j?”eY?Z]>Q”?

36
1 Farkas, J., & Neumayer, C. (2017). ‘Stop Fake Hate Profiles on Facebook’: Challenges for crowdsourced activism on social media. First Monday, 22(9). ‘Stop Fake Hate Profiles on Facebook’: Challenges for crowdsourced activism on social media Johan Farkas* and Christina Neumayer** * IT University of Copenhagen, direct comments to [email protected]. ** IT University of Copenhagen, [email protected]. Abstract This research examines how activists mobilise against fake hate profiles on Facebook. Based on six months of participant observation, the article demonstrates how Danish Facebook users organised to combat fictitious Muslim profiles that spurred hatred against ethnic minorities. The article concludes that crowdsourced action by Facebook users is insufficient as a form of sustainable resistance against fake hate profiles. A viable solution would require social media companies such as Facebook to take responsibility in the struggle against fake content used for political manipulation. Keywords: online activism; crowdsourced activism; fake profiles; hate profiles; Facebook Introduction This is how we shut his page down. We’re nearly 1300 members and if we each spend 5 seconds reporting his page, it’ll be removed in no time. (Stop Fake Hate Profiles on Facebook, post by admin, 1 July 2015) Pre-print version

Transcript of j?”eY?Z]>Q”?

Page 1: j?”eY?Z]>Q”?

1

Farkas, J., & Neumayer, C. (2017). ‘Stop Fake Hate Profiles on Facebook’: Challenges forcrowdsourcedactivismonsocialmedia.FirstMonday,22(9).

‘StopFakeHateProfilesonFacebook’:Challengesforcrowdsourced

activismonsocialmedia

JohanFarkas*andChristinaNeumayer**

* ITUniversityofCopenhagen,[email protected].

**ITUniversityofCopenhagen,[email protected].

Abstract

Thisresearchexamineshowactivistsmobiliseagainstfakehateprofileson

Facebook.Basedonsixmonthsofparticipantobservation,thearticle

demonstrateshowDanishFacebookusersorganisedtocombatfictitious

Muslimprofilesthatspurredhatredagainstethnicminorities.Thearticle

concludesthatcrowdsourcedactionbyFacebookusersisinsufficientasaform

ofsustainableresistanceagainstfakehateprofiles.Aviablesolutionwould

requiresocialmediacompaniessuchasFacebooktotakeresponsibilityinthe

struggleagainstfakecontentusedforpoliticalmanipulation.

Keywords:onlineactivism;crowdsourcedactivism;fakeprofiles;hate

profiles;Facebook

Introduction

Thisishowweshuthispagedown.We’renearly1300membersandifwe

eachspend5secondsreportinghispage,it’llberemovedinnotime.

(StopFakeHateProfilesonFacebook,postbyadmin,1July2015)

Pre-print version

Page 2: j?”eY?Z]>Q”?

2

InJune2015,aclosedFacebookgroupnamedStopFakeHateProfileson

Facebook(STOPfalskeHAD-PROFILERpåFACEBOOK)wascreatedtocombat

fakeprofilesspurringanti-MuslimdiscoursesinDenmark.Within24hours,the

groupattractedover1000membersengaginginseveralformsofcooperative

contestation.Mostnotably,thegroupusedcollectivereportingofcontentfor

violationsofFacebook’scommunitystandards(Facebook,2016).

StopFakeHateProfilesonFacebookwascreatedinreactiontoseveral

Facebookpagesthatsparkedhundredsofhatefulcommentsandsharesfrom

DanishFacebookusersinspring2015.Thesepageswereallconstructed

aroundfictitiousMuslimidentities,claimingtorepresentawiderMuslim

communityinDenmark.TheirconsistentmessagewasthatDanishMuslims

wereconspiringtotakeoverthecountry,rapeDanish(white)women,andkill

allnon-Muslims(Farkasetal.,2017).Mostuserswhoreactedtothishateful

contentdidnotrealisetheidentitieswerefakeandexpressedaggressionas

wellasxenophobicsentimentsincomments.Furthermore,userswho

contestedthepages’authorshipincommentsweresystematicallyremoved

andblockedbytheanonymouspageadministrator(s).

JournalistsfromtheDanishpublicservicebroadcaster(DanmarksRadio)

eventuallyreportedonthephenomenon,highlightingthattheFacebookpages

werefakeandlikelyconstructedbyfar-rightactiviststosmearMuslims

(Nielsen,2015).Thelatterfinding,however,couldnotbepositivelyconfirmed,

asFacebookenablespageadministratorstoremaininvisible,challengingany

Pre-print version

Page 3: j?”eY?Z]>Q”?

3

legalactionagainstthem.StopFakeHateProfilesonFacebookthusrepresented

theonlysystematicattempttoresistandcombatthefakeMuslimFacebook

pages.ThisoccurredthroughcrowdsourcedreportingofthepagestoFacebook

inordertogetthecompanytoclosethemdown.

Thepowerofcrowdsourcedonlineactivismasaformofcollectiveresistance

haslongbeenheralded,thoughparticularlyintheearlydaysofsocialmedia

(Benkler,2006;Jenkins,2006;Shirky,2009).Scholarshavearguedthat

dataficationofpersonalinformationandtheriseofmany-to-many

communicationenablesnewformsofpoliticalmobilisationbasedonapolitics

ofnumbers(LoaderandMercea,2011).Acoreaspectofsuchpolitical

mobilisationiscrowdsourcedcollectiveaction(inthestreetsandonline),often

throughsocialmediaplatformsthatenablelarge-scalecoordinationand

organisation(Lotanetal.,2011).Thereare,however,limitationstothisformof

action.Giventheincreasingrangeofopportunitiesforengagementinthe

digitalera,ithasbecomecommontolamentthatonlineparticipationisno

morethanfeel-good‘slacktivism’(Morozov,2011),‘clicktivism’(White,2010),

oraltogetherlackingacollectivealtruisticcomponent(Bauman,2001).While

thiscriticismmightringpartiallytrueinthecaseofStopFakeHateProfileson

Facebook,thisarticlearguesthatparticipationandactivismorganisedinthe

groupwasconditionedandlimitedbyFacebook’sdigitalarchitecture.Basedon

participant-observationalfindings,thearticleexploresthechallengesthatStop

FakeHateProfilesonFacebookfacedinitsstruggle.Drawinguponthese

findings,thearticlesuggeststhatcrowdsourcinguseractioncanonlymakea

marginalcontributiontosustainablypreventingfakehateprofilesonsocial

Pre-print version

Page 4: j?”eY?Z]>Q”?

4

mediaundercurrentconditions.Asustainablesolutionwouldrequirethat

Facebooktakesongreaterresponsibilityasacompanyandprovidemorethan

itscurrentlylimitedandopaqueusersupport.

