A Theory of Canceling

In Decem­ber of 2013, Jus­tine Sac­co, a senior direc­tor of cor­po­rate com­mu­ni­ca­tions at Inter­Ac­tive Corp (IAC), post­ed a tweet right before board­ing an 11-hour flight from Lon­don to Cape Town. In the tweet, which has since been delet­ed, she wrote: “Going to Africa. Hope I don’t get AIDS. Just kid­ding. I’m white!”

At the time, Sac­co only had 200 fol­low­ers, but the edi­tor of Val­ley Wag, a blog ded­i­cat­ed to Sil­i­con Val­ley, repub­lished the tweet with some crit­i­cal com­men­tary. As recount­ed by Buz­zfeed, oth­er reporters quick­ly picked up on the tweet, shar­ing it amongst them­selves on a slow Fri­day, and expressed increduli­ty that it was still up hours lat­er because, again, Sac­co was on a long-haul flight and not online. As reporters con­tin­ued to share the sto­ry, reg­u­lar Twit­ter users noticed oth­er insen­si­tive tweets Sac­co had post­ed, with a crescen­do of out­rage and calls for her ter­mi­na­tion.

By the time Sac­co land­ed in South Africa, 11 hours lat­er, she was the most talked about sub­ject on the Inter­net. Dur­ing her flight, every­one from celebri­ties to in-flight WiFi provider Indie GoGo had tweet­ed jokes at her expense; par­o­dy accounts and a satir­i­cal web domain had been set up to mock her, and her own employ­er had denounced her. With­in hours of her land­ing, IAC fired her.

Sac­co has become a kind of cau­tion­ary tale for peo­ple who are very online, from the dumb “nev­er tweet” reac­tions, to crit­i­cisms of the angry pol­i­tics of gawk­er, to white priv­i­lege, to wor­ries about the misog­y­nis­tic slant of many out­rage storms. While those all have ele­ments of truth in them, I think we can learn by exam­in­ing the many rea­sons that went into her tweet’s viral impact, how some enter­pris­ing jour­nal­ists cre­at­ed a scan­dal from noth­ing, and why the adage that things online don’t mat­ter is trag­i­cal­ly false.

A Theory of Viral Outrage

There is copi­ous aca­d­e­m­ic lit­er­a­ture about how fram­ing, agen­da-set­ting, and prim­ing in the media can cre­ate the base con­di­tions for pub­lic reac­tions to a giv­en sto­ry — even a par­tic­u­lar­ly poor­ly-timed tweet. And there is a grow­ing body of research about what makes some pieces of con­tent go viral while oth­ers get ignored (since, after all, Sac­co had, even in 2013, hard­ly post­ed the most racist con­tent on Twit­ter). And final­ly, a quick envi­ron­men­tal study can show how the evolv­ing norms of dis­course on Twit­ter have set the stage for an unend­ing churn of out­rage con­tent.

The viral­ness of Sacco’s tweet can be under­stood using Jon­ah Berger’s the­o­ry of “social epi­demics.” The viral­i­ty of con­tent has a few cru­cial ele­ments that fac­tor into how “con­ta­gious” an idea or a sto­ry online can be. Sacco’s case includes five of the six ele­ments Berg­er says fac­tor into how catch­able an idea can be: Her tweet exhib­it­ed social cur­ren­cy (a form of sig­nal­ing that occurs when peo­ple dis­cuss a top­ic — in this case, express­ing one’s anti-racism), trig­gers (when a hash­tag about her trend­ed, it formed a feed­back loop that remind­ed peo­ple to par­tic­i­pate in the con­ver­sa­tion), emo­tion (peo­ple gen­er­al­ly don’t like racism), pub­lic (again, a trend­ing hash­tag rein­forced itself), and sto­ry (a lot of broad nar­ra­tives about colo­nial­ism and racism in Africa).

