The "digital egregore" names the impression that an online crowd behaves like an autonomous entity — an emotion seizing thousands of accounts, a designated target, an outrage that seems to live and swell of its own accord. This experience is real, but it requires no supernatural entity: social media does not create a collective mind, it accelerates documented human mechanisms — emotional contagion, the viral diffusion of moral outrage, and algorithmic amplification. Networks thus revive an old occult intuition without proving it right.
Digital egregore: a contemporary application of an esoteric term (from the Greek egrêgoros, "the watcher / the awakened") to an online collective that appears endowed with a will and an emotion of its own. In the literal sense, no such entity exists. The word in fact names only a cluster of measurable collective dynamics — emotional contagion, the virality of moralised content, the hardening of group boundaries — amplified by the architecture of platforms.
In other words, the occult part of the term remains a symbolic reading, never a scientific fact. What is documented are the online phenomena themselves: they are measured on large corpora of messages, published in peer-reviewed journals, and need no ghost to be explained.
The idea of a crowd that thinks and wills as a single body was not born with the internet. As early as 1895, in The Crowd (Psychologie des foules), Gustave Le Bon described the individual dissolved into the mass as gripped by a "collective soul" that lowers critical sense and renders him suggestible. His book, dated and often criticised for its assumptions, nonetheless fixed the vocabulary from which esotericism would later draw the egregore. Where Le Bon saw a psychological metaphor, occultism reified an entity; network science, in turn, ended up measuring the real mechanisms these images covered.
Three properties of platforms feed the impression of a collective "being." First scale: thousands of reactions converge within minutes, as if a single voice were speaking. Then speed: the dynamic seems to escape each participant, who undergoes it as much as he contributes to it. Finally visibility: one sees others' emotion in real time, which invites synchronisation with it. Combined, these three properties produce the appearance of a purpose — "the crowd wants," "the crowd punishes" — when it is only connected individuals reacting to one another.
This is exactly the point the occult tradition reified. Where the occultist sees an entity "fed" by the attention of its faithful, network analysis sees an emergent phenomenon: a real collective effect, endowed with properties no single member possesses, but with no single author and no central will. Confusing "emergent" with "supernatural" is precisely the error that lucidity allows one to avoid.
The first mechanism is emotional contagion: the human tendency to synchronise one's emotions with those of others, through mimicry of expressions, posture, and tone. It has long been documented in psychology, notably by Elaine Hatfield, John Cacioppo, and Richard Rapson. Offline it presupposes physical presence; online, the text, the image, and the staging of others' emotion suffice to trigger it at a distance.
That this contagion does operate over a network, and without physical contact, has been tested at very large scale. In 2014, Adam Kramer, Jamie Guillory, and Jeffrey Hancock published in PNAS an experiment conducted on nearly 690,000 Facebook users: by reducing the share of positive or negative content in the news feed, they observed a slight corresponding shift in the emotions users subsequently expressed. The effect is small, and the study sparked a sharp ethical controversy over the absence of informed consent — but it establishes that an emotional state can be transmitted through a feed alone, without direct interaction.
More broadly, the work of Nicholas Christakis and James Fowler on large social networks (synthesised in their book Connected, 2009) showed that states such as happiness or certain behaviours propagate along social ties. These results are debated — it remains statistically difficult to distinguish true contagion from homophily (already resembling one's contacts) — but they establish a solid point: our emotions and conduct are never purely individual, they circulate within the fabric of our relationships. The digital network only widens and accelerates this fabric.
The second mechanism is the preferential diffusion of content charged with moral emotion. The study by William Brady and colleagues, published in 2017 in PNAS, analysed a large corpus of messages on divisive political topics. Documented result: each word carrying both a moral and an emotional charge added to a message significantly increased its re-diffusion. Outrage, in short, travels better than the cold argument.
The psychologist Molly Crockett extended this analysis in Nature Human Behaviour (2017) by describing how the architecture of platforms transforms moral outrage. Immediate social rewards — likes, shares, visibility — and the ease of expressing indignation can reinforce and routinise this behaviour, to the point of altering collective dynamics relative to the offline world. The mechanism is documented; here too it calls for no entity to be understood.