Thecrowdsourcingideologyonsocialmedia

JeffHowecoinedtheterm‘crowdsourcing’in2006inaWiredarticle.Theidea

ofcrowdsactingandcreatingtogetherwaspresentinearlydiscoursesabout

socialmedia.TimO’Reilly’s(2005)conceptof‘Web2.0’hadthe“wisdomofthe

crowds”asakeycomponent.Theseideasmainlyincludedcrowdsourcingina

businesscontext,focusingonbottom-upcreativeprocessesinwhich

companiesadoptideasfromcrowds,fans,andamateurs.Inadiscourse

analysisofpopularpressarticlesconcerningcrowdsourcing,Brabham(2012,

p.407)concludesthattheconceptwasalsopromotedas“apotentially

powerfultooltospurpublicparticipationandtransparencyingovernment

affairs.”Brabhamargues,however,thatthe‘amateur’labelinthiscontext

delegitimisesotherwise-worthyagentsbydevaluingtheirrolesasparticipants

andcitizensindemocraticsociety.Liberatorytechnologicaldiscourses–a

powerfulpartofthecorporateidentitiesofsocialmediacompaniessuchas

GoogleandFacebook(Turner,2006)–havethusbeenadoptedinboth

contemporarybusinessculturesanddemocraticdiscoursesandprocesses.

Basedonananalysisofthepoliticaleconomyofthedigitalmediaindustry,

Sandoval(2014,p.252)arguesthat,ratherthanbeingsocial(asassertedin

corporatesocialresponsibilitystatements),socialmediacompaniesexploit

labourand“arefeedingonthecommonsofsociety.”Socialmediaandother

Pre-print version

Page 5: j?”eY?Z]>Q”?

5

techcompaniesco-optideasoftheradicalleft,suchasparticipation,

decentralisation,spontaneousinteraction,andlackofdisciplineandhierarchy

(Žižek,2009),inconceptssuchascrowdsourcing.Thesediscoursesof

empowerment,however,shifttheobligationforactiononsocialmediatothe

users.Thiscreatespotentialsforuseractionaswellasdisempowermentsince

socialmediacompaniescandisowncorporateresponsibilityforphenomenaon

theirplatformssuchasfakehateprofiles.

Facebook’scommunitystandardsstatethatthecompanystrives“towelcome

peopletoanenvironmentthatisfreefromabusivecontent.Todothis,werely

onpeoplelikeyou”(Facebook,2016).Thecompany’smodelforhandling

abusivecontentisthusbuiltaroundfreeuserlabour.Thisiseconomically

beneficialforFacebook,asitonlyemployscommercialcontentmoderatorsto

reviewcontentreportedbycost-freeusers(Fuchs,2015;Roberts,2016).It

alsoenablesthecompanytodistanceitself,bothlegallyandcommunicatively,

fromabusivematerialonitsplatformbygrantingusersprimaryresponsibility.

ThisevasionstrategyiscentraltoFacebook,whichiscurrentlyseekingto

increasethisdelegationofresponsibility:“Theideaistogiveeveryoneinthe

communityoptionsforhowtheywouldliketosetthecontentpolicyfor

themselves”(Zuckerberg,2017).Asweshowinthisarticle,Facebook’suser-

centredapproachisproblematic,asthecompanycircumventsresponsibility

forcounteringabusewhileprovidinginadequateandopaquetoolsforuser

action.Thisdisempowersusersandlimitsthepotentialforcounteracting

phenomenasuchasfakehateprofiles.

Pre-print version

Page 6: j?”eY?Z]>Q”?

6

Activismandsocialmedialogics

Manychallengesconfrontactivistsusingcorporatesocialmediaplatformsto

counter-actdominantdiscourses,includingracism.PoellandBorra(2011,p.

695)notethatfor“crowd-sourcingalternative[news]reporting,”thecontent

oftweetsisframedbymainstreamnewstoproducevisibility.Leistert(2015)

arguesthatcorporatesocialmediahavebecomealgorithmicmassmedia,using

algorithmstocensor,normalise,andstandardiseactivistcommunications.The

silencingofcriticalvoicesreinforcesneoliberalvaluesinwhichcorporate

socialmediaplatformsareembedded(Couldry,2010).Inordertosuccessfully

achievepoliticalgoals,activistsinsocialmediaenvironmentsmustthusadapt

theirpoliticalstrategiestocorporatesocialmedialogicssuchasconnectivity,

popularity,anddatafication(vanDijckandPoell,2013).Throughthis

adaptation,activistsriskbeingco-optedbythesocialmedialogicsthatthey

attempttouseagainstthesystem(GalisandNeumayer,2016).Inotherwords,

insteadofempoweringactivists,“powerhaspartlyshiftedtothetechnological

mechanismsandalgorithmicselectionsoperatedbylargesocialmedia

corporations”(PoellandvanDijck,2015,p.534).

Inhisphilosophyoftechnology,Feenberg(2002)focusesonhumanagency,

arguingthattechnologyreinforcesprevailingpoliticalhierarchiesandpower

relations.Feenbergsuggests,however,thattechnologicalinventionalso

providesnewopportunitiesforsubversiveactorstochallengepoliticalsystems

byappropriatingnewmediatechnologiesfortheircause.Acriticalanalysisof

technologymustconsequentlybe“balancedbydescriptionofwhatpeople

actuallydoinpractice”(Mackenzie,2006,p.458).Thisrequiresustoopenthe

Pre-print version

Page 7: j?”eY?Z]>Q”?

7

blackboxofsocialmediamateriality“asactiveagentsshapingthesymbolic

andorganizationalprocessesofsocialactors”(Milan,2015,p.897).Inthe

following,weseektounpackthisblackboxbyanalysingthesocialmedia

practicesofStopFakeHateProfilesonFacebook.Insodoing,weexplorehow

thegroupnavigatessocialmedialogicsinitsstruggleagainstfakehateprofiles.

Aparticipant-observationalinquiry

Thisarticlebuildsupondatacollectedduringsixmonthsofparticipant

observationwithintheclosedFacebookgroupStopFakeHateProfileson

Facebook.ThefieldworkcommencedinlateJune2015,shortlyafterthe

creationofthegroup,andendedinearlyJanuary2016.Levelsofactivity

withinthegroupvariedoverthecourseofthesixmonths,withconcentrations

aroundoccurrencesoffakehateprofiles.Duringtheresearchperiod,StopFake

HateProfilesonFacebookcontestedeightfakehateprofiles,whichattracteda

totalofover14,000commentsand6000sharesfromDanishFacebookusers.

Priortothegroup’screation,datafromfivefakeMuslimFacebookpageshad

alreadybeencollectedinAprilandMay2015(Farkasetal.,2017).WhenStop

FakeHateProfilesonFacebookwascreatedinresponsetofakeMuslim

Facebookpages,itwasthuspossibletoinitiateresearchwithinthegroup

shortlythereafter.Dataonfakehateprofilescollectedpriortotheexistenceof

thegroupenablesacomparativeperspectiveonfictitiousprofilesbeforeand

afterinitiationofthegroup’scollectivecontestation.

