram dataset appear to bring together different conspiracy theories over time through processes of ‘convergence’ and ‘integration’.

Week 1: Examining Words and Phrases on Twitter

Team members week 1: Stefan Baak, Migle Bareikyte, Minna Horowitz, Ben Kapadia, Johanne Kalsaas, Joris Legerstee, Johanna Muhrbeck, Marcantonio Bracale Syrnikov, Mathias-Felipe de-Lima-Santos, Cesar Santos Silva, Maurice Dharampal

Link to Recipe week 1

Link to final presentation week 1

Background: verbal boundary objects

By asking what verbal tools conspiracy communities use and where they got them from, it is possible to answer the question: if conspiracy theorising is a form of social critique, what critical tools do they use, how do they use them, and what kind of social critique do they make possible?

Like conspiracy theories themselves, words can connect disparate domains and communities. For example, the word ‘gaslight’ might discursively link domains such as social justice campaigning with conspiracy theorising, and the ‘woke’ left with the libertarian right. In cases like this, words are boundary objects: “objects which are both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to maintain a common identity across sites” (Star and Griesemer, 1989, p. 393).” Verbal boundary objects, such as ‘gaslight’, tend to originate from one domain and community, and get co-opted by another. This can be a powerful disinformation tactic, illustrated by the success of Donald Trump’s co-optation of the phrases “fake news” and “big lie” for his own purposes, and confusing the discursive clarity of the terms and thus its ability to critique him. By attending to how a single word or phrase moves between communities, it is possible to interrogate the tactics used by conspiracy communities and the kinds of social critique they afford.

Data

We worked with a corpus of 15 million Twitter posts related to conspiracy theory published during the period of the pandemic. The original corpus was collected using a long list of conspiracy theory hashtags and keywords (e.g. #BillGatesisEvil). Together with domain experts a list of words was developed in four categories (see below). Using NLP (word embeddings), other words were then identified that were deemed to be used similarly to those in the expert list. The first and second “snowballed” lists of words were then used to once again scrape Twitter (when those words co-occurred with covid OR coronavirus OR covid19).

We have identified keywords in four high level themes:

1. Critical theory (eg ‘biopolitics’)
2. Social justice (eg ‘gaslight’)
3. New Age (eg ‘awakening’)
4. Epistemic (eg ‘truth’)

Methodology

The methodology of the first week’s project is a combination of qualitative and quantitative methods, following the instructions given by the tutors of the Digital Summer School 2021 as provided in a detailed recipe for the applied use of the given digital tools. By using 4Cat, it was possible to select epistemic keywords and their related snowballed terms related to four high level themes (Social Justice, Critical Theory, New Age, and Epistemic) from a greater corpus of 15M Tweets to create a dataset. Further, this dataset was ranked to create a top ten of the most retweeted posts. A short analysis of that top ten tweets could shed light on the dominant discourse surrounding an epistemic keyword and its use in certain communities. By using an ‘interactive word tree tool’, zoom-ins can be conducted for a closer reading to analyse the text for unusual and weird posts. Two forms of time-series analysis can be applied. Firstly, an analysis concerning bi-grams over time makes it possible to detect how words in posts co-occur over time. A second type of time series analysis is the analysis of authors over time, which is useful to find out how widely adopted a term might be used and how many users are posting a certain term. Further, two different types of network analysis can be conducted. First, a Co-tag network can be conducted to draw edges between hashtags with Gephi. Through the use of Gephi, it is possible to create a panoramic view of what people think they are talking about while using a certain term in their posts. A second type of network analysis that can be conducted is a Bipartite Author-tag network, which can indicate how different communities/ tribes are using a certain term and how the use of a certain term can create a common ground as well as diagonalism. Every form of clusters displayed in connection with certain terms can demonstrate a network of ‘idea entrepreneurs’. Finally, every epistemic keyword/ concept used can be positioned in a ‘political compass’. The value of - 10 on the x-axis indicates an extreme left position in the political spectrum and a value of + 10 on the same axis indicates an extreme right position. On the y-axis, a value of - 10 indicates an extreme liberitarian position and a value of + 10 indicates an extreme authoritarian position.

Findings

Critical Theory

A burgeoning effect of diagonalism is the diagonalization of culture at all levels. Discussing climate change denialism, Latour (2004) pointed out how conspiratorial actors and critical theorists may have something in common: they are “similar in the structure of the explanation, in the first movement of disbelief and, then, in the wheeling of causal explanations coming out of the deep dark below”. Conspiracy thinking often co-opts arguments from media-literacy (Starbird 2017). This implies a transition from emotivistic populism to a newly-found intellectualism. Parallel but not reducible to hash-tag hijacking (whose subject and target is the a networked affective public), diagonalism has a propensity towards a ‘theory-hijacking’ from emancipatory politics: diagonal conspiratorial actors represent themselves as a colonized and dominated subjectivity, thus seeking a new power-knowledge situation. As in post-truth’s case (see previous section), these communities don’t represent themselves as subjects, not objects, of critique: they have no role in post-truth; on the contrary, they are trying to critique it. Theory-hijacking is the epistemic side of hashtag-hijacking, and it is not enacted upon matters of concern but on the schemas of their interpretation.As we can see in the diagonalism compass below (Fig. 1) the four critical keywords we analysed present a marked diagonalism, cutting across the political compass diagram. On one hand, these keywords are contested by incompatible political orientation; on the other they are mobilized by diagonal movements that seem to destabilize received analytical categories such as ‘left-’ and right-wing’ politics.

Fig. 1 Political compass of Critical Theory terminology

Zooming in on a specific critical theory, namely Hegelian Dialectics, it can be argued that the democratisation or socialisation (cf. Birchall) of social or critical theory allows for its appearance in these environments. While the entire ‘critical theory’ data set within the corpora is not very substantial compared to others such as New Age, there is a conclusion to be drawn. Nearly all the closely read tweets illustrate that it’s become an instrument which is simplified and should adhere to the conspiracists confirmation bias. Images below illustrate how Hegelian Dialectics, something the German Idealist himself conceptualised differently in various of his works, is simplified to ‘thesis, antithesis, synthesis’ or ‘problem, reaction, solution’. It is synonymous with conspiratorial concepts such as mind control or brainwashing. Yet, in a more optimistic reading, what the conspiratorial subject seems to be aiming at is providing critique on power structures. The word tree below illustrates these two points: hermeneutic liberty (in red) and adoption, application of confirmation bias (in orange).

The hermeneutic liberty of Hegelian Dialectic concept is prevalent in the conspiracy spaces, mostly incorporated in the right-wing narrative. Yet, it does not necessarily have to be, indicating the notion of diagonalism this overall research is concerned with. This is also noticeable in the political compass above (Fig.1).

Social Justice

Red Pill

The term “red pill” is described in this article through a chronological outline of its emergence and demise. Nowadays it is claimed to mean nothing specific at all/all those things mentioned above. But this vagueness/deradicalisation of the term in a way also fits well into the narrative of “diagonalism”, where the term originally developed in The Matrix movie and corresponding to a more progressive worldview gets co-opted in such a way that its nowadays speaks to neither rightnor left.

Diagonalism is to a certain extent applicable, but also a co-optation of other fairly innocent terms or symbols has a history, e.g. in Germany, where 88 is a number used by the Nazis, for example. In addition the word itself has developed into a proverb (redpilled), in this context it is nothing more or less than a marker of interest or convincement.

Black & White privilege

Meaning and interpretation is guided by for both sides as lived experiences of audiences in an attempt for a claim on ‘’truth’’ where denial is central to (white privilege) and resignification to (black privilege). In terms of diagonalism during discourse the term itself has been interpreted and appropriated by left and right wing to add to their cause and paradigm, reasserting their ‘’truth’’ and opening up the discussion and battle for allocation.

In short, the main takeaways and findings evolve around the basis of qualitative analysis of tweets does not often suffice to determine the political leaning of the author. This is due to the, Appropriation of ambiguity through alt-right capitalisation, but also: critique of ambiguity through historicization. Also to conclude the effect of audiences trolling the whole discourse and capitalisation of terms themselves.

Demonize

Demonizing is a tool for ‘othering’, a way to create differences between somebody's own group and the opposite group. Fake news is used as a central argument by both sides of the polarized political spectrum. Any side claims to know the truth and tries to expose the lies of the other side.

The leftist narrative claims that the right wing is spreading fake news to reach certain political goals by distorting reality with unscientific polemics. The right wing narrative claims that liberals are socialist goons that create nothing of value for the society, the left only wants to destroy and ‘demonize’ political figures on the right. By using that type of narrative, the right wing is using historical anti-communist sentiments to reach the hearts of their audience. Further, the demonizing narrative of the right wing is often based on nationalism expressed in nostalgia and conservatism of a world long lost, sometimes connected to a strong anti-muslim propaganda. The most obvious difference between right and left wing is that the right wing demonizes the left with constructed lies and fake news, while the left wing is trying to demonize the right wing by pointing out the right wing's systematic use of lies and fake news. Instead of creating a common ground and ‘diagonalism’, the concept of demonizing is a tool to create strong distinction between tribes. That is possible to observe, because twitter is a platform that is mostly used by users from the USA (as seen that president trump was the most prominent and followed twitter user). It is possible that the epistemic keyword ‘demonize’ creates a certain diagonalism in different spatial contexts, as the definition of diagonalism as a concept was inspired by the strange dynamics of the protesting movement in Germany since 2015.

Woke

The term ‘woke’ was added to Oxford dictionary in 2017, according to which it refers to:

According to the article linked above, this meaning of the term emerged in the 1960s:

However, the top 10 retweeted Tweets in a dataset based on the keywords ‘woke’ and ‘awake’ show that conspiracy theorist use the term either in way that is completely removed from its origins in Afro-American culture (suggesting that you’re ‘work’ when you believe in conspiracies) or as as an ‘insult’ to discredited people who actually are ‘woke’ in its original meaning (calling it ‘woke virtue signaling’). While most of the data points to right-wing extremist conspiracy (QAnon is prominent), there are instances where the term is meant in its original form. Usually, it is used to blame others (individuals or companies) to ‘pretend’ to be woke. Example:

In the word-tree analysis, 5G conspiracy theories are quite prominent:

Bill Gates is also featured prominently because he is called ‘woke’ (in a dismissive way) or people should become ‘woke’ to the various conspiracies related to him:

Co-hashtag analysis confirmed findings from word-tree exploration, as only one cluster around QANON emerged:

The term ‘woke’ appears to be an example of a concept that originally emerged on the left but has now almost completely been taken over by the right. Instead of being ‘woke’ towards social injustices, conspiracy theorists talk about being ‘woke’ towards various conspiracies around QAnon. If the term is invoked in ways closer to its original meaning it has a negative connotation - a form of left-wing ‘virtue signaling’. If the term is used on the left, it seems to be negative too: as a way to criticize individuals or organizations that claim to be ‘woke’ but fail to do justice to it. This makes ‘woke’ an example for a term whose meaning closed down because it was diluted due to how right-wing conspiracy theorists appropriated the term.