To this moral mechanism is added a more general fact of virality: the false often circulates better than the true. In 2018, Soroush Vosoughi, Deb Roy, and Sinan Aral published in Science the analysis of about 126,000 information cascades spread on Twitter from 2006 to 2017. False news there spread farther, faster, deeper, and more broadly than the true, especially in politics — an effect the authors link to its charge of novelty and the emotions it arouses. The "life of its own" attributed to an online rumour is therefore, here too, the product of measurable properties of the content and the network.
These studies converge on a single conclusion: the "atmosphere" that seems to seize a thread is not a group soul, it is the product of measurable forces — the contagion of emotions between connected individuals, the preferential diffusion of the most morally charged content, and the heightened virality of what is new or false.
Beyond individuals, the selection performed by platforms weighs on what circulates. The popular image of the "filter bubble," popularised by Eli Pariser in 2011, holds that the algorithm locks us into our opinions. The measured reality is more nuanced. In 2015, Eytan Bakshy, Solomon Messing, and Lada Adamic published in Science a study covering more than 10 million Facebook users: algorithmic ranking did reduce exposure to politically opposing content, but to a moderate extent, and individual choices — whom one follows, what one clicks on — weighed more than the algorithm itself.
The lesson is twofold, and keeps the egregore at bay. Yes, the architecture of platforms steers the circulation of content and can favour echo chambers as well as emotional pile-ons. No, it is not a central will that thinks in our stead: it works in concert with our own choices. The collective effect arises from the encounter between a technical design and human behaviour, never from an autonomous mind.
The phenomenon popularly named the "online pile-on" (coordinated harassment) is the case where the digital egregore appears most real — and where it dissolves most clearly under analysis. It presents precisely the traits the tradition ascribed to the collective entity, each corresponding to a known mechanism:
The lesson is practical more than metaphysical: one does not soothe a ghost, but one can act on mechanisms. This runs through the design of platforms (reducing the rewards of outrage, slowing re-diffusion), through attention education, and through individual awareness of one's own suggestibility. Before relaying outraging content, a single question often defuses the pile-on: am I sharing it because it is true and useful, or because it is contagious?
Does a digital egregore really exist? Not as an entity: no scientific evidence attests to the existence of an autonomous collective mind online. The phenomena the word tries to name, however — emotional contagion, viral outrage, the coordinated pile-on — are quite real and measured by social psychology and network science.
Why does outrage go viral on social media? Because content with a strong moral and emotional charge is re-shared more (Brady et al., PNAS 2017), because the architecture of platforms socially rewards the expression of outrage (Crockett, Nature Human Behaviour 2017), and because the false and the novel circulate faster than the rest (Vosoughi et al., Science 2018). It is not a group soul, but a combination of human psychology and technical design.
How does one avoid being swept up by an online pile-on? By spotting the mechanisms in real time: distinguishing one's own emotion from the surrounding contagion, watching the tipping into "us against them," cutting the sources of amplification (notifications, feed), and slowing down before sharing outraging content. One acts on dynamics, never on a ghost.
This article covers the digital angle of a broader investigation into collective dynamics. For the full dissection of the notion — from Le Bon's crowd to Tajfel's group identity, with a strict mapping of levels of certainty — read the pillar article, then the TOME 1 investigation, The Egregore (sourced EPUB and PDF, DRM-free), the starting point of the L'Appel de l'Éveil collection.
Not as an entity: no scientific evidence attests to the existence of an autonomous collective mind online. The phenomena the word tries to name, however — emotional contagion, viral outrage, the coordinated pile-on — are quite real and measured by social psychology and network science.
Because content with a strong moral and emotional charge is re-shared more (Brady et al., PNAS 2017), because the architecture of platforms socially rewards the expression of outrage (Crockett, Nature Human Behaviour 2017), and because the false and the novel circulate faster than the rest (Vosoughi et al., Science 2018). It is not a group soul, but a combination of human psychology and technical design.
By spotting the mechanisms in real time: distinguishing one's own emotion from the surrounding contagion, watching the tipping into "us against them," cutting the sources of amplification (notifications, feed), and slowing down before sharing outraging content. One acts on dynamics, never on a ghost.
Dossier : Égrégore & souveraineté cognitive
Accueil · Collection · Journal · Dossiers · Sources