Thedatasetof13fakehateprofiles–eightofwhichwerecontestedbyStop

FakeHateProfilesonFacebook–derivesfromourqualitativeapproach.Based

Pre-print version

Page 8: j?”eY?Z]>Q”?

8

ononlineparticipantobservations(Hine,2015),ourresearchobjectiveisto

exploreandinvestigatethepeople,objects,controversies,conflicts,and

negotiationssurroundingfakehateprofilesandthestruggleagainstthem.

Throughoutthesixmonthsofresearch,wecontinuouslyobservedand

participatedintheactivitiesofStopFakeHateProfilesonFacebook.This

involvedahighdegreeofengagement.Wesupportedthegroup’scauseand

interactedregularlywithgroupmembers,particularlythegroup

administrator.Theprimarypurposeoftheseinteractions(whichcanbestbe

describedasinformaldialogues)wastounderstandthewaysinwhichthe

groupwasorganisedandoperated.Basedontheseobservations,thisarticle

seekstounravelthedelicatepracticesandtacticsofStopFakeHateProfileson

Facebookaswellasthechallengesfacingthegroup’scrowdsourceduser

action.Infuturework,quantitativemeasurescouldadvantageouslybe

includedtoexaminethescaleandproliferationoffakehatesprofilessuchas

thosecontestedbythisgroup.

InformedconsentwassecuredfrommembersofStopFakeHateProfileson

Facebookbyfirstcontactingthepageadministratorandreceivingpermission

fromhim.Wethereafteraskedtheadministratortopostastatementinthe

groupforallmemberstosee,disclosingourresearchagendaandrequesting

permissiontodofieldwork.Inthisstatement,weassuredgroupmembersthat

wewouldprotecteveryone’sanonymity.Thegrouprespondedpositivelyto

ourrequest.Useractivitywithinthegroupwasarchivedthroughscreenshots

and‘printpage’functionalitiestoensuretheexistenceofdataincasethegroup

oritscontentweredeleted.Intotal,wecollected38posts(allmadebythe

Pre-print version

Page 9: j?”eY?Z]>Q”?

9

groupadministrator)and943comments.Subsequenttoourfieldwork,all

namesofgroupmembershavebeenanonymised,andtheactoftranslation

fromDanishtoEnglishrendersthecontentunsearchable.

StopFakeHateProfilesonFacebook

Asitsnamesuggests,StopFakeHateProfilesonFacebookwascreatedwiththe

purposeoffindingandcombattingwhatitterms‘fakehateprofiles’on

Facebook.Inthegroup’smissionstatement,thistermencompasses“fake

profiles[…]groups,orpagescreatedtoincitefearandhatredtowardsspecific

groupsinDanishsociety”(Postbygroupadministrator,21June2015).The

group’sobjectivewastoexposeandcombatsuchprofilesthroughcollective

effortsusingFacebook’sdigitalarchitectureandcommunitystandards.All

groupmemberscouldinvitenewuserstothegroup,thoughtheyhadtobe

approvedbytheadministrator.Thegroupwasexplicitlynon-partisan,and

politicaldiscussionswerenotallowed.

Thefakehateprofilescombattedbythegroupwereidentifiedonthebasisofa

numberofcharacteristics,mostprominently:useofstolenprofilepictures,

falselyproclaimedaffiliationswithexistingorganisations;deletionofuser

commentsquestioningtheprofiles’authorship,lackofresponsewhen

contactingtheprofiles,andrhetoricsimilartothatofpreviousprofiles

identifiedasfake.ThefakehateprofilesusedfictitiousMuslimidentitiesto

constructanarrativeofMuslimsplottingtooverrunDanishsociety,killingand

rapingethnic(white)Danesintheprocess:

Pre-print version

Page 10: j?”eY?Z]>Q”?

10

IslamisNOTaboutpeacebutsubjugationtoAllah.Oncewegetsharialawin

Denmark,allyouinfidelpigswillhavetosubmittoIslam…It’sokaytokill,

aslongasthevictimsareinfidels.AllahuAkhbar!

(Facebookpost,MohammedEl-Sayed,30June2015)

YouDanescanlaughatmenow,butjustwaituntilwegetsharialawin

Denmark,thenallnon-Muslimswillbe‘removed’(ifyouknowwhatImean)

J.AllahuAkhbar!YoushouldbythewayknowthatItakeyourmoney,I

havesexwithyourcheapwomen,andImakethempregnant.

(Facebookpost,MehmetDawahAydemir[1],9September2015)

Mostpostsfromthesefakehateprofilescontaineddirectthreatstooppress,

rape,andkill(non-Muslim)Danes.Othersprovokedbyrejoicinginthe

September11terroristattacksorstatingthatallDanesarestupidpigsand

dogs.Onallprofiles,theaggressivestatementswerepresentedasoriginating

fromyoung,Danish-speakingMuslimslivinginDenmark(Farkasetal.,2017).

Thesefictitiousidentitieswereallconstructedaroundexistingxenophobic

stereotypesofMuslimsasviolent,hypersexual,andalienthreatstotheDanish

welfarestate(Hervik,2011).Stolenimages,text,andhyperlinkswerethusall

deployedtopersonifythesestereotypesascredibleandauthenticindividuals.

Oneachprofile,imagesofArab-lookingpeoplewerepresentedalongsidelinks

toexistingMuslimorganisations,postsaboutMuslimsdestroyingDenmark

fromwithin,andimagesofburningDanishflagsortheflagofISIS.Thefake

profilesallclaimedtospeakonbehalfofawiderMuslimcommunityin

Denmark,allparticipatinginalarge-scaleconspiracy:“WeMuslimshavecome

Pre-print version

Page 11: j?”eY?Z]>Q”?

11

tostay.Wehaven’tcomeinpeace,buttotakeoveryourshittycountry”

(Facebookposts,ZahraAl-Sayed,2July2015).Rhetoricandwordingwere

highlysimilaracrossthepages,indicatingthattheircreatorswerelikely

connectedoridentical.AsFacebookenablespageadministratorstoremain

completelyanonymous,however,theactualidentitiesandmotivesofthese

authorscannotbeestablished.Consequently,intermsofmotive,wecanonly

concludethatallfakehateprofilesdeliberatelysoughttoprovokeandspark

anti-MuslimaggressionfromDanishFacebookusers–anagendainwhichthey

largelysucceeded.

Acrossthevariousfakehateprofiles,theviolentrhetoricpromptedthousands

ofusercommentsfromDanesbelievinginthestatedauthorshipand

respondingwithhatredtowardsthefictitiousidentitiesaswellasMuslimsand

immigrantsingeneral:

Gohometoyourowncountry!Wedidn’taskyoutocomeheretoour

country”

(Usercomment,MehmetDawahAydemir[1],9September2015)

Whatthefuckisthis,youfuckingpig!!!WehelpyoucometoDenmarkand

thisishowyouthankus!

(Usercomment,MehmetDawahAydemir[1],9September2015)

NotallusersreactedwithaggressiontowardsthefictitiousMuslimidentities.