There were broad­er cul­tur­al forces at work as well, which con­tributed to the unique envi­ron­ment for Sacco’s tweet to gain such viral­i­ty. Racism is a top­ic of intense dis­cus­sion on social media, and the elec­tion of Barack Oba­ma to the White House in 2008 had raised pub­lic aware­ness of how race remained a con­stant of Amer­i­can dis­course. A fas­ci­nat­ing study by Bet­ti­na Love and Bran­de­lyn Tosolt showed a form of racial “cog­ni­tive dis­so­nance” in Amer­i­ca, where­by (most­ly white) peo­ple both express racist views while believ­ing that racism has been “solved.” At the same time he was elect­ed, psy­cho­log­i­cal research was estab­lish­ing a grow­ing under­stand­ing of how expo­sure to racism, racist imagery, and racist jokes online can have dele­te­ri­ous effects on chil­dren. So even as many Amer­i­cans felt racism was on the way out, racist con­tent remained high­ly preva­lent, and there grew increas­ing aware­ness of how destruc­tive that is for chil­dren exposed to it.

The online dis­in­hi­bi­tion effect, as explained by John Suler, can turn the inter­net into a unique­ly unfil­tered place, where peo­ple online do not feel bound by norms about their behav­ior and thus act out in var­i­ous ways. Because peo­ple feel rel­a­tive­ly anony­mous, they can feel empow­ered to make offen­sive remarks out of a false belief that those remarks will be con­se­quence-free.

In addi­tion, dur­ing Obama’s first term Sarah Sobier­aj and Jef­frey Berry iden­ti­fied both online and in tra­di­tion­al media the rapid rise of what they called “out­rage dis­course.” While indi­vid­ual out­rage can be an impor­tant mech­a­nism of rein­forc­ing social norms, they define out­rage dis­course as “a par­tic­u­lar form of polit­i­cal dis­course involv­ing efforts to pro­voke vis­cer­al respons­es.” Indi­vid­ual out­rage can also be con­trast­ed with viral out­rage, where­by thou­sands of ran­dom strangers who have no direct rela­tion­ship to an inci­dent “dog­pile” the tar­get of out­rage – pre­cise­ly what hap­pened to Sac­co. It cre­ates what Takuya Sawao­ka and Benôit Monin call a “para­dox that the same indi­vid­ual out­rage that seems laud­able and nec­es­sary in iso­la­tion may be viewed more as exces­sive and bul­ly­ing when echoed by mul­ti­tudi­nous oth­er users.” On social media, we call it can­celling.

But Sacco’s sto­ry is not as sim­ple as a bad tweet going viral on its own and get­ting her can­celled. It also played into a spe­cif­ic cul­tur­al moment, ampli­fied a grow­ing trend of media fram­ing, and was sub­ject to agen­da-set­ting by enter­pris­ing jour­nal­ists. In Sacco’s case, it seems clear that Sam Bid­dle, an enter­pris­ing and aggres­sive blog­ger, suc­cess­ful­ly set an agen­da in how to cov­er the tweet. This isn’t new to social media: As Stephen Reese wrote in 1991, three pre­con­di­tions need to exist for a long-stand­ing issue to become the sub­ject of intense pub­lic scruti­ny: “a) the media and pub­lic agen­da cor­re­late; b) the media agen­da pre­ced­ed the pub­lic agen­da in time; and c) alter­na­tive expla­na­tions that respond to third fac­tors, such as objec­tive con­di­tions, have been ruled out.”

Reese places this into the con­text of a pow­er-rela­tion­ship between jour­nal­ist and sub­ject for agen­da-set­ting, which is salient to the media envi­ron­ment in which Sac­co tweet­ed in 2013. Spiro Kiousis argued that there is a feed­back loop of sorts at work in how agen­das set by the media play into pub­lic atti­tudes, which fur­ther encour­age more agen­da set­ting to pla­cate pub­lic atti­tudes. In oth­er words, “agen­da-set­ting involves the trans­fer of salience from the media to the pub­lic,” which is how Bid­dle could “push” the sto­ry to a far wider audi­ence than it would have reached organ­i­cal­ly. Matthew Ragas and Spiro Kiousis con­duct­ed a study that sug­gest­ed this type of agen­da set­ting is more of a “pack­ag­ing process, which involves select­ing cer­tain aspects of real­i­ty and mak­ing them more salient.”