New Age

The “New Age” movement is a range of para-religious beliefs and practices mixing oriental, ancient, and folk cultures which came to be in the Western world in the late 1960s from the hippie movement (Brown, 1997). Although analytically often considered to be religious, those involved in it typically prefer the designation of spiritual, mystical, mind, body, spirit and rarely use the term New Age (Satin, 1978). For this reason, many researchers contest calling it the “New Age movement,” as it is better seen as a broad set of cultural practices and beliefs (Kemp & James, 2007). Scholarly definitions of “New Age” mostly agree it does not have a specific core of beliefs, and is highly eclectic, with ancient peoples wisdom, natural elements as mystical entities, and cosmological interconnectedness, as some of the most common (Robertson, 2016). More recently, due to its boom online, the term has been associated with conspiracy theories, which resulted in a new term “conspirituality.” Coined in 2011 by Ward & Voas (2011), it describes people who are attracted to conspiracy theories about the spirituality space. Our analysis looked at several words and their variations that could be associated with “New Age,” such as vibrate, wake up, ascension, and free think. These words gave broadly very similar results, with the most prominent themes being awareness of political and corporate power, the makings of secret organizations like NWO, QAnon, and the Deep State. Our analysis also found a series of conspiracy theories associating the New Age to the COVID-19 pandemic. For example, in the dataset, the concept of vibrational (vibration and vibrate were also put under this category) is mostly used to refer to a person's body connection with its surroundings and somehow is disturbed by changes in the political realm. The most prominent claims are made in conjunction with the idea of “pandemic”—how covid pandemic is planned by secret organizations, notably by changing the ‘frequencies’ of this said vibration.Fig. 2 illustrates the word three of these tweets.

Fig. 2

Tweets about the term “vibrational” show the influence on a person's body connection with its surroundings due to unavoided frequencies and energies. In the case of “ascension,” it was possible to see a spike of the use in November 2020, following the US elections. Most of the tweets were tapping into the evolution concept of dimension to achieve superiority. These tweets were very associated with the principles of yoga that were commonly linked to the left (Klein, 2020; Schofield, 2021). However, these terms were used as a ​​coping mechanism to deal with an uncertain world as well as to deny the COVID-19 pandemic and to reject vaccines. One of the images found on these tweets that highlighted our attention described that this experience that we are facing is a path toward a fifth dimension (5D), which will be a better world (see Fig. 3). According to it, our current dimension (called 3D) is full of false flags, fake news, pollutants, etc and the “free think” will only be achieved in the fifth dimension when people will be enlightened and have their “third eye opened”. In other tweets, this moment is called the Great Awakening. Our analysis understands that these messages are hard for social media platforms to detect as they use blatantly conspiracy language and spiritual wellness messaging, requiring human intelligence to distinguish.

Fig. 3

The image illustrates the most common use of the word “ascension” as a path toward a fifth dimension (5D). On the political compass, the positions of these tweets were mainly on the libertarian spectrum, but with some tendency to the left political space. This is an interesting finding as these conspiracy theories were about idealistic worlds, very linked to yoga and “flower power,” which was historically closer to the “left,” but it has recently been reported as blended with anti-vaccination fantasies and outright coronavirus denialism disseminated by the alt-right groups (Klein, 2020; Schofield, 2021).

Fig. 4 The political compass of New Age

Generally speaking, our analysis found that people express less interest in politics, and more a desire to escape (to “ascend” or “wake”) and to live beyond the mundane, which they consider being spoiled by the interference of secret organizations (NWO, Deep State, etc.), people with bad energies, that are asleep (the sheeple), or with altered frequencies.

Epistemic keywords

The fourth and final sub-project focused on epistemic keywords. Against the background of a deepening ‘epistemic crisis’ (Dahlgren 2018; Zuboff 2020) where notions of knowledge and truth are increasingly contested, our assumption is that such keywords serve as a rich source for exploring the hermeneutics of online conspiracism. A small number of keywords were selected from the previously composed keyword list to explore to what extent notions of ‘diagonalism’ (Slobodian & Callison 2021) may provide fruitful perspectives for the study of conspiracy theories in online spaces. For the present research, we focused on a set of interconnected keywords serving as a heuristic for the epistemic condition (i.e. crisis) conspiracy communities perceive themselves to be in:

• Epistemic environment: ‘deception’
• Epistemic quality: ‘hidden'
• Epistemic action: ‘uncover '
• Epistemic goal: ‘truth’

Tweets in our dataset reflected users’ experience of living in a state of deception, where the truth is hidden and needs to be actively uncovered.Our epistemic sub-corpora were structured as follows:

• deception | deceptions: 9509 posts with 35738 retweets.
• uncover | uncovers | uncovered | uncovering: 10639 posts with 10639 retweets.
• hidden: 22171 posts with 83736 retweets
• truth: 177951 posts with 996017 retweets

Deception

A distant reading of deception-discourse on Twitter reveals that the term is intrinsically linked to notions of epistemic crisis. Words such as ‘lies’, ‘fake (news)’, ‘conspiracy’ and ‘‘manipulation’ dominate the corpus. The concept appears integral to the prolific conspiracy narratives of QAnon, New World Order, Obamagate, deep state, and ‘plandemic’. These connections are especially apparent when looking at the most commonly used hashtags in the corpus.

Fig. 5 Wordtree for ‘deception’

An analysis of the most common bi-grams over time shows that ‘deception’ also connects to a more recent conspiracy narrative, that of the Great Reset. Labelled a “conspiracy smoothie” for its ability to mix and blend a near infinite number of critiques into one (Klein, 2020), the powerful Davos elites emerge as central actors in the deception.

Fig. 6 Rank flow diagram for ‘deception’

While ‘deception’ undoubtedly seems to serve as a verbal ‘integrator’ for conspiratorial discourse, the “true” nature of this conspiracy remains highly contested. Close readings reveal that while tweets might mention conspiracy narratives they do not necessarily support them. Rather, users from different sides of the political spectrum contest each others’ notion of what the great deception -- a central phrase in the corpus --truly is. The core conflict centers on whether the deception consists of the conspiracy (e.g. widespread election fraud), or the conspiracy theory (e.g. QAnon).

Fig. 7 Conflicting tweets on the nature of “the great deception”.

A closer glance at the corpus’ top tweets support the notion that the deception-concept is remarkably complex and fluid. Appropriated by a diverging set of actors with a great deal of ideological, social and cultural diversity, the idea of a great deception resonates with political movements not only across the political spectrum, but the world: The American democratic, republican and alt-right movements (#1, #3, #5, #8) as well as British progressives and conservatives (#2, #4, #10) appear alongside a Filippino human rights activist (#6), a Thai influencer (#7) and a Kenyan opposition politician (#9). More than a cohesive conspiratorial concept, then, the deception-term functions as a polyphonous rallying cry and discursive weapon in the fight against (perceived) epistemic injustice.

Hidden

The overview of the keyword ‘hidden’ shows that while it is linked to a great variety of conspiratorial themes, three distinct stands of discourse prevail. First, the keyword is strongly linked to Covid-19. This is to a great extent due to the viral YouTube video Plandemic: The Hidden Agenda Behind Covid-19, the impact of which on conspiracy tweets is clearly visible in Fig. 8. (see also the keyword Deception). The video, released on May 4 2020, had within a week been viewed more than eight million times on YouTube, Facebook, Twitter and Instagram (Frenkel et al. 2020).

Fig. 8 Wordtree for ‘hidden’

Even so, while the video dominates the corpus on ‘hidden’, Covid-19 is featured only in two of the ten most retweeted tweets. The rest of the most popular tweets highlight the second discursive trait characterizing the corpus, and a feature typical to conspiracies in general: the political and others of the ‘epistemic elite’ – i.e., those with privilege regarding knowledge and power to disseminate it – purposely hiding the truth. While most top tweets, whether originating from the U.S., India, Nigeria or China, address political and government secrecy and corruption, including Obamagate (see also the keyword Uncover), this notion of distrusting not only the political but also the cultural order has prevailed in the Western culture for half-a-century (e.g., Fenster 1999). Related is the recurring phrase in the corpus that what is hidden is in fact right in front of us, all around us if we only knew how to look, ‘hidden in plain sight’. The above is exemplified in a top tweet (#5, Fig. 9) that mocks conspiracies about Hollywood and other cultural elites who are allegedly using the drug adrenochrome for satanic rites and equivalent. The tweet combines the conspiracy discourse with the humorous meme that the cartoon Simpsons has the canny ability to predict the future:

Fig. 9 Tweet about adrenochrome

Finally, the keyword ‘hidden’ underlines a curious discursive battle over epistemic authority. Unsurprisingly, what is ‘hidden’ is kept from ‘us’, ‘the people’ by those in the know. While some conspiracy theories seek to elevate and legitimize particular groups as morally sound and superior (see account e.g., in Douglas 2021), the corpus shows a strong populist collectivist discourse that positions ordinary people opposite to certain elites. Yet, typical to the contemporary ‘economies of visibility’ (Banet-Weiser 2018) and ‘attention inequality’ (Zhu & Lehrman 2016), where most misinformation is spread by a handful of powerful disseminators (CCDH 2021), the most those critical of the mainstream power and knowledge institutions and elites are also a part of them. In the corpus, the ‘Plandemic’ video is often framed and shared as an investigative documentary. The two most retweeted tweets, both claiming to uncover secrets of Obamagate, are by a former Fox News journalist and a well-known actor, and the top ten includes two tweets by a doctor questioning the mainstream view on Covid-19. Arguably, this dynamics can be said to be a foundational characteristic of epistemic diagonality of conspiracies: visibility and virality attract additional, sometimes contradictory and opposing meanings.

