Can we develop our intuition to counter misinformation?

Pro­pa­gan­da is a glob­al strat­e­gy put in place by a state, an insti­tu­tion or a com­mu­ni­ty to desta­bilise a tar­get. Mis­in­for­ma­tion is the unin­ten­tion­al shar­ing of fake news, erro­neous or obso­lete infor­ma­tion, through error, or lack of vig­i­lance or knowl­edge of the sub­ject: there is no actu­al inten­tion to cause harm here. Dis­in­for­ma­tion, on the oth­er hand, is a pro­pa­gan­da tool that works by delib­er­ate­ly cre­at­ing and shar­ing false infor­ma­tion with the inten­tion of caus­ing harm. This arti­cle focus­es on dis­in­for­ma­tion because, in addi­tion to the ques­tion of the truth of infor­ma­tion, this con­cept rais­es the issue of verac­i­ty, and there­fore of trust in infor­ma­tion sources. We will argue that com­bat­ing mis­in­for­ma­tion means ask­ing three ques­tions about knowl­edge: what to trust, how to trust and who to trust.

The rise of dis­in­for­ma­tion on social net­works is gen­er­at­ing infor­ma­tion­al chaos that threat­ens demo­c­ra­t­ic life: sat­u­ra­tion of auto­mat­ed adver­tis­ing and har­vest­ed data, pro­mo­tion of shock­ing and con­spir­a­cy-ori­ent­ed infor­ma­tion, dis­cred­it­ing of author­i­ty fig­ures, algo­rith­mic log­ic at the source of thought bub­bles. “For exam­ple, 120,000 years’ worth of videos are viewed on YouTube every day. Of these, 70% are watched because the platform’s arti­fi­cial intel­li­gence rec­om­mends them”¹. Social net­works have also come to be seen as one of the most reli­able means of con­sult­ing the news². The mis­in­for­ma­tion of young peo­ple in par­tic­u­lar sends out wor­ry­ing sig­nals: one in 4 young French peo­ple sub­scribe to cre­ation­ist the­o­ries, 16% think that the earth could well be flat, 20% that the Amer­i­cans have nev­er been to the moon, and 49% that astrol­o­gy is a sci­ence. A large pro­por­tion of them believe that an influencer’s pop­u­lar­i­ty is a guar­an­tee of reli­a­bil­i­ty (rep­re­sen­ta­tive sam­ple aged between 18 and 24)³. Trust in sci­ence is strong and sta­ble in all Euro­pean coun­tries, except in France where it has fall­en by 20 per­cent­age points in 18 months⁴. This drop in con­fi­dence in sci­ence is cor­re­lat­ed with sup­port for fake news and con­spir­a­cy the­o­ries⁵. At the same time, illec­tro­n­ism (a con­trac­tion of illit­er­a­cy and elec­tron­ics) is cre­at­ing a new area of exclu­sion, with 14 mil­lion French peo­ple expe­ri­enc­ing dif­fi­cul­ties in using dig­i­tal tools, at a time when dema­te­ri­al­i­sa­tion is becom­ing wide­spread⁶.

These vul­ner­a­bil­i­ties, com­bined with pow­er­ful forces of influ­ence, have dam­ag­ing effects on our democ­ra­cies: reduced crit­i­cal think­ing and creduli­ty on the part of cit­i­zens, inabil­i­ty to resist seduc­tion by, and sup­port for, dubi­ous ideas, selec­tive expo­sure to infor­ma­tion and preva­lence of hypoth­e­sis-con­fir­ma­tion bias, dichoto­mous think­ing and reduced abil­i­ty to make argu­ments⁷. Admit­ted­ly, these flaws are noth­ing new (cf. Orson Welles’ radio hoax “War of the Worlds”), but the infil­tra­tion of supra-nation­al pow­ers, the pow­er of tech­no­log­i­cal tools and the avail­abil­i­ty of our slum­ber­ing brains make this a crit­i­cal risk.

The levers for com­bat­ing dis­in­for­ma­tion and mis­in­for­ma­tion are there­fore a pri­or­i­ty for our democ­ra­cies. They fall into two dis­tinct cat­e­gories: lim­it­ing the pro­duc­tion and dis­sem­i­na­tion of fake news, and lim­it­ing its impact.