Numeroususersactivelytriedtodismantlethehatredandwarnothersthat

Pre-print version

Page 12: j?”eY?Z]>Q”?

12

theprofileswerefake.Theanonymouspageadministrators,however,

systematicallyobstructedsuchattempts,asweshowbelow.

StopFakeHateProfilesonFacebookwasformedinJune2015toorganiseand

increasecontestationoffakehateprofiles.Thiscontestationinvolvedfour

distinct,concurrentprocesses:(1)findingandreportingpages,(2)alerting

users,(3)alertingjournalistsandauthorities,and(4)speculatingabout

culprits.Theseprocesseswerecontinuouslynegotiatedanditeratively

developedbygroupmembersinordertoincreasetheeffectivenessoftheir

efforts.Inthefollowingsections,weexplorethegroup’scrowdsourced

contestation,focusingonthesocio-technicaltacticsdeployedintheirstruggle.

Basedonthisexamination,wediscussthelimitationsandopportunitiesfor

crowdsourceduseractiononsocialmediaandtheirimplicationsforthe

preventionoffakehateprofiles.

(1)FindingandreportingfakehateprofilesonFacebook

Thefirststepinthegroup’scontestationwastosearchforFacebookprofiles,

groups,orpagesusingfictitiousidentitiestodisseminatehatespeech.When

memberslocatedsuchcontent,theywouldcontactthegroupadministrator

andgethimtosharealinkwithinthegroupalongsideashortstatement,for

example:

We’vereceivedatipfromamemberanditseemsthisprofileisfake.The

rhetoricissimilartopreviousprofiles,andIwillthereforeencourageyouall

toreportthepage,sowecanshutitdown.

Pre-print version

Page 13: j?”eY?Z]>Q”?

13

(Postbygroupadministrator,5January2016).

UserswouldfollowthelinkandreporttheprofiletoFacebookforviolationsof

thecompany’scommunitystandards,whichprohibitbothfakeidentitiesand

hatespeech(Facebook,2016).KeytothisoperationwasFacebook’s‘report’

button,whichcanbefoundonallprofilesandpagesaswellasposts,pictures,

andvideos.WhenreportingviolationstoFacebook,groupmemberswould

subsequentlypostcommentswithinthegroup,oftensimplywriting:

‘Reported’.Memberswouldtherebycontinuouslymaketheir(otherwise-

invisible)actionsvisibletoeachother.Someusersdeliberatelyreportedthe

sameprofilefornumerousviolations(e.g.fakeidentity,hatespeech,

harassment)andalsoreporteditsindividualposts.Thiswasdoneinthehope

thatlargerquantitiesofreportswouldcauseFacebooktopaymoreattention.

Facebook’sprocessingoffileduserreportsisahighlyopaqueprocess

(Roberts,2016),makingitdifficulttodiscernhowthecompanyoperates.

Consequently,groupmemberswoulditerativelyexchangepersonal

experiencesandhypothesesinanattempttomaximisetheeffectivenessof

theircrowdsourcedcontestation.Arecurrentfindingbygroupmemberswas

thatthequantityofreportsplayedamajorroleinFacebook’sresponse,

althoughthecompanyofficiallydeniesthis(Facebook,2016).Often,when

filingreports,groupmemberswouldinitiallyreceiveastandardresponsefrom

Facebook,statingthatthereportedprofile(s)didnotviolateFacebook’s

communitystandards.Groupmemberswouldtakescreenshotsofthesereplies

andpostthemwithinthegroupaccompaniedwithstatementsofdisbelief:

Pre-print version

Page 14: j?”eY?Z]>Q”?

14

Really!?They’vecheckedthepageandwon’tshutitdown…!!!”

(Commentbygroupmember,22June2015)

Ican’tbelieveFacebookclaimsthisisn’tviolatingtheircommunity

standards?Afakeprofilespreadinghatespeech,thismustbeaviolationof

therules?

(Commentbygroupmember,1July2015)

Afternumerousadditionalreports,Facebook’sverdictwouldoftenbe

reversed,causinguserstopostnewscreenshotsaccompaniedwithstatements

ofcelebration:“TogetherWEARESTRONG…evilwillbeconqueredinthisway!

<3”(Commentbygroupmember,2July2015).Thepatternofreversed

verdictsfromFacebookcausedmemberstospeculatethatthecompanyatfirst

respondsalgorithmicallytofiledreportsandonlylaterinvolvesactualhuman

staff:“Keepreportingtheprofiles.Facebookusesrobotstogothroughthe

complaints.Realhumanswillonlylookintoitiftherearelotsofreports”

(Commentbygroupmember,22June2015).

Asexemplifiedbytheabovequotes,groupmembersfeltempoweredthrough

theircollectivecontestation,asitenabledthemtoinfluence(whatwere

otherwisefelttobe)unwaveringdecisionsmadebyFacebook.Simultaneously,

however,groupmembersalsofeltdisempoweredbyFacebook’ssecrecyand

lackofcollaboration,withnoapparentinterestinthegroup’scrowdsourced

activism.Thegroup’spowerseemedtoliesolelyinitssize.Groupmembers

andthegroupadministratorwouldthereforerepeatedlyemphasisethe

Pre-print version

Page 15: j?”eY?Z]>Q”?

15

importanceofallmembersfilingasmanyreportsaspossibleandcomplaining

ifFacebookdidnotrespondpositivelytotheirrequest(s):

Weneedtokeepreportinghis[theanonymousadministrator’s]page.At

somepoint,Facebookwillgettiredandlookatwhathe’sactuallywritten.

Thisishowweshuthispagedown.We’realmost1300members,andifwe

allspend5secondsreportinghispage,it’llberemovedinnotime.

(Postbygroupadministrator,1July2015).

ThecontestationsurroundingFacebook’s‘report’buttonshowshowStopFake

HateProfilesonFacebookengagedintacticalsocio-technicalnegotiations,

continuouslyattemptingtounlockFacebook’ssecretivedigitalarchitecture

anduseitstrategicallytofurtheritscause.Thesestrategiesprovedlargely

successful,ascontestedhateprofilesoftenonlyexistedforafewdaysbefore

Facebookremovedthem(seeTable1).

<InsertTable1-Overviewoffakehateprofilesandtheirdurationsof

existence.>

(2)Alertingusers

Onseveraloccasions,hateprofilescontestedbyStopFakeHateProfileson

FacebookreceivedhundredsoreventhousandsofcommentsfromDanish

Facebookusers.MostusersacceptedtheproclaimedMuslimidentitiesand

expressedanger,hostility,andevenracism:

Pre-print version

Page 16: j?”eY?Z]>Q”?

16

Fuckyou,youfuckingmonkey

(Usercomment,MohammedEl-Sayed,1July2015)

Disgustinganimal!Getthefuckoutofmycountry…youdon’tbelonghere!

(Usercomment,MehmetDawahAydemir[1],11September2015)

It’sbecauseofpeoplelikeyouthatmoreandmorepeopleturnracist

(Usercomments,MehmetDawahAydemir[1],11September2015).