Uncover

The exploration of the keyword ‘uncover’ showed that the use of the word revolved around claims of how previously secret or hidden undertakings are, or are about to be, revealed. For example, the uncovering of ObamaGate is recurrent in the 10 most retweeted tweets in the corpus (#2, #4, #8). In line with Gray’s notion of a conspiracy theory as “a counter-discourse of some sort; that is, a discourse seeking to challenge the orthodox or dominant explanation for an event” (2010, 29. See also Buenting & Taylor 2010), the uncovering may be interpreted as a form of counteraction where the uncovering represents an act against those in power aimed at exposing their deception, that is, revealing the truth. Bringing misdeeds of powerful groups or persons into the light, conspiracy theories may be used to unsettle or invalidate dominant hierarchies and reconceptualize (conspiratorial) understandings of reality (Sapountzis & Condor 2013). Along these lines, the actions that are uncovered typically concern powerful actors’ reprehensible behavior. In this case, and as mentioned above, former president of the United States Barack Obama seems to be the powerful agent whose political power and legitimacy is being challenged. What is more, the majority of the 10 most retweeted tweets include links to different media outlets which, in turn, can be placed at different points on the political spectrum (#2 and #1, Fig. 10). This may give some clues about political orientation, but cannot conclusively tell whether political views expressed in the tweets align with those of the linked media outlets.

Fig. 10 Tweets featuring examples of 'uncovering'

The word tree showed that some of the most frequent words that follow ‘uncover’ include ‘truth’ and ‘conspiracy’, and that some of the most frequent words that precede ‘uncover’ include for example ‘to’ and ‘will’ which indicate that something has been, or is about to, be uncovered by someone (e.g., ‘police’ or ‘researchers’). . Zooming in on specific branches of the tree showed that ‘uncover’ was used in relation to a variety of claims including, for example, ObamaGate, PizzaGate, Trump, Big Pharma, 5G, QAnon. But the word tree also showed opposing claims including, for example, how Trump has given credence to ‘wacky’ QAnon, and further ridicule QAnon’s mission to uncover the truth. Although the latter use of the word appears to be significantly less common in the corpus, the word tree shows that the word ‘uncover’ is used by different actors who do not necessarily sympathize with the same political actors, or endorse conspiracy theories. While these observations are insufficient to draw any decisive conclusions, it nonetheless gives some valuable clues of the fluid or ‘diagonal’ use of the word.

Fig. 11 Wordtree for ‘uncover’

Truth

Among the epistemic keywords analysed, ‘truth’ is the most frequently used (170k occurrences). While the epistemic environment, quality and action (‘deception’, ‘hidden’, ‘uncover’ - respectively) point to a clear articulation of the epistemic condition in which the actors are embedded, ‘truth’ is more elusive. Being one of the ultimate horizons of epistemic activity, ‘truth’ has no specific orientation to a single nexus of environment and action. The overall wordtree (Fig. 12) shows identifiable pragmatic and doxastic keywords (‘know’, ‘tell’) and, more importantly, a subjectivity: the ‘truthseeker’. This subjectivity is diagonal: it seeks a particular truth-effect, to uncover the hidden, to unmask deception.

Fig. 12 Wordtree for "Truth'

Navigating the wordtree we visualize what truth ‘is’. It is a ‘revolutionary act’, lived through with a temporality of messianic expectation: ‘the truth is coming out’. This truth-to-come covers a series of interrelated matters of concern (Latour, 2004): the ‘deep state’, ‘5G’, the ‘genocide’, ‘covid’ and ‘vaccines.

Fig. 13 Exploring the wordtree.

The following wordtree composition (Fig. 14) tells a more complex narration, an almost paradoxical formula that seems to capture an hermeneutics: “there is a conspiracy to hide the truth about this conspiracy”. As we can see, the antagonized actors are the ‘deep state’ and ‘they’ (Tuters & Hagen, 2018), defined by pragmatic keywords such as ‘trying’, ‘want’, ‘plan’. Yet, ‘you can’t hide the truth’, because it is coming.

Fig. 14 Composing wordtrees.

In Fig. 15, we can see a series of pragmatic keywords such as ‘stand’, ‘fight, ‘search, ‘check content’, sketching the set of actions required when it’s ‘time’ for the ‘battle for truth’. Interestingly, post-truth appears in the overall wordtree: these conspiratorial actors recognize they are living in a post-truth media ecology. Nevertheless, these actors do not frame themselves as involved in post-truth, because ‘their regime has brought’ it. This use of post-truth may signal a diagonalization of ‘critical theory’ theory concepts blurring the line between producers and consumers of post-truth. In this sense, these communities live through an epistemic condition in which truth is hidden and post-truth is evident. Yet, post-post-truth is coming, accelerated by a networked hermeneutics that enrols actors in searching, sharing and uncovering the deception.

Fig. 15 Before and after truth.

Fig. 16 is a network visualization of the co-tag network of the ‘truth’ subcorpora. Node and edge color encode the weighted degree of connectivity of the ndoses. As we can see, it’s conspiritual in content: religious and mundane entities converge in the field of concern generated by ‘covid’, ‘coronavirus’, ‘scamdemic’, ‘plandemic’. QAnon-adjacent keywords populate most of the network. #QAnon and #WW1WGA are the most frequently co-occurring hashtags, reflecting the notorious capacity for amplification that characterizes that community. It is clear that QAnon was able to drive network convergence and integration, but this primacy should not be interpreted as stability. As we can see, established QAnon matters of concern (pizzagate) are not central in the network, dethroned by a new source of conspiratorial discourse: the pandemic. This may signal a ‘diagonalization of QAnon’: while communities reassemble around a new core concern, they import the schemas of interpretation of their predecessors. Actors freshly initiated to systematic conspiratorial thinking find themselves in an ecology that provides them with a tradition of epistemic keywords and hermeneutic practices. This may signal that QAnon’s epistemic style can survive while the newsworthiness of its classic antagonists (the Clintons) fades, thus traveling to new diagonal movements. While the analysis of the sociopolitical base of those movements can be incredibly disorienting, the epistemic keyword approach may enhance the traceability of these political trajectories.

Fig. 16 Network visualization

In sum, our analysis shows that the epistemic keywords under scrutiny tap into elements of deception: uncovering hidden truths implies that deception is taking, or has taken, place. What is more, the analysis was informed by the notion of ‘diagonalism’ (Slobodian & Callison 2021) and the findings presented above seem to indicate that the epistemic keywords explored in this sub-project are used as a discursive device employed by actors/entities across the political spectrum (see Fig. 17, below). The difference lies in who is portrayed as the deceiver(s), or put differently, whose deception or hidden plans are, or are about to be, uncovered. In summary, while these initial findings warrant further study they seem to substantiate that notions of ‘diagonalism’ (Slobodian & Callison 2021) may offer a fruitful route into a better understanding of the mutli-faceted nature of online conspiracies. Further research might take into account media effects and the problem of digital bias (see e.g. Marres 2015)

Fig. 17 Political compass of epistemic keywords

Week 2: Detecting Convergence & Integration in Instagram

Link to Recipe week 2

Link to final presentation week 2

Subproject 1: hashtag analysis

Initial Data Sets

The dataset comprised ~800k Instagram posts with 4.3 million comments with the earliest post being from late 2011. We focussed only on 2020, which also happened to be by far the biggest year in the data. Posts were gathered from 66 known conspiracy influencers/accounts as well as posts containing at least one of the following 82 hashtags related to conspiracies.

"wakeupsheeple", "davidicke", "wedonotconsent", "rfidchip",

"cabal", "attilahildmann", "thestormisuponus", "markofthebeast",

"governmentlies", "wherewegoonewegoall", "hollyweirdisevil",

"idonotconsent", "event201", "fearmongering", "populationcontrol",

"nonewnormal", "saynotobillgates", "healthfreedom", "believemothers",

"gibgateskeinechance", "filmyourhospital", "weareq", "fuckbillgates",

"fucknwo", "id2020", "hisnamewassethrich", "firefauci", "truther",

"plannedemic", "widerstand2020", "plandemic2020", "billgatesisevil",

"fakevirus", "stopbillgates", "qanonarmy", "scamdemic", "arrestbillgates",

"protruth", "wearethenewsnow", "coronafake", "andrenochrome",

"coronalüge", "projectbluebeam", "outofshadows", "darktolight",

"qarmy", "bodoschiffmann", "coronalies", "givegatesnochance",

"freedomkeepers", "stop5gflorida", "betweenmeandmydoctor", "stop5gaustralia",

"medicalexemption", "reopenusa", "medicalrights", "stop5gbarcelona",

"medicalfreedomofchoice", "pizzagate", "stop5guk", "stop5gcalifornia",

"parentalrights", "georgesoros", "stop5grollout", "readtheinsert",

"stop5gitalia", "plandemic", "stop5gespana", "stop5gpennsylvania",

"stop5gcentralcoast", "informedconsent", "stop5geverywhere",

"stop5gtowers", "stop5gglobal", "stop5ginternational", "stop5ghawaii",

"stop5gtoday", "stop5gworldwide", "vaccinationchoice", "stop5gusa",

"stop5gitaly", "mindcontrol"

Research Questions

Can we detect convergence or integration in conspiracy theories, at the level of hashtag use?

Methodology

We broke the dataset into three quarters (the fourth quarter did not appear complete). We manually categorized hashtags into communities. With these quarters created maps of co-hashtag occurrence using gephi, and created coloured overlays in order to try to detect convergence or integration.

Discussion

In the graphs below one can see that QAnon (blue) stays remarkably stable over time, apart from the Pizzagate (purple) which grows and somewhat overlaps somewhat with communities in the top of the graph. At the top of the graphs we see the growth of Covid-related hashtags (green), which do not overlap much with QAnon, though they do with Bill Gates (red), NWO (yellow), 5G (brown) and Conspirituality (orange). Oddly Bill Gates (red) is not present in the first quarter, though he is in other levels of analysis, see below.

Fig. 1 Hashtag co-occurence networks for instagram data Q1-Q3, coloured by conspiracy tribe

Preliminary conclusions

As a community QAnon (blue) was remarkably stable across 2020 and its issues had relatively little to do with the pandemic. There was little integration and convergence on display in this community, except for the Pizzagate (purple) subcommunity which came increasingly to focus on pedophelia. By contrast the other communities—Bill Gates (red), NWO (yellow), 5G (brown) and Conspirituality (orange)—featured both integration and convergence leading over the course of 2020 to the emergence of what Naomi Klein has called the Great Reset conspiracy smoothie.

Subproject 2: Named Entities

Contributors: Stefan Baack, Marcantonio Bracale Syrnikov, Shirley Chan, Johanne Kalsaas, Johanna Muhrbeck, Martin Roth, Yanling Zhao

Overview of networks over time

Fig.2 Co-occurence networks of named entities in instagram posts over time

Bill Gates

Main contributors: Johanne Kalsaas, Johanna Muhrbeck

This part examined antagonistic discourses on Instagram which may be conceptualized as a site of platformed antagonism (Farkas, Scho & Neumayer 2018). Through qualitative and quantitative textual analysis of Instagram posts in English in each dataset (Q1, Q2, Q3, Q4) it explored whether, and if so how, Bill Gates was antagonized over time. Throughout our data we found that the Instagram posts continually reproduced a discursive construction of Bill Gates as an antagonistic enemy of humanity.