350,000 mes­sages are post­ed on X (for­mer­ly Twit­ter) every minute, for 250 mil­lion active users. There are an esti­mat­ed 2,000 mod­er­a­tors, i.e. one mod­er­a­tor for every 175,000 users⁸. The same infla­tion is observed for oth­er social net­works. These fig­ures call into ques­tion the very pos­si­bil­i­ty of mod­er­at­ing infor­ma­tion, which is increas­ing­ly man­aged by algo­rithms, a black box whose trans­paren­cy is often ques­tioned⁹. Elon Musk, via his com­pa­ny X, filed a suit against Cal­i­for­nia on 8 Sep­tem­ber 2023, accus­ing the Amer­i­can state of hin­der­ing free­dom of expres­sion by forc­ing plat­forms to be trans­par­ent about con­tent moderation.

To be a sci­en­tist is to fight one’s brain

Copié !

Legal reg­u­la­tion (ARCOM, DSA) is now being debat­ed, and polit­i­cal insti­tu­tions are tak­ing up the issue, but the bal­ance between free­dom of expres­sion and cen­sor­ship has not yet been achieved. In France, the Autorité de Régu­la­tion de la Com­mu­ni­ca­tion Audio­vi­suelle et Numérique (ARCOM) acts effec­tive­ly but remains lim­it­ed in terms of resources, since it has 355 employ­ees work­ing on a wide range of issues (pro­tec­tion of audi­ences, media edu­ca­tion, respect for copy­right, infor­ma­tion ethics, super­vi­sion of online plat­forms, devel­op­ments in radio and dig­i­tal audio, VOD dis­tri­b­u­tion). With the Dig­i­tal Social Act, Europe is putting a sys­tem of account­abil­i­ty in place for the major plat­forms as from 2024, based on a sim­ple prin­ci­ple: what is ille­gal offline is ille­gal online. The aim is to pro­tect Inter­net users by a num­ber of prac­ti­cal means: mak­ing the way in which the rec­om­men­da­tion algo­rithm works acces­si­ble to users, as well as the pos­si­bil­i­ty of deac­ti­vat­ing it, jus­ti­fy­ing mod­er­a­tion deci­sions, set­ting up an explic­it mech­a­nism for report­ing con­tent, and allow­ing appeals. Cer­tain types of tar­get­ed adver­tis­ing will be banned. Penal­ties for non-com­pli­ant plat­forms are set to match the stat­ed ambi­tions: 6% of glob­al turnover.

The fact remains, how­ev­er, that if we take into account the vul­ner­a­bil­i­ties men­tioned above, the rapid growth in the amount of infor­ma­tion exchanged and the dif­fi­cul­ties in reg­u­lat­ing and mod­er­at­ing plat­forms, a com­ple­men­tary approach is need­ed: not just lim­it­ing mis­in­for­ma­tion, but reduc­ing its impact on its tar­gets by strength­en­ing their capac­i­ty to resist. But how do we know whether a piece of infor­ma­tion is true?

Epis­temic beliefs relate to the ideas we have about knowl­edge and the process­es by which knowl­edge is cre­at­ed: what makes us think we know things? What fac­tors con­tribute to a mis­per­cep­tion of knowl­edge? These ques­tions are cen­tral to under­stand­ing the spread and impact of mis­in­for­ma­tion, as well as ways of coun­ter­ing it.