StopFakeHateProfilesonFacebooksoughttodismantlethishatredtowards

Muslimsandimmigrantsbyalertingusersthattheprofileswerefakeand

deliberatelycreatedtoinciteaggression.Groupmemberswouldpost

commentsontheprofiles,warningusersnottobelieveintheproclaimed

identitiesandpoliticalmanipulation.Aftermakingsuchcomments,members

wouldnotifyeachotheroftheiractionswithintheclosedgroup:“Wroteonhis

page,awarningandalinktothisgroup”(Commentbygroupmember,1July

2015).

Theanonymouspageadministrator(s)runningthefakehateprofiles,however,

continuouslysabotagedtheseefforts.OnallFacebookprofilesandpages,

administratorscanremoveanycontentwithoutnotifyingitsauthorandcan

blockanyuserfrommaking(additional)comments.Theadministrator(s)of

thefakehateprofilessystematicallyusedthistechnologicalfeaturetotheir

advantagebydeletingallcommentsandblockingalluserswhocontestedtheir

proclaimedauthorship.Newusersencounteringthehateprofileswouldthus

beexposedexclusivelytousercommentsaffirmingthelegitimacyofthe

Pre-print version

Page 17: j?”eY?Z]>Q”?

17

sources.Groupmembersandtheirwarningswerecontinuouslydeletedand

blockedeventhoughtheystillattemptedtoalertusers:

Yougetblockedsofastinthere,butatleastIgottopost20timesthatthe

pagewasfakebeforeitwasover.

(Commentbygroupmember,1July2015);

Iwasremovedrightaway!!Thepersonbehindmustknowthatwework

togetherandareonhistrail!!

(Commentbygroupmember,1July2015).

Duetothesystematicmoderationperformedbytheanonymouspage

administrator(s),theeffectivenessofthegroup’seffortstoalertusersastothe

existenceoffakeMuslimhateprofilesseemstohavebeenlimited.The

administratorsofthehateprofilestacticallyexploitedFacebook’sdigital

architecturetosilenceanycontestation.Nevertheless,afewgroupmembers

reportedthattheyhadinfactfirstbelievedinthefakeauthorshipandonly

laterbecameawareofitsdeceptivenatureduetocommentsmadebygroup

members:“Yesterday,Ireallythoughtthatsomeonewasbeingthishostileand

Ijumpedinfeetfirstandcursedhimback.I’mgladsomeonetoldmeitwas

fake.”(Commentbygroupmember,12September2015).Thishighlightshow

thestrugglebetweenStopFakeHateProfilesonFacebookandvariousfakehate

profilesfundamentallyconcernedvisibilityandawareness.Thehateprofiles

soughttorenderallcontestationinvisible,leavingonlycommentsaccepting

theproclaimedauthorship.Thegroup’sgoal,incontrast,wastomakeits

Pre-print version

Page 18: j?”eY?Z]>Q”?

18

contestationasvisibleaspossibletowarnuserswhilesimultaneouslymaking

thepagesinvisible(throughdeletionbyFacebook).

Inseveralrespects,Facebook’sdigitalarchitectureseemstohavesupported

thehateprofiles’agendabyprovidingunlimitedanonymitytotheir

administrator(s)aswellasasymmetricalpowerrelationsbetween

administrator(s)andusers(e.g.throughtheabilitytoremoveanycomment).

Thecountergroup’seffortstoalertusersregardingfakehateprofilesmight

havefurthermorehadtheunforeseenconsequenceofcontributingtotheir

proliferation.Facebook’salgorithmscontinuouslyevaluatecontentand

‘decide’howfaritshouldspreadbasedonanumberofparameters.Acentral

parameterinthisprocessisthenumberoflikes,comments,andshares

receivedbythecontentinquestion(Bucher,2012).Commentspostedby

membersofStopFakeHateProfilesonFacebooktowarnusersmightthushave

indirectlyincreasedthefakehateprofiles’reach,potentiallydeceiving

additionalFacebookusers.Thus,despitethegroup’scollectiveefforts,fake

hateprofilescontinuedtoposeacomplexchallenge.Aswediscussbelow,

however,thegroupalsopursuedthegoalofmakingtheircontestationvisible

outsideofFacebook.

(3)Alertingjournalistsandauthorities

AlthoughStopFakeHateProfilesonFacebookprimarilyoperatedwithinthe

boundariesofsocialmedia,thegroupalsosoughttoreachoutandinvolve

journalistsandauthoritiesintheirstruggle.Thegroupmanagedtoattractthe

attentionofseveralmajorDanishmediainstitutions,includingthenational

Pre-print version

Page 19: j?”eY?Z]>Q”?

19

tabloidEkstraBladet(Ryde,2015),thenewspaperInformation(Skovhus,

2015),andtheTVbroadcasterTV2.Thesemediaoutletsallreportedonthe

phenomenonoffakehateprofilesonFacebook,thelattertwointerviewingthe

group’sadministratoraspartoftheircoverage.Thepublicoutreachagenda

pursuedbythegroupwasprimarilyundertakentowarntheDanishpublic

aboutpotentialdemocraticdangersposedbyfakehateprofiles.

Simultaneously,itenabledthegrouptoattractmorememberstoparticipatein

theirstruggle.Theseeffortslargelyprovedsuccessful.Yetaswiththegroup’s

effortstowarnusersonFacebook,theincreasedattentiontofakehateprofiles

achievedthroughmassmediacouldpotentiallyalsohaveledmoreusersto

engagewiththeprofiles,indirectlyincreasingtheirproliferationonFacebook.

Inparallelwiththegroup’seffortstoreachjournalists,groupmembersalso

contactedtheDanishpoliceandtheintelligenceagency(PET)inorderto

instigateinvestigationsintotheoriginatorsofthefakehateprofiles.The

ephemeralityofthecontestedcontent,however,presentedanobstacletothis

agenda.Theshorttimeperiodsinwhichthefakehateprofilesexistedmeant

thatarchivedmaterialwasnecessaryinordertofilepolicereports.Thegroup

addressedthischallengebyworkingcollectivelytocompilesuchmaterial:

REQUEST:Amemberisaskingforscreenshotsfromthehateprofilesthat

havebeenshutdownsincethepolicewanttolookintothecase…please

sendthemtomeinaprivatemessageorpostthembelow,sothey’revisible.

(Postbygroupadministrator,2July2015)

Pre-print version

Page 20: j?”eY?Z]>Q”?

20

Inadditiontothechallengeofpiecingtogetherdeletedevidence,the

ephemeralityandanonymityoffakehateprofilesprovedproblematic.

Ephemeralityofcontentmeantthatauthoritiescouldneverobservethe

consequencesoffakehateprofilesastheyunfolded.Furthermore,the

completeanonymityoffakehateprofilecreators,enabledbyFacebook’s

design,meantthatnochargescouldbefileddirectlyagainstanyone.Doingso

wouldfirstrequireathoroughinvestigationandclosecontactwithFacebook.