Fig. 3: Histogram of posts mentioning Bill Gates on Instagram in 2020

Textual analysis

For the textual analysis, we used 4Cat to create a sub-corpus of all Instagram-posts mentioning Bill Gates. The corpus was consequently divided into quarters (Q1=353 posts, Q2=4181, Q3=1889, Q4=286), so we could study changes over time. For each quarter we went on to construct the Jason Davies wordtrees, as well as looking in depth at the top 10 most-liked posts. We thereby aimed to answer a threefold research question: 1

1. How was Bill Gates antagonized on Instagram in 2020?
2. To what extent was Bill Gates connected to other conspiracy narratives?
3. Can we observe a degree of convergence and/or integration of Instagram conspiracism around the figure of Bill Gates?

The histogram of our corpus shows an explosive increase in Instagram posts mentioning Bill Gates at the start of the Covid-pandemic. This pattern is illustrative of how Gates quickly was construed as the main agent of the pandemic by online conspiracists. Although the main conspiratorial claim - that Bill Gates has a hidden vaccine agenda - remained central over time, the content and focus of the posts, and hence the basis upon which Bill Gates was constructed as an antagonist in the posts, varied somewhat across the datasets. Overall, a close reading suggests that the top posts in Q1 emphasized Bill Gates' evil and reprehensible motives, while in Q2 the posts concentrated on portraying Bill Gates as a criminal and often called for legislative measures. In Q3 these types of allegations seem to escalate: Bill Gates is not only labeled as a criminal, pedophile monster and murderer, but as the devil or Antichrist himself. In Q4, the content appears to enter a less emotional, more ‘analytic’ stage. What is more, while Melinda and the Gates foundation are central entities in the first quarter, they all but disappear in the last three. As the pandemic progresses, ‘Gates’ becomes increasingly synonymous with Bill. Another, final key finding can be exemplified by a post that was published only a few days after the US lockdown was implemented:

Fig. 4: Wordtree Q1

Fig. 5: Wordtree Q2

Fig. 6: Wordtree Q3

Fig. 7: Wordtree Q4

Q1

One key feature of how Bill Gates is constructed in the first quarter of 2020 is the idea of his involvement in vaccine development as a pretext for controlling the world’s population through microchip implants:

Following the narrative “vaccines as depopulation” (which preceded and was amplified by the corona pandemic), Gates is frequently cast as a eugenicist:

This type of allegations are often accompanied by claims that construct Bill Gates as in charge of a ‘global elites’ whose objective is to establish the ‘New World Order’:

Bill Gates is also constructed as fundamentally evil and connected to satan:

Q2

In the second quarter of 2020, Bill Gates remains connected to the notion of the alleged mandatory vaccination programs as an evil plot for world domination. However, two new themes emerge during this period. First, claims made by Robert F. Kennedy Jr. about Bill Gates were repeatedly posted: :

In this and in other similar posts, Bill Gates is connected to power ambitions and strategic experimentation with human lives. The second recurrent theme is the construction of Bill (and to some extent Melinda) Gates as guilty of crimes against humanity and medical malpractice:

A petition to investigate Bill Gates for crimes against humanity and medical malpractice has amassed a staggering 410,000 signatures from concerned citizens, more than four times the number required for an official response from the White House. (Instagram post J, 2020-04-23)

Q3

In the third quarter of 2020, claims relating to an elite vaccine agenda, depopulation of the world’ and crimes against humanity from the previous quarters are reproduced. The construction of Bill Gates as an eugenist is also repeated:

Q4

Finally in the fourth quarter of 2020, the posts appear focused on Bill Gates evil motives behind his goal of global vaccination and places emphasis on the construction of Bill Gates as the head of the whole operation, reflecting on his position in the tyranny of elites’ and unfold the mechanics behind his power:

Network analysis

The findings from the textual analysis could be verified with a network analysis, where we analyzed how various keywords co-occurred in the text of Instagram posts. While already being prominent in Q1, Bill Gates starts to strongly attract various Covid-19 related conspiracy theories from Q2 onwards:

Fig. 8: Q2 network showing persons and conspiracy theories (Force Atlas 2).

In the network in Fig. 8 edges are weighted based on co-occurrence, the size of the nodes represents the degree and the colors indicate the type of node (person vs. conspiracy theory). As you can see, Bill Gates has a very high degree and is thus very connected (second highest degree behind ‘trump’). Moreover, Fig. 8 shows how he is surrounded by various conspiracies and this connectedness with conspiracies removes Bill Gates slightly from the rest of nodes representing persons. Two main conclusions can be drawn from the observations made: the discourse in Instagram posts reveal a continuous production of Bill Gates as representing an antagonistic enemy full of self-importance and engaged in wrongful and criminal activities, and an overlap between content relating to Bill Gates and other, often anti-vaccine, conspiracy narratives. The discourse on Instagram about Bill Gates remains consistently and forcefully antagonistic throughout our corpus. The majority of the analyzed posts clearly expresses antagonism by reference to the alleged motives of Bill Gates and his wrongful, even criminal, activities which commonly relates to this alleged global vaccine agenda. Bill Gates is described as evil, mad with power, inhumane and is repeatedly connected to criminal behavior including crimes against humanity and pedophilia. Moreover, Bill Gates is constructed as the responsible mastermind of an elitist New World Order whose objective is global population control.

The second conclusion relates to the observation that many of the analyzed Instagram posts often include and merge a vast range of various conspiracy narratives, especially through an active use of seemingly unstructured hashtagging. For example, the #WWG1WGA which can be associated with QAnon (Morrish 2020), and the #WEAREFARRAKHAN which can be linked to the Nation of Islam (Mason 2021), can be observed in the posts. The posts do not necessarily push a coherent conspiracy theory, but rather various conspiracy narratives that seem to align with the notion of ‘conspiracy without theory’ as proposed by Rosenblum & Muirhead (2020). Such conspiratorial, discursive practices in online contexts may be fruitfully explored from the perspective of secondary orality in which ritualistic, repetitive performativity is a central feature (Venturini, forthcoming). Finally, some caution should be taken in interpreting this kind of online conspiracism. As has been previously reported, various actors have exploited trending hashtags in order to reach wider audiences. For example, Daesh has relied on this tactic (EUvsDisinfo 2020) and it has been discovered that QAnon hijacked #SaveTheChildren to reach new audiences after being banned on various platforms (Roose 2020). In other words, the extensive use of hashtags may in some situations thus be a mere tactic to build an audience.

Microsoft

Maincontributors: Shirley Chan, Marcantonio Bracale

A sub-corpus was also created of instagram-posts with Microsoft mentions. Subsequently, the corpus was divided into quarters on 4cat similar to the sub-corpus of instagram-posts mentioning Bill Gates. The textual analysis focused on constructing word trees and close reading of the top 10 most-liked posts for each quarter. We thereby addressed the following research questions:

1. How is Microsoft portrayed within the instagram dataset from 2020?
2. What actors and organisations are mentioned in conjunction with Microsoft?
3. How is Microsoft situated within the conspiratorial narratives in the dataset?

Textual analysis

To gain more insights into the conspiratorial narratives centered around Bill Gates, we also looked into instagram-posts mentioning Microsoft. The decision to investigate Microsoft further was fuelled by the findings from the network analysis of organisations where Microsoft along with the entity “New World Order” was identified as the primary integrating nodes. The textual analysis of Microsoft followed the similar process for the textual analysis of posts mentioning Bill Gates.

Q1

The connection between microsoft and vaccine/health/pharma discourses is present in Q1. Microsoft is often mentioned in connection to Bill Gates and other initiatives that he is involved in, for example, ID2020 and event 201 where the latter is hosted by the Bill and Melinda foundation. Microsoft is primarily mentioned in conjunction with Gates, highlighted as one platform that he utilises to gain control over people. Gates and Microsoft are perceived to be working in cohorts with tech-, pharma companies, media and the government to implement social control and global surveillance. The digital ID (ID 2020), vaccine, 5G and wealth of social media data are considered as resources that Gates utilises to gain global control:

Other interesting findings consist of several posts that highlight the Netflix documentary “Inside Bill's Brain: Decoding Bill Gates” (2019), claiming that one can see signs of Gates plotting the pandemic.

Fig. 9 Word tree of Microsoft mentions in Q1

Q2

Microsoft is discussed aligned with Gates who is perceived to be in charge of the pandemic and the vaccine. Several posts highlight that Gates utilises his economic capital to gain control over influential actors and organisations, for instance US health official Anthony Fauci and the WHO (World Health Organisation):

Claims are also made that Gates is utilising the pandemic to gain social control globally. For example, Microsoft is mentioned in conjunction with ID 2020 and biohacking where claims are made that Bill Gates through vaccination and technology tries to implement a “microchip” to control population growth, fertility and human behaviour in general (for example, sell and buy cryptocurrency). Several posts highlight that Gates has separated himself from Microsoft to become a philanthropist. However, they still suspect the company and Gates to be in cohort.

Fig. 10 Word tree of Microsoft mentions in Q2

Our textual analysis reveals that there was a prevalent focus on Bill Gates rather than Microsoft within the dataset. Microsoft is highlighted as a “partner in crime” or as a platform/resource for Gates to gain global control. One main take from the textual analysis of Microsoft mentions on Instagram is similar to the findings from the network explorations of organisations where several different spheres become intertwined. In this entanglement, Bill Gates (and Microsoft to a lesser extent) is shown to act as an integrating force. Additionally, several conspiratorial narratives within the corpus also show a clear emphasis on the pandemic, which could be argued to amplify the conspiratorial narratives of Bill Gates.

Network analysis

Network analysis is a useful methodological tool to complement textual analysis. Named Entity Recognition enables the researcher to extract named persons and institutions from a textual corpus (in this case, Instagram posts) and analyse their connectedness over time. By following the actors while they model the social structure they inhabit, their efforts to assemble a ‘cognitive mapping’ of their own epistemic crisis, we can trace network integration and convergence. The following co-occurrence network visualizes the most connected organisation (weighted degree > 10); edge width encodes the weighted degree of a node. We consider the transition from Q1 to Q2 2020 as a transformative moment in a conspiratorial dataset. The beginning of the COVID-19 pandemic parallels an increase of belief in CTs and their normalisation in the context of a generalized societal crisis and its psychological stressors (De Coninck et al., 2021).