Kel­ly Gar­rett and Bri­an Weeks, from Ohio and Michi­gan Uni­ver­si­ties, car­ried out a vast study in the Unit­ed States in 2017 with the aim of gain­ing a bet­ter under­stand­ing of some of the deter­min­ing fac­tors in adher­ence to mis­in­for­ma­tion and con­spir­a­cy the­o­ries. Ini­tial­ly, they mea­sured the opin­ions of par­tic­i­pants on con­tro­ver­sial sub­jects in cer­tain con­spir­a­cy net­works: the fact that the Apol­lo Mis­sion nev­er went to the moon, that AIDS was an inten­tion­al cre­ation to harm the homo­sex­u­al com­mu­ni­ty, that the 9/11 attacks were autho­rised by the US admin­is­tra­tion to jus­ti­fy polit­i­cal deci­sions (mil­i­tary inva­sion and reduc­tion of civ­il rights), or that JFK, Luther King or Princess Diana were assas­si­nat­ed on the orders of insti­tu­tions (gov­ern­ments or secret agen­cies). They also mea­sured par­tic­i­pants’ opin­ions on high­ly sen­si­tive con­tem­po­rary social issues where there is a counter-dis­course to the cur­rent sci­en­tif­ic con­sen­sus: the role of human activ­i­ty in glob­al warm­ing, or the fact that cer­tain vac­cines cause dis­eases such as autism.

This data was cor­re­lat­ed with oth­er mea­sures of the same par­tic­i­pants’ epis­temic beliefs. The results are indis­putable: par­tic­i­pants are more like­ly to sub­scribe to con­spir­a­cy the­o­ries and are more sus­pi­cious of sci­en­tif­ic dis­course the more:

• they trust their intu­itions to “feel” the truth of things,
• they believe that facts are not suf­fi­cient to call into ques­tion what they believe to be true,
• they con­sid­er that all truth is rel­a­tive to a polit­i­cal context.

Since this study, a great deal of research has shown the extent to which these three ele­ments con­sti­tute vul­ner­a­bil­i­ties in the fight against dis­in­for­ma­tion. In the next arti­cle, we will explain how each of these epis­temic beliefs works, in order to iden­ti­fy the psy­choso­cial skills we need to devel­op to sharp­en our crit­i­cal think­ing skills.

The first impor­tant result of the study by Kel­ly Gar­rett and Bri­an Weeks con­cerns the trust placed in our intu­ition to under­stand the world around us, with the strong idea that cer­tain truths are not acces­si­ble ratio­nal­ly. Instinct, first impres­sions and a dif­fuse “gut” feel­ing are said to be excel­lent indi­ca­tors for guid­ing our judge­ments and deci­sions. This epis­temic belief is wide­ly held today in main­stream pub­li­ca­tions and per­son­al devel­op­ment meth­ods: “enter the mag­ic of intu­ition”; “devel­op your 6th sense”; “man­age with intu­ition”; “the pow­ers of intu­ition”: These titles sup­port the idea that there is a “lit­tle je ne sais quoi” that allows us to access hid­den truths and under­stand the world direct­ly by “recon­nect­ing” with our­selves and our envi­ron­ment (the cos­mos, pseu­do quan­tum vibra­tions, etc.). Inspired by the New Age¹⁰, these approach­es, which often relate to health and well-being, do not shy away from advo­cat­ing a return to com­mon sense and our abil­i­ty to know things in an emo­tion­al way, with­out the need for proof, thanks to a “gift”. Yet sci­ence has often devel­oped con­trary to com­mon sense and first intu­itions: a heavy body does not fall faster than a light one, hot water freezes faster than cold water…

Admit­ted­ly, sci­en­tif­ic research does not ques­tion the role of intu­itive knowl­edge, and numer­ous works and pub­li­ca­tions are devot­ed to it¹¹, many of them in med­i­cine under the aegis of “Gut Feel­ing”¹². But what this cog­ni­tive sci­ence research says is very dif­fer­ent from what we find in per­son­al devel­op­ment books, pri­mar­i­ly because intu­ition is described as a form of rea­son­ing that is part of a fair­ly ratio­nal process. In fact, sci­en­tists have shown (with the help of empir­i­cal research car­ried out with pro­fes­sion­als who have devel­oped intu­itive knowl­edge, such as com­pa­ny direc­tors, doc­tors, fire­men, chess play­ers, sports­men and sol­diers) that intu­ition is all the more effec­tive in experts who have had a great deal of pre­vi­ous expe­ri­ence, thanks to oppor­tu­ni­ties to make hypothe­ses based on the analy­sis of their envi­ron­ment, to test them in a real sit­u­a­tion, to ben­e­fit from feed­back (suc­cess or fail­ure), to make cor­rec­tions, to retest, etc.… until they arrive at an implic­it, effi­cient and rapid know-how known as intu­ition. There’s noth­ing eso­teric or “quan­tum” about it, but prac­tice, dis­ci­pline, and feed­back¹³ enable us to make rapid deci­sions when the con­text demands it. If 82% of Nobel Prize win­ners acknowl­edge that their dis­cov­er­ies were made thanks to their intu­ition¹⁴, it is above all because they have accu­mu­lat­ed such a wealth of sci­en­tif­ic knowl­edge and method­olog­i­cal expe­ri­ence that they end up aggre­gat­ing clus­ters of clues to arrive at an insight: “eure­ka”!