Thus, Biddle’s use of fram­ing cre­at­ed a spe­cif­ic slant on Sacco’s tweet — and because of the spe­cif­ic cul­tur­al moment of Amer­i­ca in Obama’s sec­ond term, the inter­net pub­lic was already primed to respond to racist jokes. Kirk Hallahan’s mod­els of fram­ing sug­gest that peo­ple can “use asso­ci­a­tion and expec­ta­tion to make infer­ences about events and to impute mean­ing not man­i­fest­ed in the mes­sage itself.” In oth­er words, peo­ple were already expect­ing to react to a racist joke, so when a joke that could plau­si­bly be called racist was put in front of them, they filled in the miss­ing con­text and react­ed.

Sac­co had a lot going against her. Being stuck in the air, with lim­it­ed inter­net access at best left her unable to address the grow­ing out­rage. This essen­tial­ly left her defense­less, as she was unable to main­tain a rela­tion­ship with the stake­hold­ers push­ing the scan­dal. “One of the most impor­tant aspects of cri­sis man­age­ment is com­mu­ni­ca­tion.” The reac­tion to her tweet grew so large because she couldn’t address it for so long.

Last­ly, online out­rage can boil over into dras­tic, real world con­se­quences. While online inter­ac­tions are often writ­ten off as being not real in some way, this has nev­er been the case. In the ear­ly 90s, the jour­nal­ist Julian Dibbel wrote A Rape in Cyber­space, which described how a play­er of an ear­ly online mul­ti­play­er game was vir­tu­al­ly raped using the inter­ac­tiv­i­ty tools pro­vid­ed by the game, result­ing in real-life trau­ma. Research in the two decades since estab­lished that online mul­ti­play­er games have impor­tant psy­cho­log­i­cal con­se­quences for peo­ple, both pos­i­tive and neg­a­tive, in part because of how much time the play­ers spend in the game. Online is real life, in oth­er words — sep­a­rat­ing the two doesn’t make any sense giv­en how much of our social inter­ac­tions are medi­at­ed by the inter­net. And because rela­tion­ships formed through these com­mu­ni­ties and oth­er forms of social net­work­ing can be as pow­er­ful as real-world friend­ships, it seems indis­putable that events online can have pow­er­ful con­se­quences offline as well.

How Trad & Soc Media Made a Mountain From a Molehill

Before her tweet went viral, Jus­tine Sacco’s twit­ter feed was most­ly wry obser­va­tions about her dai­ly life – often sar­cas­tic, some­times caus­tic, but gen­er­al­ly bland. It wouldn’t be out of place in the group chats of most edu­cat­ed upper mid­dle class peo­ple liv­ing in a big city. Despite her sta­tus as a high-rank­ing com­mu­ni­ca­tions pro­fes­sion­al for a promi­nent firm, she treat­ed her pub­lic feed like a pri­vate gath­er­ing of friends, where there was lee­way to be sar­cas­tic and edgy in a way that sim­ply would not be appro­pri­ate in pub­lic.

But Sacco’s tweet would not have gone any­where had it not been sig­nal-boost­ed — and that required pro­mot­ing it to a larg­er audi­ence. The jour­nal­ist who first pub­li­cized Sacco’s tweet was Sam Bid­dle, a tech­nol­o­gy jour­nal­ist high­ly crit­i­cal of Sil­i­con Val­ley. He was a pio­neer of using jour­nal­is­tic snark to not just mock celebri­ties (how it is often used in tabloids), but to dri­ve home the­o­ries of social and crim­i­nal jus­tice. The con­cept of snark (acer­bic sar­casm) was defend­ed most promi­nent­ly by Gawk­er writer Tom Soc­ca, who framed it as an authen­tic reac­tion to the inau­then­tic smarm that dom­i­nates “uplift­ing” and “objec­tive” media. Snark is the dom­i­nant tone on social media, espe­cial­ly Twit­ter, and peo­ple often are far snarki­er there than they ever would be in per­son. Bid­dle was a mas­ter of snark and con­tin­ued to mock her for months after she was job­less and made into a pari­ah, which exac­er­bat­ed how much she was affect­ed by the out­rage.

A selec­tion of reac­tion tweets can also explain how and why peo­ple were so eager to demon­strate their dis­gust with Sacco’s tweet. A cru­cial aspect of viral con­tent is the way it sig­nals social cur­ren­cy on the part of the per­son shar­ing — the act of acknowl­edg­ing a piece of viral con­tent, to say noth­ing of com­ment­ing on it, let’s the poster demon­strate cur­ren­cy and rel­e­vance.