Fig. 11 Organisation Network

The most evident network convergence that characterises the transition from Q1 to Q2 is the integration between two archetypal organisations of the “covert sphere” (Melley, 2012), the FBI and the CIA, with the Big Pharma conspiracy theory (Blaskiewicz, 2013). Moreover, we can observe how in Q2 the names of pharmaceutical companies involved in vaccine research (Bayer, Pfizer, etc) start to appear. The force-directed layout we used (MultiGravity Atlas 2), and the resulting spatialization of named entity co-occurrence, shows a marked difference between the two quarters: in Q1, ‘CIA’ and ‘FBI’ occupy an antipodal position in respect to ‘big pharma; in Q2, they are strongly connected. This network dynamic reflects the acceleration of conspiratorial narratives centering on the pandemic and their ability to assemble ever-expanding conspiratorial weavings that schematize real-world events.

The network analysis of the ‘Microsoft’ named entity produces similar findings. In this case we focus on the ego-network of ‘Microsoft’, including both persons and organisations. Edge weight and color encodes the weighted degree on the node. A tendency of this network is to show increasing convergence towards conspirituality (Ward & Voas, 2011): from Q1 to Q2, ‘God’ and ‘Jesus’ gain network centrality, together with ‘evil’ and ‘David Icke’. In other words, a conspiratorial networking reassembled the narratives around ‘Microsoft’, moving from the distant domain of multinational capitalism and intergovernmental organizations to the pulsating center of the conspiratorial toolkit and ‘name dropping’. Another interesting feature is the increasing network antagonism between ‘Trump’ and ‘Bill Gates’: both strongly connected to ‘God’, they seem to confront each other as the network authorities, moral figures and political options.

Fig. 12 Microsoft Ego-Network

As in the previous graph, the ‘Microsoft’ ego-network shows integration and convergence in which previously distant nodes are now associated by strong connections. In this case, the integration of ‘Microsoft’ in the conspiratorial network is realized by the conspiritual cluster: once the ‘God’ node is associated with ‘Bill Gates’ (reflecting increasing textual co-occurrence) a new narrative can be brought forth, producing matters of concern on a completely different scale, moving from the merely polemical to the existential. In this sense, when conspiratorial convergence is realized it’s because it is able to enrol actors from different domains. Reconfiguring the global connectivity of the collective and connective ‘cognitive mapping’ of the current epistemic crisis, convergence brings actors from difference scales and modes of existence together: when ‘CIA’ and ‘big pharma’ converge a layer of depth and secrecy is added to the hermeneutical style of conspiratorial actors. When ‘Bill Gates’ and ‘Trump’ are triangulated by ‘God’, a sense of moral and spiritual urgency feeds into the interpretation schema. While epistemic keywords tried to capture how actors collect entities to name, entity recognition reports which actors are named. In this sense, network convergence implies network expansion: while integration provides disparate and diagonal communities a common ground through the networking of shared keywords, convergence restructures global connectivity by pushing the network authorities of these communities in the foreground. In our case, convergence is mediated and promoted by the ‘God’ node. Once convergence is realized, all the chains of association from a network can be transferred to another one. Suddenly, from Q1 to Q2, ‘Microsoft’ is at the center of a story that belongs to a cosmic, religious scale.

Subproject 3: Images of Bill Gates

Main contributors: Olga Bialy, Golnar Gishnizjani, Ben Kapadia, Tong WU, Yiran Zhao

Visual analysis of Bill Gates images from January to September 2020 demonstrate he has been antagonized under the umbrella of conspiracy theory on Instagram. In the first nine months of 2020, images related to Gates have been integrated into various communities/tribes of conspiracy theory from bioweapon to global government. However, a wide range of visual narratives around Gates did not construct a convergent narrative. Indeed, we have seen fragmentary visual narratives that do not necessarily have shared entities even though they occur together or simultaneously. We argue that along with textual narratives surrounding Bill Gates, visual content is mainly accompanied by a chain of hashtags highly affected by the oral culture on Instagram. Although images on Instagram antagonized Gates prevalently, existing visual narratives are evanescence. Thus, conspiratorial images of Bill Gates are more provocative rather than persuasive (Venturini, upcoming). Our analysis is based on a dataset of 500 Instagram images about Bill Gates, including the original images, post texts, hashtags, and author information, covering the first three quarters of 2020. Both the original dataset on Pixplot and the one converted into spreadsheets were used. We thereby addressed the follwoing research questions:

1. How is Bill Gates represented or antagonized in the images?
2. To what extent is Bill Gates the target of convergence between different communities or tribes?

Categorization

To analyze if a possible antagonism towards Bill Gates is performed by different tribes and communities, this part of the analysis aims at the identification of convergence. 500 images of Bill Gates collected from Instagram and posted in the first three quarters of 2020 (January till September) were the object of this analysis. According to the definitions of conspiracy theories relating to Covid 19 developed by the Infodemics project (2020), the dataset was thematically coded. The seven categories created by the Infodemics project were used, whereupon the category Bill Gates was subcoded in further categories, resulting in 22 subcategories used. The subcategories were defined jointly by the group members, combining the image content and the original post text and hashtags. Each image was put into one single category according to its content. A dump chart of the categories was then created using RawGraphs to show the overall trend of dominant/marginal communities around Bill Gates as a common antagonist. Two similar charts were also created to separately show the changes in likes and comments.

Pixplot image clusters

To understand in which degree Bill Gates is antagonized in Instagram posts by different groups, a cluster analysis with Pixplot was conducted. Pixplot helps to visualize the 500 images of Bill Gates as thematic clusters by using machine learning techniques (processing with an Inception Convolutional Neural Network and projected with the UMAP algorithm) to project similar and related images nearby one another. The use of Pixplot makes it possible to visually find out if tribes with different backgrounds make use of the same images in their posts to antagonize Bill Gates and further, if these actions create a certain convergence of different tribes. Inside Pixplot, the projected images are linked to the authors which posted the images. In addition to the original image clusters, we also filtered the images according to alternative parameters. Five most-used hashtags were selected to show the articulation between images and hashtags, including #billgates, #coronavirus, #billgatesevil, #Id2020, and #covid19. Furthermore, the most repetitive images within these hashtags were filtered so that we can gain a basic cognition about the discourse around the topics.

Network analysis

A bipartite author-category network was constructed based on the aforementioned categorization of Bill Gates images, showing the relations between users and categorized images. The index of modularity was calculated to show the co-occurence of certain users and subcategories, as well as the connections between different subcategories, which allows us to further analyze the communities around specific topics.

Discussion

In the following part, dominant and marginal communities/tribes in which Bill Gates is antagonized over the first nine months of 2020 on Instagram are discussed.

Fig. 13 Dominant and marginal communities around Bill Gates

1. Bioweapon is the main tribe of how Bill Gates is represented on Instagram. The results indicate that Gates is depicted as a person who intends to control and depopulate the world by promoting the idea of getting the vaccine and involving in the vaccine production process. Even though no significant change was observed in the prevalence of this idea over the period of the study, users mainly targeted Gates for implementing microchips (this discussion was heated from March to June 2020). Relying on screenshots of news headlines from alt-media that covered stories about Gates's actions in India and some African tribes, users consistently accused Gates of controlling the population and creating a "New World." Amid this group, the memes and news headlines circulate to emphasize that Gates deliberates to modify human species by selectively mating people. In fact, Instagram users who fit in the bioweapon community/tribe claim Gates practices eugenics.

Fig. 14 Dominant and marginal communities over time

2. In relation to the bioweapon community/tribe in which people highlight the Covid-19 vaccine as a way of governing people and shaping a new order globally, a notable group of users manifests their opposition to the vaccine by targeting Bill Gates. They do not merely claim the vaccine is made for breeding but also defend the idea that the virus is released intentionally. This group accused Gates and other people involved in making the vaccine, including Dr. Fauci, of profiting from the process. Even though the activities of anti-vaccine people initiate from the beginning of the Covid-19 outbreak, especially in the US, the debate about Bill Gates has popped up from June to September. Like the previous community/tribe we observed, using Bill Gates images with the text inside them is the most common way of maintaining him as one of the key faces in conspiracy narratives on Instagram.

Fig 15. Network of topic and user clusters

3. Among the community/tribe that accuses Gates of human rights abuses, the world's ideas of breeding and depopulation are also reproduced. Furthermore, users depict Gates as a pedophile and murder. Images shared by this community/tribe portray Gates as an evil and monster. Indeed, users implemented stereotypical facial characteristics of the demon in media to illustrate the real face of Gates that is hidden behind his seeming humanitarians' activities. Through making connections between all these stories and promoting mandatory vaccinations by Gates, this community/tribe asserts that Gates is a deceiver who intends to take the world under control. According to the graph, April and May were the heated months for this debate; however, he has constantly been a target of such allegations.

4. Another community/tribe that antagonized Gates is QAnon. The supporters of this theory expressed their opposition to Bill Gates in various ways throughout the first nine months of 2020. In another way, we have not observed a significant change regarding escalating this issue over time. Moreover, QAnon includes a wide range of topics that supporters propagate; due to that, images categorized in this community/tribe are interconnected with others. The footprint of this community/tribe is clearly visible among Trumpist and anti-vaccine groups. Comparable to other categories, news headlines and memes are the most circulated images within the QAnon community/tribe. 5. To put it concisely, we placed images in 22 categories; however, they cannot all be named as a community/tribe. As the findings indicate, some categories are connected and shape the larger community that can interpret under conspiracy theory discourse on Instagram. Even though thematic dispersion among the images is significant, they do not merely belong to a specific category. Indeed, one image can point to Gates’ inhuman face and simultaneously raise the issue of depopulation. Since images do not solely indicate how Bill Gates is constructed on Instagram, textual content and hashtags also should be considered to provide a comprehensive understanding. Correspondingly, the practice of implanting a chain of hashtags to share images is a meaningful sign that shows the overlap of images themes. Interestingly, the variety of images is not as much as the themes. Clustering images in Pixplot based on the most popular hashtags evidence that similar images circulated in different communities/tribes. Mediated practices of users on making an antagonistic face from Gates through sharing images is repetitive.

Fig. 16: overview of 5 top hashtags images/ #billgates, #coronavirus, #billgatesevil, #Id2020, and #covid19

Preliminary conclusions

The results demonstrate an uninterrupted representation of Bill Gates on Instagram as an antagonistic face that engages in various evil actions. By sharing images on Instagram that most of them are news headline screenshots, memes, and illustrated monster images, Bill Gates is depicted as an evil, imposter, greedy, inhuman who commit several crimes from killing people to pedophilia. It is declared that the ultimate aim of Bill Gates and the social bubble around (elite people) him is to create a new world order and dominate the world's population. The other significant point is the overlapping of various conspiracy visual narratives. Observing the similar images in different communities/tribes throughout the first nine months of 2020 shows there is no convergent conspiratorial narrative about this key face of conspiracy theory at the time of the pandemic. However, the existence of a remarkable integrated visual narrative from Gates does not contribute to a concrete and well-structured theory on Instagram. As Venturini (upcoming) asserts, features of secondary orality, including ritualistic and repetitive performativity, are detectable amid those who make an antagonistic face of Bill Gates on Instagram. To sum up, it should be noticed further research is needed to recognize the visual characteristics of Bill Gates in photo-based social media as one of the pivotal faces of conspiracy theory in digitally saturated culture.

