From enduring rumours around Jeffrey Epstein’s death to beliefs that vaccines cause autism, conspiracy theories are a common part of the daily discourse, especially on online platforms like TikTok and Reddit. Typically dismissed as delusional or resulting from personality flaws, such theories are not a sign of poor judgement but actually an attempt to regain control at times of stress. 

That's according to Henrich Greve, Professor of Entrepreneurship at INSEAD, speaking on the INSEAD Knowledge podcast about his new book, Ctrl+Alt+Doubt: Decoding the Language of Online Conspiracy Talk. Co-authored with Stanford University's Huggy Rao, the book began as a response to try and find ways to curb the spread of harmful conspiracy theories during the Covid19 pandemic.

Made up of a "bricolage" of familiar cultural fragments, such theories are put together as a form of problem solving: the story fits events that the creators are already observing. Constructed from a shared language, they offer a means for someone to try and make sense of an uncertain world. 

Interestingly, the most impactful theories are those that use neutral language. Convincing conspiracy talk reads as rational, not angry, since it is presented as a considered explanation, not an emotional outburst. What’s more, an individual often holds several overlapping theories, which can act as a hedge. If one is debunked, the others remain intact, which makes the whole belief system harder to shift.

The problem is that such flights of fancy are not harmless. Vaccine conspiracy theories have contributed to the return of measles and cuts to mRNA research funding in the United States, while state-sponsored disinformation has fuelled violence and undermined trust in elections. 

According to Greve, there is no single fix when it comes to trying to limit the spread of such harmful theories. Counterarguments, priming or inoculation have some effect, but their impact is inconsistent and conspiracy theories are notoriously resistant to fact checking. 

The more promising route lies with the platforms where these stories are spread, using language analysis to identify bots and “superspreaders” early, and prompting people to pause before they share something questionable, a reminder that what gets posted has real consequences.