The social cur­ren­cy of keep­ing up with Sac­co between the time her tweet was high­light­ed to the time of her land­ing in South Africa became a form of gameshow, as #HasJustineLand­edYet became, briefly, the most-used hash­tag on the inter­net. Peo­ple were thrilled to demon­strate social cur­ren­cy by post­ing updates on her flight, to sig­nal their own beliefs by con­demn­ing her joke, and to rev­el in the inher­ent dra­ma of some­one being cut off from the world at the exact moment they became a fig­ure of infamy.

And those reac­tion tweets were a prob­lem. Jus­tine Sacco’s con­tent land­ed in a per­fect storm of cul­tur­al move­ments, where a white woman mak­ing a racial­ly insen­si­tive joke as she trav­els to Africa would bind togeth­er many social and media forces, destroy­ing her career in the process.

A year after start­ing the inter­net storm, Bid­dle apol­o­gized to Sac­co, and in that apol­o­gy, he revealed some of his moti­va­tion for push­ing the out­rage storm. “Twit­ter dis­as­ters are the quick­est source of out­rage, and out­rage is traf­fic. I did­n’t think about whether or not I might be ruin­ing Sac­co’s life. The tweet was a bad tweet, and see­ing it would make peo­ple feel good and angry—a sim­ple social and emo­tion­al trans­ac­tion that had hap­pened before and would hap­pen again and again.”

There are three things to unpack from this quote. The first is that Bid­dle was aware of the forces that Berg­er iden­ti­fied as dri­ving viral­i­ty online and try­ing to delib­er­ate­ly trig­ger them. He also was required to do this, as gen­er­at­ing viral con­tent was a rev­enue strat­e­gy for Gawk­er. Last­ly, he sim­ply didn’t care about the per­son he tar­get­ed – a com­mon theme for viral out­rage, where the tar­get of out­rage is rarely viewed as wor­thy of con­sid­er­a­tion. Sociopa­thy and the right cul­tur­al moment came togeth­er to cre­ate dis­as­ter for Sac­co.

Bid­dle had already primed his fol­low­ers to react, angri­ly, to his con­tent – that was his entire beat at Gawk­er, where he prid­ed him­self on col­lec­tion “scalps” from ran­dom fig­ures in the indus­tries he tar­get­ed with efforts at gen­er­at­ing viral out­rage. The way Bid­dle framed these sto­ries (“a crit­i­cal activ­i­ty in the con­struc­tion of social real­i­ty because it helps shape the per­spec­tives through which peo­ple see the world”) cre­at­ed momen­tum for cre­at­ing viral out­rage. Thus, a mis­step by the man­age­ment at a start­up firm wasn’t just a mis­step – in Biddle’s frame, it was evi­dence of men­dac­i­ty and fraud. Because his audi­ence always saw sto­ries framed in such an unchar­i­ta­ble way, when­ev­er he would post about exec­u­tives or oth­er wealthy peo­ple in coastal Cal­i­for­nia, his thou­sands of twit­ter fol­low­ers would pile on with snarky com­men­tary and often per­son­al jabs at the sub­ject.

As a rule, pub­licly racist com­men­tary is frowned upon. How­ev­er, com­men­tary that appears racist can be either retract­ed or cor­rect­ed in a time­ly man­ner if the per­son who made it is avail­able to par­tic­i­pate in the dis­cus­sion. Sac­co, how­ev­er, post­ed her racial­ly insen­si­tive joke dur­ing a peri­od where she would be unable to man­age any fall out. While this detail was not in her tweet, she post­ed the joke imme­di­ate­ly before board­ing an 11-hour flight, which lim­it­ed her abil­i­ty to either apol­o­gize or delete the joke. This pro­vid­ed a gut-lev­el impres­sion, even if under­stood on an intel­lec­tu­al lev­el to be unfair, that she did not care and was sim­ply obliv­i­ous to the effects of her post. Sac­co was unable to com­mu­ni­cate on her flight (though many raised the ques­tion of why she wasn’t on in-flight wifi, con­sid­er­ing this was often an expen­sive ser­vice it isn’t nec­es­sar­i­ly indica­tive of neg­li­gence on her part).