Subproject 4: Case studies

#TheGreatReset: Conspiratorial influencers and deplatforming on Instagram

Main contributors: Yelyzaveta Glybchenko, Craig Ryder, Kola Heyward-Rotimi & Kathryn Murphy

Main findings

Overall, we found that conspiratorial influencers exist in a much bigger web of accounts and hashtags. The data identified a co-occurence of other conspiracies and related hashtags in the influencers’ content including 5G, satanism, Pedogate and the New World Order (NWO). The influencer accounts identified were not outliers in the network, suggesting convergence between conspiracy content on Instagram. This suggests that individuals who share conspiracy content are likely to be interested in a broad range of theories, however further analysis is necessary to determine whether they are contributing to the wider convergence of conspiracies. What we can conclude, however, is that it is difficult to sustain prominence as a conspiratorial influencer with accounts facing difficulties as content moderation systems work to reduce disinformation online.

Introduction

This research focuses on influencers of conspiracy content online relating to The Great Reset. Literature generally uses the term influencer to refer to ‘everyday, ordinary Internet users who accumulate a relatively large following on blogs and social media through the textual and visual narration of their personal lives and lifestyles, engage with their following in "digital" and "physical" spaces, and monetize their following by integrating "advertorials" into their blog or social media posts and making physical appearances at events.' (Abidin, 2015). However, for conspiratorial influencers, the notion of sharing their personal lives and integrating advertorials is less relevant. Alternatively, the term microcelebrity is often used to describe ‘a state of being famous to a niche group of people, but it is also a behavior: the presentation of oneself as a celebrity regardless of who is paying attention’ (Marwick, 2013: 114). According to Marwick, ‘There are two ways of achieving internet fame--by consciously arranging the self to achieve recognition, or by being ascribed fame by others due to one's accomplishments.' (Marwick, 2013: 114). Whilst influencer implies that individuals have accrued a level of status, microcelebrity focuses more on mindset and behaviour, and is less concerned with actual audience size. For conspiracy influencers, the notion of ‘being famous to a niche group of people’ is particularly relevant and can help understand the ways figures are able to gain recognition from a small group, but yet generally lack any sense of mainstream recognition outside of their conspiracy community. Within networks, influencers can be viewed as ‘hubs that relay messages to others, often lending their credibility to these messages’ (Ingenhoff et al., 2021: 1). Thus, influencers can be considered ‘opinion leaders’ in a two-step flow model of communication wherein information flows through personal networks via influential individuals (opinion leaders) (Katz & Lazarsfeld, 1955). It is this ability to disseminate messages to others with a level of credibility, and therefore influence, that makes these users interesting and worthwhile to study, particularly when they may be responsible for the widespread communication of misleading conspiratorial narratives. In particular, exploring influential accounts alongside hashtags enables the notion of symbolic power to be interrogated generally, and in relation to individual hashtags (Niederer & Columbo in Rogers, 2021). Whilst conspiracy theorists are not commonly positioned as influencers within existing literature, we consider this a valuable way to conceptualise the way they operate. Thus, this research aims to contribute to existing literature around conspiracy theorists by positioning key actors within conspiracy networks as influencers of ideas. The research seeks to interrogate the significance of these actors in the dissemination of content around The Great Reset on Instagram, considering which users possess the most symbolic power, as well as how these accounts relate with other users and hashtags in the broader conspiracy discourse (Rogers, 2021). We also consider Venturini’s assertion that conspiracy theories mimic the linguistic traditions of oral cultures in what he terms “secondary orality” (2021). Venturini argues that conspiracy narratives are allowed to “bake” on 4chan because the ephemerality of the message boards encourages users “assure perpetuation-without-preservation of their folklore through the classic oral technique of tireless repetition” (2021, para 19). Truth gathers veracity, on 4chan, through socio-technical affordances that 1) remove content that does have the approval of the crowd and 2) encourages the regurgitation of successful, or influential, ideas, regardless of their sincerity. We predict that Instagram curates a different base of ideas and communities, but no less informed by the discretion of its opaque platform affordances.

Data

1315 Instagram posts from the initial dataset relating to The Great Reset

Research questions

The research focused on ‘actors and entrepreneurs’ of conspiracy theories (Venturini, forthcoming) with the aim to identify key actors in the dissemination of #TheGreatReset content on Instagram in 2020. As such, our research questions were:

1. Who are the key influencers of conspiratorial ideas relating to #TheGreatReset?
2. What types of content are they sharing?
3. How does their content on #TheGreatReset interact with other conspiracy theories?

Methodology

To properly analyze #TheGreatReset within the larger Instagram conspiracy network, we had to filter the data to look at only the actors relevant to the hashtag/theory. Achieving this took multiple attempts, using 4CAT’s Fabio_instagram_post_text to narrow the network to our project’s focus. The first lexicon filter we implemented used the term “the great reset,” which restricted the network too far, returning only 100 results and all from 2020. Widening the timeframe for our lexicon term search did not result in better results. Using multiple terms along with “the great reset” was most effective. On one iteration, the great reset, thegreatreset, and #thegreatreset were used which provided 888 results across multiple years. Because of limitations on date filtering in 4CAT, all of our searches resulted in accounts that spanned more than one year. To restrict this information to 2020 only, we had to manually select those posts. This was achieved by deleting rows in our data spreadsheet that corresponded with accounts outside of the 2020 timeframe. The final iteration that was interpreted through a Gephi visualization used the great reset, thegreatreset, #thegreatreset, and greatreset and provided 1315 results. With the 4CAT results, we were able to visualize the network through Gephi. This process was achieved by partitioning the nodes into two colors, with pink equating to hashtags and green to accounts. After delineating the visualization through these color partitions, providing higher readability to assess which hashtags were being used by who, multiple layouts were applied to the data. This includes Yifan Hu and Fruchterman Reingold, both of which are displayed in the following pages. Both are force-directed layouts, so they position nodes through a set of rules based on tension between edges and nodes. Yifan Hu was more helpful to visualize the distinct clusters of hashtags that are particular to individual users, while Fruchterman Reingold clustered all of the most popular accounts and hashtags in the center which also provides useful data about the network’s hierarchies. We also conducted content analysis of the filtered 4CAT dataset to identify the most prominent users across the dataset as a whole, and in each quarter of 2020. This involved calculating users who posted most frequently, and received the most likes and comments on their posts within each time frame, and drawing up a list of the top 10 usernames for each measure. To identify key influencers within the dataset, we then looked for usernames that recurred across more than one top 10 list, suggesting they had a sustained presence and level of audience engagement on their content. This allowed us to identify four top influencers across 2020, and the top influencer in each quarter. Combining insights from the content analysis and the Gephi visualisations enabled us to draw the overall conclusions from this research about the ways conspiratorial influencers operate.

Findings

Network analysis

To explore the network, we created multiple visualizations using Gephi. We used several network visualization layouts to interpret the relationship between accounts and the hashtags they use. Presented below are the Yifan Hu and Fruchterman Reingold visualisations showing how hashtags were distributed throughout accounts discussing The Great Reset. This identified a wider set of hashtags that occurred across a wide range of accounts such as #plandemic and #scamdemic, as well as other hashtags that are more specific to individual accounts.

Fig. 17 Yifan Hu Layout

These individual hashtags are represented through clusters tied to the main network solely through one account, instead of a web of accounts connecting to the same hashtags. While the Yifan Hu visualisation shows the unique hashtag clusters better, the Fruchterman Reingold layout provides a clearer view of the primary accounts and hashtags in the network as they appear clustered in the center.

Fig. 18 Fruchterman Reingold Layout

Having explored the overall network, we chose to zoom in on individual accounts to find top users for the year of 2020 according to the following measures: ‘most posts’, ‘most likes’ and ‘most comments’. The findings are presented in the visual below:

Fig. 19. Top Users for 2020

Conducting the time series analysis of the top accounts, we calculated the most influential accounts across quarters 2-4 of 2020 and discovered the following:

1. None of the influencers identified were active across all 3 quarters;
2. most prominent influencers differ from quarter to quarter;
3. the influencers were most active during quarter 3 (July-September) and quarter 4 (October-December);
4. Content received the most engagement (likes and comments) in quarter 3.

Presented below are the top influencers in quarters 2-4 of 2020:

Fig. 20 Top Users by Quarter, 2020.

By combining insights from the top influencers in each quarter of 2020 with the top overall influencers, we identified the top five conspiratorial influencers:

1. @criticalthinking101
2. @perceptionreconfiguration
3. @menschmodus
4. @miri.p.swiss.truths
5. @chingonaxpress

Of these five accounts, we selected the 2 most prominent influencers for further analysis using qualitative and network analytical approaches, as discussed in the next section.

Discussion of individual accounts

@criticalthinking101

Fig. 21. Deplatformed?

This user is ranked in top 5 lists for most posts, likes and comments across 2020. However, they were active only during quarter 3, when they shared 48 posts. On these 48 posts, they received 14,300 likes and 780 comments. Their account is no longer available, suggesting they may have been de-platformed or they closed their account. While the account was not available for further analysis on Instagram, an analysis through Gephi provided more insights. The one-step ego-network of the user showed the hashtags used. Some of the examples across the network include #followthewhiterabbit, #itsaboutcompliance, #5gtechnology, #satanicworship, #youarebeingliedto.

Fig. 22 One-step ego network of @criticalthinking101

While the ego network above only gives insights into the hashtagging of @criticalthinking101, the two-step ego network below situates the user in the wider network of influencers. It shows connections of @criticalthinking101 to other users using the same hashtags in the network.

Fig. 23 Two-step ego network of @criticalthinking101

@perceptionreconfiguration

Unlike the previous user, this account is still active. perceptionreconfiguration started their account on 10th July 2020 and shares content mainly relating to the Covid-19 pandemic, but occasionally also other conspiracies such as pedogate, pizzagate and satanism. Their account is solely focused on conspiratorial content, featuring no personal posts or identifying information. They first mention The Great Reset on 10th July 2020 in a video where they show the World Economic Forum’s Great Reset initiative model and use it to make a case for covid being a ‘plandemic’.

The user created a total of 53 posts between July and November 2020 using #TheGreatReset, with the hashtag often appearing alongside other popular conspiratorial hashtags (e.g. #scamdemic and #plandemic). The content showing police heavy-handedness at anti-lockdown protests was the most popular with their audience, achieving high levels of engagement. Overall, perceptionreconfiguration demonstrated an activist/leader mentality, printing posters and encouraging their followers to unite against the ‘nanny state’.