ThiscausedfrustrationandfeelingsofdisempowermentformembersofStop

FakeHateProfilesonFacebook,asthesocialmediacompanyshowedno

apparentinterestincollaboratingwiththem.Thegroupwasthustotallyreliant

onDanishauthoritiesforconductinginvestigations,yetthegroupalso

experiencedalackofsupportfromauthoritiesinidentifyingandinvestigating

thecreatorsoffakehateprofiles.Thiscauseddistress:

Idon’tunderstandwhyITspecialistsinthepolicecan’tfindtheir[the

administrator’s]IPaddress…Thesefakeprofilesaresofarout…

(Commentbygroupmember,13September2015)

Idon’tthinkwecanachieveanythingthroughpolicereports.

(Commentbygroupmember,1July2015)

ThemostpowerfulmeansavailabletomembersofStopFakeHateProfileson

Facebookthuscontinuedtobetheircollectiveeffortstoreportfakehate

profilestoFacebookandgetthemdeleted.Yetthisstrategyhadsevere

limitations,asthegroupcouldnevergettotherootoftheproblemdueto

Facebook’sdigitalarchitectureand(apparent)lackofinterestincollaboration.

Pre-print version

Page 21: j?”eY?Z]>Q”?

21

Theanonymouscreatorsoffakehateprofilescouldcontinuously(re-)create

newfictitiousidentitieseachtimeoldoneswereremoved.Formembersof

StopFakeHateProfilesonFacebook,thiscausedfrustration,eveninsituations

inwhichFacebookdeletedfakehateprofiles:“Yes:)!Finally,they[thefake

hateprofiles]areremoved..buthe[theanonymousadministrator]willjust

createanewone:(:(“(Commentbygroupmember,1July2015).Thelackof

collaborationfromauthoritiesandFacebookledtoinvestigationsbygroup

memberstoidentifytheanonymouscontentcreators.

(4)Speculatingaboutculprits

ArecurringthemewithinStopFakeHateProfilesonFacebookwasspeculations

astowhowerebehindthefakehateprofilescombattedbythegroup.On

Facebook,allpageandprofileadministratorscanremaincompletely

anonymous.Evenifapageorprofileisremoved,noinformationisprovidedas

towhocreatedit.Duetonumeroussimilaritiesacrossdifferentfakehate

profiles,groupmembersbecameconvincedthatseveralprofileswerecreated

bythesameadministrator(s):“Thisisexactlythesamerhetoricasonthelast

one.It’sthesamepersonwho’sbehindit,fuckingcoward”(Commentbygroup

member,24October2015);“Youjustknowit’sa20-year-oldkidwithno

friendsandNazitendencieswho’sbehindthekeyboard.”(Commentbygroup

member,1July2015).Severalmembersexpressedfrustrationattheabilityof

theanonymousadministrator(s)tocontinuallyconstructnewfakehate

profilesandsparkaggression,eventhoughFacebookcontinuallydeletedthe

pages.Membersalsoexpressedhopethatauthoritieswouldreactand

investigatetheculprits:“Ireallyhopehe[theadministrator]willbepunished

Pre-print version

Page 22: j?”eY?Z]>Q”?

22

forthehatredhecreates”(Commentbygroupmember,12September2015).

Othersconductedtheirowndetectiveworkandformulatedhypothesesabout

specificpeoplewhocouldbebehindtheprofiles,includingfar-rightactivists.

Suchspeculationswere,however,criticisedbyothermembers,whoargued

thatStopFakeHateProfilesonFacebookshouldnotbecomeavigilantegroup:

“ThisisexactlywhatImean.Asuspicionisn’tenoughtoaccusepeople”

(Commentbygroupmember,12September2015).Thegroupnever

successfullyidentifiedanyhatecontentcreators,thoughtherewerestrong

indicationsthatseveralofthefakehateprofilescombattedbythegrouphad

thesameadministrator(s).

Crowdsourcedsocialmediaactivism

HavingexploredtheactivistpracticesofStopFakeHateProfilesonFacebook,

wenowaddresshowgroupmembersnavigatedFacebook’ssocialmedialogics

intheirstruggleagainstfakehateprofiles.Wefurthermorediscusswhether

thegroup’scrowdsourcedactivismprovedsuccessfulintermsofits

overarchinggoalofstoppingfakehateprofiles.Thegroup’scrowdsourced

contestationandreportingdoesindeedseemtohavesucceedinshorteningthe

lifespansoffakehateprofiles(seeFigure1).Fakehateprofilesstudiedinour

researchthatexistedpriortothegroup’sformationexistedsignificantlylonger

thandidprofilesthatwerecreatedafterthegroup’sformation.Facebook,

however,maintainsinitscommunitystandardsthatquantityofreportsdoes

notinfluencetheevaluationofflaggedcontent(Facebook,2016).Facebook’s

opaqueevaluationproceduresmeanthatdifferentfactorscouldhavehadan

impact.

Pre-print version

Page 23: j?”eY?Z]>Q”?

23

<InsertFigure1-Lifespansoffakehateprofilesbeforeandafterthe

formationofStopFakeHateProfilesonFacebook.>

Temporalityisnottheonlyavailableparameterforevaluatingthegroups’

struggleagainstfakehateprofiles.Someprofilesthatexistedfortheshortest

periodsoftime(andwerecreatedafterthegroup’sformation)werealsothose

thatreceivedthemostcommentsandsharesfromDanishFacebookusers(see

Figure2).Forexample,aprofilenamedMehmetDawahAydemir[1],which

existedforlessthantwodaysinSeptember2015,managedtoattract10,426

commentsand4954shareswithinthisperiod.Mostcommentingusersdidnot

recognisethepage’sdeceptivecharacter.Itsrapidproliferationcanpartiallybe

explainedbyFacebook’salgorithmicprioritisationofshorttimedecayswhen

assessingtheimportanceofcontentand‘deciding’itsreach(Bucher,2012).

Posts,images,andvideoscaninotherwordsreachthousandsofuserswithin

hoursiftheysparkalargenumberofinteractions.Thisseemstohavebeenthe

casewithMehmetDawahAydemir[1].

<InsertFigure2-Timesofexistenceoffakehateprofilesandnumbersof

commentsandshares.>

AlthoughStopFakeHateProfilesonFacebooksuccessfullyreducedthe

lifespansoffakehateprofiles,theprofilescouldstilldeceiveandprovoke

thousandsofuserswithinthisperiod.Lifespan,then,doesnotseemtobethe

bestindicatorofthegroup’ssuccessindismantlingfakehateprofiles.This

Pre-print version

Page 24: j?”eY?Z]>Q”?

24

raisesthequestionofwhetherthisformofcrowdsourcedactivismrepresentsa

viabletrajectoryforstoppinghatredandmanipulationonsocialmedia.Should

itbeuptouserstostopphenomenasuchasfakehateprofilesonFacebook?Or

shouldsocialmediacompaniestakegreaterresponsibility?