So What?

Was Sac­co wrong to tweet that joke? Yes. How­ev­er, while the spe­cif­ic patholo­gies of viral out­rage or, as it’s now known, “can­cel cul­ture,” are worth exam­in­ing in detail, there are more imme­di­ate lessons that can be drawn. The inter­net has pre­sent­ed an inter­est­ing chal­lenge for peo­ple and orga­ni­za­tions try­ing to man­age their rep­u­ta­tions online — the very well known process­es of fram­ing and agen­da-set­ting can com­bine with the unique attrib­ut­es of a social epi­dem­ic and draw glob­al atten­tion to an oth­er­wise obscure racist joke. By par­tic­i­pat­ing in a spe­cif­ic media cul­ture with a dis­parag­ing point of view to those it con­sid­ered more pow­er­ful, Gawk­er had laid the foun­da­tion for an explo­sion of out­rage. In addi­tion, the nature of Sacco’s tweet, which cut through sev­er­al extreme­ly con­tro­ver­sial top­ics in a very short amount of space, had mul­ti­ple build­ing blocks of viral con­tent.

Gawk­er, where Bid­dle wrote at the time, was part of an inten­tion­al agen­da-set­ting by a small group of online jour­nal­ists. In a famous 2007 con­fronta­tion, the tele­vi­sion host Jim­my Kim­mel con­front­ed an edi­tor at Gawk­er about her post­ing images of him being intox­i­cat­ed. When he accused her of throw­ing rhetor­i­cal rocks at celebri­ties, she respond­ed, “aren’t they kind of pro­tect­ed by piles of mon­ey from those rocks.”

The con­cept of “punch­ing up” – attack­ing those per­ceived to have more pow­er than you do because they deserve it in some way – in jour­nal­ism has its roots in Fin­ley Peter Dunne’s char­ac­ter Mr. Doo­ley. His mock­ery of the press is often quot­ed as “The job of the news­pa­per is to com­fort the afflict­ed and afflict the com­fort­able,” even though Dunne’s book presents Doo­ley as being high­ly crit­i­cal of the muck­rak­ing yel­low jour­nal­ist of the late 19th cen­tu­ry. How­ev­er, in this case, the “piles of mon­ey” do not pro­tect the sup­pos­ed­ly pow­er­ful exec­u­tives from the barbs thrown their way. Ulti­mate­ly the high­ly-traf­ficked blog Bid­dle wrote for had far more pow­er to gen­er­ate pub­lic intrigue than the vir­tu­al­ly anony­mous cor­po­rate pub­lic rela­tions work­er he had tar­get­ed. While tra­di­tion­al agen­da-set­ting the­o­ry places the largest amount of pow­er in the cor­po­ra­tion, in real­i­ty the pow­er dynam­ic was flipped. In a very real way, Bid­dle was punch­ing down — though he would nev­er admit such a thing.

That isn’t to say agen­da-set­ting is dis­card­ed, or can explain every­thing. The tra­di­tion­al media have always had a role to play in shap­ing the public’s per­cep­tion, and that involves a cal­cu­la­tion between a jour­nal­ist and their source. But social media doesn’t real­ly work that way — most of the time, the sources are also the tar­gets of cov­er­age, and quite often both are fel­low jour­nal­ists.

Much of the media dis­course on social media is per­for­ma­tive — jour­nal­ists are per­form­ing for each oth­er to frame and cap­ture real­i­ty in a way that is mutu­al­ly impres­sive. You can see this as an act of cura­tion (and in fact, many jour­nal­ists describe their social media pro­files as cura­tion), where bits of real­i­ty are select­ed or des­e­lect­ed for ampli­fi­ca­tion. The tem­pera­ment and judg­ment of the jour­nal­ist online then becomes vital to under­stand­ing why and how they will fix­ate on one piece of con­tent but not anoth­er. Thus, when you take Gawker’s insti­tu­tion­al belief that any and all cor­po­rate exec­u­tives above an unspec­i­fied lev­el of income are fair game to sin­gle out and mock, it becomes clear that Sacco’s tweet was not des­tined for infamy – she just had incred­i­bly bad luck in that some­one hap­pened to find her bad con­tent.