The first psy­choso­cial skill to devel­op in the fight against mis­in­for­ma­tion is there­fore to dis­trust one’s own intu­itions by resist­ing one­self¹⁵: “to be a sci­en­tist is to fight one’s brain” said Gas­ton Bachelard. It’s not a ques­tion of sup­press­ing our intu­itions, but rather of tak­ing the nec­es­sary time to ques­tion them, audit them and val­i­date their basis, and in this way to car­ry out metacog­ni­tive work on our­selves in a non-indul­gent and mod­est way: on what past expe­ri­ence is my intu­ition based, have I had the oppor­tu­ni­ty to have a lot of feed­back on the effects of my actions linked to this intu­ition, and to what extent am I not being influ­enced by my desires, my emo­tions or my envi­ron­ment? This is all the more dif­fi­cult because an impres­sion is above all… impres­sive: what mat­ters most is not so much its con­tent as the men­tal process of its con­struc­tion and its con­se­quences on the way we think and act¹⁶.

The sec­ond impor­tant result of Kel­ly Gar­rett and Bri­an Weeks’ study relates to the impor­tance we attach to con­sis­ten­cy between facts and opin­ions. Put anoth­er way, can we main­tain a belief in the face of a demon­stra­tion that con­tra­dicts it? Some of us need fac­tu­al evi­dence to form an opin­ion, dis­trust appear­ances and are con­cerned about the method used to pro­duce data. Oth­ers not so much: the study men­tioned above shows that the lat­ter are much more like­ly to sub­scribe to false infor­ma­tion and con­spir­a­cy the­o­ries. We remem­ber the “alter­na­tive facts” the day after Trump’s elec­tion, symp­to­matic of the post-truth era. These strate­gies for dis­tort­ing real­i­ty are only pos­si­ble because they find an audi­ence who, while not being fooled by them, do not feel the need for con­sis­ten­cy between facts and beliefs. On the con­trary, the coher­ence they seek tends to adjust the facts in favour of their beliefs, a ratio­nal­i­sa­tion effect that is well known from work on cog­ni­tive dis­so­nance. Hugo Merci­er and Dan Sper­ber¹⁷have recent­ly exam­ined this issue in a book which defends the the­sis that our rea­son serves us above all… to be right, not only in rela­tion to oth­ers, but also in rela­tion to our­selves! Hence the cog­ni­tive bias­es with a self-jus­ti­fy­ing func­tion: hypoth­e­sis con­fir­ma­tion, anchor­ing, loss aver­sion, ret­ro­spec­tive bias, etc.¹⁸. It’s easy to see why com­bat­ing this is a daunt­ing­ly com­plex task, yet one that is both nec­es­sary and pos­si­ble if we make the effort to teach the sci­en­tif­ic method and its com­po­nents, and not just to stu­dents des­tined for sci­en­tif­ic careers! These alter­na­tive facts call into ques­tion the very notion of truth and knowl­edge recog­nised as fair¹⁹, and lead to the sor­did con­clu­sion that sci­ence is an opin­ion like any oth­er²⁰: this pos­ture under­mines the very foun­da­tions of our demo­c­ra­t­ic insti­tu­tions, which is why knowl­edge of the sci­en­tif­ic method has now become a com­mon good and a gen­uine psy­cho-social skill accord­ing to the WHO: “abil­i­ties that enable the devel­op­ment not only of indi­vid­ual well-being, but also of con­struc­tive social interactions”