Basedonthechallengesandlimitationsfacingthecrowdsourcedactivismof

StopFakeHateProfilesonFacebook,wearguethatFacebook’sdelegationto

usersofresponsibilityforreportingviolationsisproblematic.Unlesssocial

mediacorporationstakegreaterresponsibilityincombattingfacelesshatred

andracismproducedbyanonymousadministrators,noactioncangobeyond

solelyclosingdownsuchhateprofiles.Reachingouttoandcollaboratingwith

authorities–findingcontentandidentifyingitscreators–couldbepartofa

solution.Removingunlimitedanonymityforpageandprofileadministrators

couldbeanother.Suchefforts,however,wouldrequireFacebooktochangeits

self-image,whichiscurrentlythatofatechcompanyandnotamediacompany

(Seetharaman,2016).IfFacebookistostopfakehateprofilesonitsplatform,

thecompanymustacknowledgethatproblemsassociatedwithfakeidentities

andhatredarepartiallyitsresponsibilityandnotonlythatofusers.This

argumenthasrecentlybeenraisedindebatesconcerning‘fakenews’(Stromer-

Galley,2016).Hopefully,futureresearchcanhelpaddresstheseissuesby

expandingcurrentknowledgeontheextensivenessoffakehateprofileson

socialmediaaswellasrelatedphenomenasuchasfakenewsspreadbysocial

bots(seeShaoetal.,2017).Sucheffortscouldadvantageouslydrawuponboth

bigdataanalysisandmachinelearning(seeFerraraetal.,2016).

Pre-print version

Page 25: j?”eY?Z]>Q”?

25

Onthebasisofthepresentarticle’sfindings,Facebook’slimitedresponseto

thephenomenonoffakehateprofileshighlightsadiscrepancybetweenthe

company’sbusinessmodelanditscorporateideals.Facebook’sbusinessmodel

isbuiltaroundcommodificationofuser-generateddataanduserattention,

makingthequalityofcontenteconomicallysecondarytotheattentionit

receives.Atthesametime,Facebook’scorporateidentity,whichhijacksleft-

wingideasofparticipationanddecentralisation(Žižek,2009),burdensusers

withresponsibilityfortacklingproblemssuchasfakehateprofiles.Inthis

process,thecompanyprovidesonlylimitedopportunitiesforuserstoengage

incrowdsourcedactivism.EventhoughFacebookreferstoitsplatformasa

“globalcommunity”(Facebook,2016),thecompanyseemstoprioritise

commodificationofuserattentionovertheempowermentofusersandquality

ofcontent.

Inthecurrentstateofaffairs,crowdsourcingofresponsibilityleavesuserswith

a‘report’buttonastheironlyweapon.Eveniffakehateprofilesonlyexistfor

shortperiodsoftime,theirvisibilitycanstillbegreatduetosocialmedialogics

thatalgorithmicallyprivilegecontentthatquicklyattractscomments,likes,and

shares–evenifthesereactionsexpresshatredandracism.AlthoughStopFake

HateProfilesonFacebookcontinuouslyperformedcrowdsourcedresistance,

Facebook’sarchitecturedisempoweredthegroupbylimitingitspossibilities

foraction,whilefakehateprofilescouldcontinuetospurhatred,aggression,

hostility,andracism.Ifthisistochange,socialmediacompaniesmustreduce

hierarchicalpowerrelations,increasethepotentialforuseraction,andtake

responsibilityforhatredandracismontheirplatforms.

Pre-print version

Page 26: j?”eY?Z]>Q”?

26

Abouttheauthors

JohanFarkasisAssistantLecturerattheITUniversityofCopenhagen.His

researchinterestsincludepoliticalparticipationanddisguisedpropagandain

digitalmedia.

ChristinaNeumayer(PhD,ITUniversityofCopenhagen/MA,Universityof

Salzburg)isAssociateProfessorofdigitalmediaandcommunicationinthe

DigitalDesigndepartmentattheITUniversityofCopenhagen.Herresearch

interestsincludedigitalmediaandradicalpolitics,socialmediaandactivism,

socialmovements,civicengagement,publicsandcounterpublics,surveillance

andmonitoring,andbigdataandcitizenship.

Acknowledgements

TheauthorswouldliketothankStopFakeHateProfilesonFacebook,

particularlyitsadministrator,formakingthisworkpossible.Theauthors

wouldalsoliketothanktheanonymousreviewersofthisjournal.

References

ZygmundBauman,2001.Theindividualizedsociety.Cambridge,UK:Polity

Press.

YochaiBenkler,2006.TheWealthofNetworks:HowSocialProduction

TransformsMarketsandFreedom(Vol.7).NewHaven:YalePress.

Pre-print version

Page 27: j?”eY?Z]>Q”?

27

DarenC.Brabham,2011.“Themythofamateurcrowds:Acriticaldiscourse

analysisofcrowdsourcingcoverage”.Information,Communication&Society,

volume15,number3,pp.394-410.

TainaBucher,2012.“Wanttobeonthetop?Algorithmicpowerandthethreat

ofinvisibilityonFacebook”.NewMedia&Society,volume14,number7,pp.

1164–1180.

NickCouldry,2010.Whyvoicematters:Cultureandpoliticsafterneoliberalism.

London:SAGEPublications.

Facebook,2016.“FacebookCommunityStandards”at

https://www.facebook.com/communitystandards,accessed13January2016.

JohanFarkas,JannickSchou,andChristinaNeumayer,2017.“Cloaked

FacebookPages:ExploringFakeIslamistPropagandainSocialMedia”.New

Media&Society.https://doi.org/10.1177/1461444817707759

AndrewFeenberg,2002.Transformingtechnology:Acriticaltheoryrevisited.

Oxford,UK:OxfordUniversityPress.

EmilioFerrara,Wen-QiangWang,OnurVarol,AlessandroFlammini,andAram

Galstyan,2016.“Predictingonlineextremism,contentadopters,and

interactionreciprocity”.In:E.SpiroandYY.Ahn(editors),SocialInformatics.

Pre-print version

Page 28: j?”eY?Z]>Q”?

28

SocInfo2016.LectureNotesinComputerScience,Springer.

https://doi.org/10.1007/978-3-319-47874-6_3

ChristianFuchs,2015.Cultureandeconomyintheageofsocialmedia.London:

Routledge.

VasilisGalisandChristinaNeumayer,2016.“Layingclaimtosocialmediaby

activists:acyber-materialdétournement”.SocialMedia+Society,volume2,

number3,pp.1-14.

PeterHervik,2011.TheAnnoyingDifference:TheEmergenceofDanish

Neonationalism,Neoracism,andPopulisminthePost-1989World.NewYork:

BerghahnBooks.

ChristinaHine,2015.EthnographyfortheInternet:Embedded,Embodied,and

Everyday.London:Bloomsbury.

JeffHowe,2006.“TheRiseofCrowdsourcing”.Wired,Issue14,number6.

https://www.wired.com/2006/06/crowds/,accessed19July2017.