There are two obser­va­tions about this inci­dent that are worth not­ing. The first is that real-world con­se­quences for online events are sim­ply a fact of life and won’t stop. The end­less hand-wave that online out­rage is lim­it­ed and con­tained and there­fore not “real” is, sim­ply, false. Sac­co was an ear­ly vic­tim of viral out­rage lead­ing to career ruin, but there is a grow­ing list of peo­ple who have expe­ri­enced a sim­i­lar process. In fact, that grow­ing list is part­ly why and how Sac­co was able to move on with her life, even­tu­al­ly – because the inter­net did first.

The sec­ond obser­va­tion is a vari­a­tion of the cliche “nev­er tweet.” Sac­co con­fused the pub­lic for the pri­vate and did not real­ize her pub­lic twit­ter feed could be strip mined for con­tent for a muck­rak­ing blog. Even in 2013, this demon­strates a lack of aware­ness and flu­en­cy in how online media is gen­er­at­ed and con­sumed (which is, to be unspar­ing toward Sac­co, not a great place to be as a PR exec­u­tive). It is also worth not­ing that cer­tain types of ver­bal sar­casm trans­late extreme­ly poor­ly to the writ­ten word, espe­cial­ly when expe­ri­enced by peo­ple out­side of a closed social group.

But we can still take lessons from her dread­ful expe­ri­ence and hope­ful­ly avoid a sim­i­lar fate. The mod­ern social media plat­forms are designed very specif­i­cal­ly to hack your dopamine cycle and make you crave the val­i­da­tion of strangers for post­ing emo­tion­al­ly charged con­tent. It is a sys­tem designed to coax ever more per­son­al infor­ma­tion about you into the open where it can be strip-mined for mar­ket­ing insights. While we obvi­ous­ly can­not con­trol all of the data about us float­ing in the aether, every sin­gle thing that we inten­tion­al­ly post online will be exploit­ed and used in some way that affects us.

Sec­ond­ly, devel­op­ing offline rela­tion­ships is real­ly impor­tant. Bid­dle did not change his opin­ion of Sac­co until he both­ered to talk to her in per­son. It was direct­ly see­ing the cost his treat­ment had imposed that he recon­sid­ered how he had treat­ed her and what sort of pow­er he held over peo­ple. And it wasn’t until more and more infor­ma­tion about the severe per­son­al toll this inci­dent imposed on Sac­co emerged that peo­ple real­ized just how hor­rif­ic her expe­ri­ence was.

Sacco’s “reha­bil­i­ta­tion,” as one crit­ic put it, did not always reflect well on her. While her human­iza­tion was undoubt­ed­ly a per­son­al relief, it doesn’t change the fact that her Twit­ter post was in very poor taste. Some­one get­ting ham­mered for post­ing tru­ly bad con­tent makes for a very poor poster child, as one crit­ic put it: “The plain mean­ing of what she said still remains, res­onates beyond her inten­tions, and ampli­fies the vol­ume of an already cor­ro­sive dis­course.” Sacco’s tweet drew on a lot of very painful his­to­ry for an entire con­ti­nent of peo­ple, and that his­to­ry is inescapable. While the exact treat­ment to which she was sub­ject­ed turns out to be pro­found­ly bad luck, she would nev­er have had to face it if she hadn’t made a racist joke in the first place.

Even so, “nev­er tweet” is a shal­low con­clu­sion to come to. It is unre­al­is­tic to expect peo­ple to exist in pub­lic, online, for the sake of employ­ment and social cur­ren­cy, but also shrug when one is sub­sumed into a storm of viral out­rage. Peo­ple need room to make mis­takes and to be able to apol­o­gize and come back from that mis­take, but the cur­rent norms of social media don’t allow that sort of per­son­al growth to take place. That doesn’t mean con­dem­na­tion for bad behav­ior gets tak­en off the table, by any means (and hid­ing behind “free speech” to jus­ti­fy racist speech is not okay, either), but it should mean there is a lit­tle bit more grace for peo­ple who do want to change and grow. Sac­co has, luck­i­ly, had that oppor­tu­ni­ty to learn from her expe­ri­ence. Maybe oth­ers need that chance, too.

Joshua Foust used to be a foreign policy maven. Now he helps organizations communicate strategically and build audiences.