Fig. 24 One-step ego network of @perceptionreconfiguration.

Some of the hashtags they used, as seen in the one-step ego-network, include: #thebiggestscamever, #justsayno, #hoaxvirus, #fuckthenewworldorder, and #wearebeingplayed. The user’s hashtagging in the network can be seen in the 2-step ego network below.

Fig. 25 Two-step ego network of @perceptionreconfiguration

Discussion

Our research finds that conspiratorial influencers on instagram are entangled in a much bigger web of conspiratorial accounts and hashtags, integrating disparate ideas, including 5G, NWO, Pedogate and satanism into one mega-theory, or “conspiracy smoothie” (Klein, 2020). However, it is difficult for individuals to successfully influence conspiracy narratives because the recommendation algorithm treats content like a carousel and posts that do not attract engagement are marginalised in favour of more ‘popular’ content, which Instagram deems to be making greater impact (Hootsuite, 2021). Not unlike how 4chan’s automated system guillotines unpopular content, motivating users to swamp message boards with repetitive content or ‘swarm’ around posts they want to survive (Venturini, 2021), the “black box” of Instagram’s recommendation algorithm encourages users to engage in algorithmic self-optimisation practices in order to increase their reach and influence (Bishop, 2018). For example, experimenting with thematically related hashtags, that, evidently, helps triangulate conspiracists into the same digital spaces and encourages disparate ideas to coalesce. As such, the smoothie is at its most nutritious under the titular #thegreatreset because anti vaxxers and 5G’ers, despite different genealogies, are comfortable bedfellows in the attention economy (Fairchild, 2007; Marwick, 2015) of Instagram, which privileges eyeballs, followers and likes over factual accuracy. Moreover, conspiracists are faced with an increasingly aggressive and ambiguous platform that seems to deplatform influencers before they gather notable momentum. While this may merely be platforms playing politics and paying lip service to calls for tighter regulation on disinformation, we infer that whilst deplatforming derails the influencer, it also galvanises the conspiracist by confirming their suspicions that a powerful elite is curtailing freedom of speech and orchestrating what the public is allowed to know. To epitomise this ongoing struggle with the platform, we note influencer content on Instagram can have a desperate tone, with wannabe influencers not so much influencing, but pleading for support and solidarity. When one conspiracy entrepreneur posts the same post four times in a day—incurring very few likes—it reads less like a signal of resistance and closer to a sign of resignation. As such, whilst the characteristics of a ‘secondary orality’ were on show, there was little to support the notion that repetition of content brings gravitas. The study of conspiracy theories is an epistemological challenge because it is against one’s cognitive conditioning to not interpret conspiratorial ideas as irrational and turn to debunking efforts. Harambam (2021) convincingly argues that discrediting conspiracy theories is not the role of social science because debunking is 1) unlikely (contra-facts have a tendency to be reappropriated as a evidence of a larger conspiracy), 2) unacademic (the role of the human ethnographer is to go “native” and try to understand a worldview from within) and 3) unproductive (counter arguments may dissociate the conspiracy community further and intensify the existing bifurcation of knowledge). With this in mind, our research has resisted making value judgements on the conspiracy theories we encountered and we recommend deeper qualitative research, specifically with the influencers behind the accounts, to really get under the hood of how instagram encourages conspiratorial interlocutors to converge. However, it seems it is very difficult to be a conspiratorial influencer sustaining a level of prominence over time due to platforms’ ongoing attempts to eliminate disinformation online, which puts those at the centre of conspiratorial networks on the top of content moderators’ hit lists; thus conspiratorial influencer status can be characterised as fleeting and ephemeral.

Digital Visuality of the Great Reset

Main contributors: Flavia Saxler, Megan Leal Causton, Maurice Dharampal

Introduction

Conspiracy theories are on the rise and its research has taken two major directions to deal with the spread of misinformation (Venturini, 2021). One strand deals with the computational detection of fakery and the other with interpretations of pseudo-theories. However, both miss to deal with the key features of conspiracism and their popularity. It is the very nature of for once big data and new communicative forms arising with it that rely on old and new narratives and practices. Venturini proclaims that “if online conspiracism seems hollow, it is because its core is far removed from the kind of analytic thinking we use to question it.” Our project starts with this thought that we need to rethink our methods to deal with data that - at least for us - seems messy, nonsense and random. Perhaps it is our very positionality as researchers who cannot access the field without the proper methods that shape the production of knowledge.

Hence, in our project we looked into the digital visuality of Instagram images of The Great Reset. We investigated the relationship between Venturini’s notion of digital orality to digital visuality. Venturini coins digital orality as formalised patterns of communication that are flimsy and repetitive and often appear to be mundane and hence not of - at least academic - importance. Venturini puts forward digital orality as a trope to understand conspiracy theories through their attentive structure and flagrancy. He thereby relates to McLuhan ’s (2004/1964) and Innis (1986/1950) historical approaches to understand media through their medium. Orality thereby originally referred to pre-digital ages, where culture transferred through word of mouth. Oral culture is shaped by repetitiveness and its relation to popular culture and folklore. Ong’s (1982) notion of secondary orality refers to a post-literate orality through electronic media technologies that bring back cultural practices of the past. Secondary orality is based on formalised patterns, of rites and passages that travel through communicative practices, those are easy and accessible and refer to popular culture.

Papacharissi (2015) uses the term of digital orality to show how new modalities of storytelling rely on the affordances of the digital but also referencing practices of the past. The latter relates to Jenkins (2006) work on ‘convergence culture’, namely, how new and old media collide and afford new forms of knowledge production. Papacharassi (2015) puts forward that “big data'' may not change the definition of knowledge, but they do modify how we communicate knowledge” (p.2).

The project itself is based on a large data set of 30 thousand images relating to conspiracy theories. It affords narratives of knowledge by weaving image assemblages together for understanding broader themes of conspiracy and their subsequent spaces. Our sub-project seeks to answer the following questions:

1. What are the visual tropes of images circulated relating to #newworldorder
2. How can we trace those tropes back via reverse-image-search and what does it tell us about its digital visuality?
3. How does digital visuality relate to digital orality?

In the next section we shine a light on how to understand a large data set of images relating to the great reset. We reflect on our epistemic standpoint, the data set, and propose to bring back Law’s (2004) thought on method of assemblage to deal with ambiguous data surrounding discourses of conspiracism.

Methodology

Our methodology follows Law’s (2004) criticism on social science to look for dominant narratives and to not deal with the messiness of data on that basis . He asked in 2004, a time pre-social media, how might methods deal with mess? And how to represent messiness without looking for dominant patterns and lines, but embracing its ambiguity? Hence, we argue that we need to embrace data ambiguity for understanding the complexity of digital experiences and not try to find a linear story (a similar experience even to that of an everyday social media user). Data is not as neutral and objective–methods and data do not simply describe social realities, but recreate them. D’Ignazio and Klein (2020) put forward how data visualisation always relates to power and often dismisses to deal with the unknown, the mess, the margins and peripheries.

Law (2004) calls this a “method assemblage” (Coleman & Ringrose, 2013; Lury & Wakeford, 2013). Methods do not only describe but also produce and enact the realities they seek to understand (Law, 2004). Law’s (2004) principle not to clean up data but to embrace its messiness. He argues that data should be tackled in its fluidity and multiplicity to embrace the modern experience of the world that is ambiguous and ambivalent.

Our sampling is intuitive and highlights the randomness of human intelligence, we navigated through images with the PixPlot provided by our wider team using the top hashtags that relate to the great reset (#newworldorder) while also keeping in mind our group’s division of the main theme into the subclusters Globalism, Covid-19, Technocracy and Dictatorship. From this we followed Law’s three-folded principle of crafting and bundling relations:

: “(a) whatever is in-here or present (for instance a representation or an object); (b) whatever is absent but also manifest (that is, it can be seen, is described, is manifestly relevant to presence); and (c) whatever is absent but is Other because, while necessary to presence, is also hidden, repressed or uninteresting.” (Law,2004, p. 161) We crafted relations through the digital practice of clicking, zooming in and out to see what appears, and what images stick and ‘prick’ (Barthes, 1981). ‘Pricking’ of images refers to when a photograph affects one personally and elicits an immediate, visceral response. This pricking is a vulnerable experience that constructs knowledge about the viewed (Page, 2017). Through discussions and debates on what appeared to us as very present, as absent and as uninteresting to see where we meet and diverge. On the basis of this we decided on forty top images that should represent the great reset. Furthermore, we tracked narratives through a manual reverse-image search to show convergence in such that images are reconfigured and recontextualised in the digital space. Through reverse-image-search we are able to highlight how visually similar images appear in different spaces and discourses and how they are re-fabricated through reverberations in the digital space.

Furthermore, we tracked narratives through a manual reverse-image search to show convergence in such that images are reconfigured and recontextualised in the digital space. Through reverse-image-search we are able to highlight how visually similar images appear in different spaces and discourses and how they are re-fabricated through reverberations in the digital space.

Thereby, we were aware of the basis of our knowledge production. Our view is partial and constructed from a specific point of view and on the basis of specific skills. We are both researchers that identify as female, which can be an asset to understand discourses of conspiracy that are predominantly shaped by white males. Furthermore, we are both qualitative researchers with a lack of skills in digital methods, which pushed us to think more creatively of how to deal with a large data set without mining and analysing the data through digital tools and methods. The latter can be seen as a limit, however, is an important mention to show the cracks and limits of knowledge production.

Findings and discussion: Visual similarity of The Great Reset and visual fabrications

Fig. 26 Top forty images compiled in the research assembled alongside some of the results of the reverse-image-search queries

Our overarching major finding is that digital visuality relates to digital orality in such a way that it is repetitive through aesthetic codes (white font, short sentences, colorful comics), is memorable (iconic images, famous people) and shocking (absurdity, war images, WW2 references, historic images).