HenryJenkins,2006.Convergenceculture:Whereoldandnewmediacollide.

NewYork:NewYorkUniversity.

OliverLeistert,2015.“Therevolutionwillnotbeliked:Onthesystematic

constrainsofcorporatesocialmediaplatformsforprotest”.In:L.DencikandO.

Pre-print version

Page 29: j?”eY?Z]>Q”?

29

Leistert(editors),Criticalperspectivesonsocialmediaandprotest:Between

controlandemancipation.London:Rowman&LittlefieldPublishersInc,pp.5–

52.

BrianD.LoaderandDanMercea,2011.“NetworkingDemocracy?”.Information,

Communication&Society,volume14,number6,pp.757–769.

http://doi.org/10.1080/1369118X.2011.592648

GiladLotan,ErhardtGraeff,MikeAnanny,DevinGaffney,IanPearce,danah

boyd,2011.“TheArabSpring|TheRevolutionsWereTweeted:Information

Flowsduringthe2011TunisianandEgyptianRevolutions”.International

JournalofCommunication,volume5,pp.1375–1405.http://doi.org/1932–

8036/2011FEA1375

AdrianMackenzie,2006.“Innumerabletransmissions:Wi-Fi®fromspectacle

tomovement”.Information,Communication&Society,volume9,number6,pp.

781–802.http://dx.doi.org/10.1080/13691180601064139

StefaniaMilan,2015.“Fromsocialmovementstocloudprotesting:The

evolutionofcollectiveidentity”.Information,Communication&Society,volume

18,number8,pp.887–900.

http://dx.doi.org/10.1080/1369118X.2015.1043135

EvgenyMorozov,2011.Thenetdelusion:Hownottoliberatetheworld.London:

PenguinBooks.

Pre-print version

Page 30: j?”eY?Z]>Q”?

30

StineBødkerNielsen,2015,19May.“ViovertagerDanmark:Falskefacebook-

sidersættermuslimeridårligtlys”.DRNyheder.

http://www.dr.dk/Nyheder/Indland/2015/05/18/110828.htm,accessed19

July2017.

TimO’Reilly,2005.“Whatisweb2.0”at

http://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html,,

accessed19July2017.

ThomasPoellandErikBorra,2011.“Twitter,YouTube,andFlickrasplatforms

ofalternativejournalism:Thesocialmediaaccountofthe2010TorontoG20

protests”.Journalism,volume13,number6,pp.695–713.

http://doi.org/10.1177/1464884911431533

ThomasPoellandJosevanDijck,2015.“SocialMediaandActivist

Communication,”In:C.Atton(editor).TheRoutledgeCompaniontoAlternative

andCommunityMedia.London:Routledge,pp.527-537.

SarahT.Roberts,2016.“CommercialContentModeration:DigitalLaborers’

DirtyWork,”In:S.U.Noble&B.Tynes(editors),TheIntersectionalInternet:

Race,Sex,ClassandCultureOnline.NewYork:PeterLang,pp.147–160.

http://doi.org/10.1007/s13398-014-0173-7.2

Pre-print version

Page 31: j?”eY?Z]>Q”?

31

RonjaRyde,2015.12September.EkspertomMehmet-profil:Forargende

Dannebrogs-pissererformentligfalsk.EkstraBladet.

http://ekstrabladet.dk/nyheder/samfund/ekspert-om-mehmet-profil-

forargende-dannebrogs-pisser-er-formentlig-falsk/5729027,accessed19July

2017.

MarisolSandoval,2014.Fromcorporatetosocialmedia:Criticalperspectiveson

corporatesocialresponsibilityinmediaandcommunicationindustries.

Abingdon,UK:Routledge.

DeepaSeetharaman,2016.25October.“FacebookLeadersCallItaTech

Company,NotMediaCompany”.TheWallStreetJournal.

https://www.wsj.com/articles/facebook-leaders-call-it-a-tech-company-not-

media-company-1477432140,accessed19July2017.

ClayShirky,2009.11December.Thenetadvantage.ProspectMagazine.

http://www.prospectmagazine.co.uk/2009/12/the-net-advantage,accessed

19July2017.

ChengchengShao,GiovanniL.Ciampaglia,OnurVarol,AlessandroFlammini,

andFilippoMenczer,2017.“Thespreadoffakenewsbysocialbots”.Eprint

arXiv:1707.07592.

Pre-print version

Page 32: j?”eY?Z]>Q”?

32

PhillipR.Skovhus,2015.22September.ӮgtehadfrafalskeFacebook-

profiler”.Information.http://www.information.dk/546054,accessed19July

2017.

JenniferStromer-Galley,2016.2December.“ThreewaysFacebookcould

reducefakenewswithoutresortingtocensorship”.TheConversation.

http://theconversation.com/three-ways-facebook-could-reduce-fake-news-

without-resorting-to-censorship-69033,accessed19July2017.

FredTurner,2010.Fromcounterculturetocyberculture:StewartBrand,the

WholeEarthNetwork,andtheriseofdigitalutopianism.Chicago:Universityof

ChicagoPress.

JoséVanDijckandThomasPoell,2013."Understandingsocialmedialogic."

MediaandCommunication,volume1,number1,pp.2-14.

http://www.cogitatiopress.com/mediaandcommunication/article/view/70

MicahWhite,2010,12August.“Clicktivismisruiningleftistactivism”.The

Guardian.

http://www.guardian.co.uk/commentisfree/2010/aug/12/clicktivism-

ruining-leftist-activism,accessed19July2017.

SlavojŽižek,2009.Violence:Sixsidewaysreflections.London:ProfileBooks.

Pre-print version

Page 33: j?”eY?Z]>Q”?

33

MarkZuckerberg.2017,“BuildingaGlobalCommunity”at

https://www.facebook.com/notes/mark-zuckerberg/building-global-

community/10154544292806634/,accessed7august2017.

Pre-print version

Page 34: j?”eY?Z]>Q”?

34

Name Time period Number of days

Ali El-Yussuf [3] 16/06-22/06 2015 7

Mohammed El-Sayed 30/06-02/07 2015 3

Fatimah El-Sayed 01/07-02/07 2015 2

Zarah Al-Sayed 02/07-02/07 2015 1

Mehmet Dawah Aydemir [1] 09/09-12/09 2015 4

Mehmet Dawah Aydemir [2] 13/09-15/09 2015 2

Ebrahim Said 24/10-25/10 2015 2

Mohammed Al-Dawah 05/01-07/01 2016 3

Table1:Overviewoffakehateprofilesandtheirdurationsofexistence.

Pre-print version

Page 35: j?”eY?Z]>Q”?

35

Figure1:LifespansoffakehateprofilesbeforeandaftertheformationofStopFake

HateProfilesonFacebook.

Pre-print version

Page 36: j?”eY?Z]>Q”?

36

Figure2:Timesofexistenceoffakehateprofilesandnumbersofcommentsandshares.

Pre-print version