The lat­est result from Kel­ly Gar­rett and Bri­an Weeks’ study shows that the more indi­vid­u­als believe that facts are depen­dent on the polit­i­cal pow­er in place or the socio-polit­i­cal con­text in which they are pro­duced, the more read­i­ly they adhere to dis­in­for­ma­tion and con­spir­a­cy the­o­ries. This type of epis­temic belief, which is res­olute­ly rel­a­tivis­tic, is facil­i­tat­ed by the fact that our beliefs also serve to rein­force our iden­ti­fi­ca­tions with the groups to which we belong: we eval­u­ate the infor­ma­tion to which we are exposed accord­ing to our socio-ide­o­log­i­cal prox­im­i­ty to its source. The under­ly­ing prob­lem here is there­fore one of verac­i­ty rather than truth: it is a ques­tion of the moral qual­i­ty of the author of a piece of infor­ma­tion and there­fore of the trust we place in him or her. Fran­cis Wolff²¹ shows that this rel­a­tivist stance is now a stum­bling block in the fight against the risks com­mon to human­i­ty as a whole (glob­al warm­ing, eco­nom­ic cri­sis, short­age of resources, extinc­tion of species, epi­demics, ter­ror­ism, etc.) because of local demands (iden­ti­ty-based, com­mu­ni­tar­i­an, nation­al­ist, xeno­pho­bic, reli­gious rad­i­cal­ism, etc.) that ham­per our abil­i­ty to engage in dia­logue and find ways of mov­ing for­ward col­lec­tive­ly. So what psy­cho-social skills do we need to devel­op if we are to know who we can trust and build com­mon projects that tran­scend com­mu­ni­tar­i­an divi­sions? To answer this ques­tion, Philippe Bre­ton²² car­ried out a num­ber of empir­i­cal stud­ies dur­ing exper­i­men­tal argu­men­ta­tion work­shops. His results sug­gest that we need to devel­op what he calls a “demo­c­ra­t­ic abil­i­ty”, which is cur­rent­ly far too lack­ing to build trust, and which is based on three skills:

• Speak­ing in front of oth­ers: prac­tis­ing over­com­ing the fear of speak­ing in front of an unfa­mil­iar group. Sci­en­tif­ic research shows that this fear is one of the most wide­spread among adults (55%²³). This fear hin­ders the very pos­si­bil­i­ty of estab­lish­ing the con­di­tions for cooperation.
• Cog­ni­tive empa­thy: prac­tis­ing defend­ing opin­ions con­trary to one’s own. The aim is to learn to iden­ti­fy the qual­i­ty of the argu­ments and thus reg­u­late one’s less sol­id epis­temic beliefs. This strat­e­gy is part of the psy­cho­log­i­cal inoc­u­la­tion meth­ods²⁴ designed to strength­en men­tal immunity.
• Com­bat “con­sen­su­al palaver”: soft con­sen­sus is a way of avoid­ing debate that gives the illu­sion of bring­ing peo­ple togeth­er. Prac­tis­ing “frank and peace­ful con­flict­ual­i­ty”²⁵ is not easy, but it does pro­vide the nec­es­sary demo­c­ra­t­ic vitality.

“Il faut voir comme on se par­le. Man­i­feste pour les arts de la parole” (We need to look at how we talk to each oth­er. A Man­i­festo for the Arts of Speech) is the title of the lat­est book by Gérald Garut­ti, founder of the “Cen­tre des Arts de la Parole”, a third-par­ty cen­tre that restores the psy­cho-social skills need­ed to build a space for shared dia­logue and com­bat the mis­in­for­ma­tion that is under­min­ing our democ­ra­cies. These third-par­ty cen­tres, spaces for sci­ence and dis­cov­ery, and cit­i­zens’ lab­o­ra­to­ries for exper­i­men­ta­tion, share the com­mon goal of devel­op­ing demo­c­ra­t­ic skills in the form of oper­a­tional know-how: know­ing how to argue and make a coun­ter­ar­gu­ment, know­ing how to lis­ten, sus­pend­ing one’s judge­ment and elic­it­ing that of oth­ers. They also help us to under­stand how sci­en­tif­ic truth is con­struct­ed and how this knowl­edge can be biased: these are the levers of free will and liv­ing together.