Fig. 27 Images containing religious connotations connected to our sub-clusters of Globalism, Covid-19, Technocracy and Dictatorship and are indicative of other patterns observed

One observation which became an apparent pattern within images which contained or represented religious connotations. This assemblage of images could be further divided into subgroupings that were based on related themes, their aesthetics and narrative. Two of the most significant were based on a theme of nature or “stripping back” society back to the basics provided by religious beliefs. This is seen in references to the Garden of Eden–specifically with sin and forbidden fruit. This was observed frequently in the PixPlot in the circulation of various stock imagery related to toxicity, the government and Covid-19. On the other extreme there was a presence of influencer-esque fruit imagery which varied from advertising its purity alongside religion and conspiracy. The second is the juxtaposition of Covid-19 versus “God’s nature”. This coincides with a narrative of a religious reset that is forecast, however what is interesting is the presence of memes that contrast each other (either/or’s) based on government regulations during Covid-19. The features of digital orality that Venturini (2021) highlights in online communication as hashtags to bundle information, image templates to communicate information easily and video challenges that have a specific formalised pattern shows ritual conditions of communication. Those patterns also show in the digital visuality: the usage of meme structure and font, the use of famous people for remembrance, the use of iconic images with new writing to refer to the past and inscribe the present. As Venturini mentions, conspiratorial narratives have two crucial features that relate them to secondary orality - repeatability and memorability. Both features show in visual tropes in the great reset - similar images appear and images that prick and relate to discourses seen before. The latter we detected through reverse-image-search. Those images do not exist in a vacuum but travel through the digital web and are shaped by new experiences, thoughts and bits of information.

Conclusion

Through this intuitive and manual project we are able to show an imminent critique on how methods are crafting and producing realities. We put forward to critically reflect how we deal with data, what the data represents and how our very own methodological approaches produce knowledge that are not unbiased but come from specific standpoints, experiences, and skills. Qualitative approaches to big data are often neglected in times of optimism for artificial intelligence. However, we conclude that qualitative approaches to data need to be valued and further can set a path for assemblage methods.

For further analysis, Venturini’s (2021) proposal to integrate ethnography with computational methods seems to be vital to understand attention regimes of digital orality. Furthermore, techniques of ethnography allow a deep understanding of the field site through living within it. A worthwhile analysis to undertake would be a network analysis based on our preliminary common assumptions by breaking down each theme found and visualise its trajectory and its common connectors which evolve these narratives and aesthetics which now end up monopolising social media. This would allow for a more concrete understanding of the assemblage of the great reset and surrounding conspiracies.

Overall conclusion

This project proposed and enacted a ‘data hermeneutical’ approach to the study of online conspiracy narratives. We thereby empirically tested two hypotheses relevant to the study of online conspiracy theories in the humanities.

Our first hypothesis concerned the emergence of so-called ‘diagonalist movements’ that cut across traditional left/right distinctions (Slobodian & Callison 2021). In order to test this hypothesis, we combined ‘close’ and ‘distant’ readings of a curated list of keywords from the domains of critical theory, new age, epistemology (words related to truth and knowledge), and social justice terminology. Our analyses of the twitter data revealed a series of terms that are being pulled on by actors across the political spectrum, thus supporting the hypothesis of ‘diagonalism’.

Related to this first hypothesis, our second hypothesis was that the coronavirus pandemic was a driver behind the convergence and integration of conspiracy theories into a so-called ‘conspiracy singularity’. We tested this hypothesis by looking for evidence of convergence and integration on the level of Instagram hashtags, the co-occurrence of named entities in Instagram posts, and the analysis of images.

Comparing our Twitter and Instagram datasets offers the possibility of a tentative cross platform analysis of “conspiracy theory” in the time of the pandemic. It would appear that in both Twitter and Instagram political discussions of Covid-19 are markedly diagonal. Our research into Twitter demonstrated this pattern by looking at how different communities shared abstract concepts, through using the political compass to locate different communities in a two dimensional matrix (left/right, libertarian/authoritarian). In contrast to common research practice, our approach with twitter was thus to look into posts’ content as opposed to their hashtags.

On Instagram we also focussed on posts textual content as well as their visual content and their hashtags. There our findings also support the hypothesis of diagonalism in political discussions of Covid-19. Additionally we identified significant dynamism in those discussions which featured both integration and convergence leading over the course of 2020 to the emergence of what Naomi Klein has called the Great Reset conspiracy smoothie. This stood in contrast with the QAnon conspiracy theory which remained remarkably stable across 2020, its issues having relatively little to do with the pandemic.

References

Abidin, C. (2015). Communicative ❤️ Intimacies: Influencers and perceived
interconnectedness. Ada: A Journal of Gender, New Media, and Technology, (8)
doi://dx.doi.org/10.7264/N3MW2FFG

Banet-Weiser, Sarah (2018). Empowered. Popular Feminism and Popular Misogyny. Duke University Press.

Bishop, S. (2018). Anxiety, panic and self-optimization: Inequalities and the YouTube
algorithm. Convergence: The International Journal of Research into New Media
Technologies, 24(1), 69-84. doi:10.1177/1354856517736978

Blaskiewicz, Robert. "The big Pharma conspiracy theory." Medical Writing 22.4 (2013): 259-261.

Buenting, J., & Taylor, J. (2010). Conspiracy theories and fortuitous data. Philosophy of the Social Sciences, 40(4), 567-578.

CCDH (2021) The Disinformation Dozen: Why platforms must act on twelve leading online anti-vaxxers https://www.counterhate.com/disinformationdozen

Dahlgren, P. (2018) Media, Knowledge and Trust: The Deepening Epistemic Crisis of Democracy, Javnost - The Public, 25:1-2, 20-27, DOI: 10.1080/13183222.2018.1418819

De Coninck, David, et al. "Beliefs in conspiracy theories and misinformation about COVID-19: Comparative perspectives on the role of anxiety, depression and exposure to and trust in information sources." Frontiers in psychology 12 (2021).

Douglas Karen M. (2021). COVID-19 conspiracy theories. Group Processes & Intergroup Relations. 2021;24(2):270-275. doi:10.1177/1368430220982068

EUvsDisinfo 2020. EEAS Special Report: Disinformation on the coronavirus – short assessment of the information environment. March 19, 2020

https://euvsdisinfo.eu/eeas-special-report-disinformation-on-the-coronavirus-short-assessment-of-the-information-environment/ (accessed 2021-07-15)

Fairchild, C. (2007). Building the authentic celebrity: The "idol" phenomenon in the attention economy. Popular Music and Society, 30(3), 355-375. doi:10.1080/03007760600835306

Farkas, J., Schou, J., & Neumayer, C. (2018). Platformed antagonism: racist discourses on fake Muslim Facebook pages. Critical Discourse Studies, 15(5), 463-480.

Fenster, Mark 1999. Conspiracy Theories: Secrecy and Power in American Culture. Minneapolis: University of Minnesota Press.

Frenkel et al. (2020). How the ‘Plandemic’ Movie and Its Falsehoods Spread Widely Online. NY Times https://www.nytimes.com/2020/05/20/technology/plandemic-movie-youtube-facebook-coronavirus.html

Gray, M. (2010). Revisiting Saddam Hussein's political language: The sources and roles of conspiracy theories. Arab Studies Quarterly, 32(1), 28-46.

Harambam, J. (2021). Against modernist illusions: why we need more democratic and constructivist alternatives to debunking conspiracy theories, Journal for Cultural Research, 25:1, 104-122, https://doi.org/10.1080/14797585.2021.1886424

Hootsuite. (2021). 22 simple ways to increase instagram engagement (free calculator). Retrieved from https://blog.hootsuite.com/instagram-engagement/

Ingenhoff, D., Calamai, G., & Sevin, E. (2021). Key influencers in public diplomacy 2.0: A country-based social network analysis. Social Media + Society, , 1-12. doi:10.1177/2056305120981053

Katz, E., & Lazarsfeld, P. F. (1955). Personal influence: The part played by people in the flow of mass communications. Glencoe, IL: Free Press.

Klein, Naomi (2018). The Great Reset Conspiracy Smoothie. The Intercept. https://theintercept.com/2020/12/08/great-reset-conspiracy/ (accessed 2021-07-15)

Latour, Bruno. "Why has critique run out of steam? From matters of fact to matters of concern." Critical inquiry 30.2 (2004): 225-248.

Marres, N. (2015). Why map issues? On controversy analysis as a digital method. Science, Technology, & Human Values, 40(5), 655-686.

Marwick, A. E. (2013). Status update. New Haven: Yale University Press.

Marwick, A. E. (2015). Instafame: Luxury selfies in the attention economy. Public Culture, 27(1), 137-160. doi:10.1215/08992363-2798379

Mason, J. (2021). How the extremist organization disseminates vaccine misinformation through social media. First Draft. February 18, 2021. https://firstdraftnews.org/articles/the-nation-of-islam-and-anti-vaccine-rhetoric/ (accessed 2021-07-20)

Melley, Timothy. The Covert Sphere. Cornell University Press, 2012.

Morrish, L. (2020) How QAnon content endures on social media through visuals and code words. First Draft. December 3, 2020. https://firstdraftnews.org/articles/how-qanon-content-endures-on-social-media-through-visuals-and-code-words/ (accessed 2021-07-20)

Norton, R. (2014). Agenda 21 and Its Discontents: Is Sustainable Development a Global Imperative or Globalizing Conspiracy? The Urban Lawyer, 46(2), 325-360. Retrieved July 14, 2021, from http://www.jstor.org/stable/24392808

Rogers, R. (2021). Visual media analysis for instagram and other media platforms. Big Data & Society, 2021(January-June), 1-23. doi:10.1177/20539517211022370

Rosenblum, N. L., & Muirhead, R. (2020). A Lot of People Are Saying. Princeton University Press.

Roose, K. (2020). How ‘Save the Children’ Is Keeping QAnon Alive. The New York Times. https://www.nytimes.com/2020/09/28/technology/save-the-children-qanon.html (accessed 2021-07-15).

Sapountzis, A., & Condor, S. (2013). Conspiracy accounts as intergroup theories: Challenging dominant understandings of social power and political legitimacy. Political Psychology, 34(5), 731-752.

Slobodian, Quinn & William Callison. 2021. Coronapolitics from the Reichstag to the Capitol. Boston Review http://bostonreview.net/politics/william-callison-quinn-slobodian-coronapolitics-reichstag-capitol (accessed 2021-07-13)

Starbird, Kate. "Information wars: A window into the alternative media ecosystem." Medium.(Mar. 2017) (2017).

Tuters, Marc, and Sal Hagen. "(((They))) rule: Memetic antagonism and nebulous othering on 4chan." New Media & Society 22.12 (2020): 2218-2237.

Venturini (2021, forthcoming).

Ward, Charlotte, and David Voas. "The emergence of conspirituality." Journal of Contemporary Religion 26.1 (2011): 103-121.

Zhu, Linhong & Lerman, Kristina. (2016). Attention Inequality in Social Media. Association for the Advancement of Artificial Intelligence. arXiv:1601.07200 [cs.SI]

Zuboff, S. (2020). Caveat Usor: Surveillance Capitalism as Epistemic Inequality. In K. Werbach (Ed.), After the Digital Tornado: Networks, Algorithms, Humanity (pp. 174-214). Cambridge: Cambridge University Press.