Moral Outrage in the Digital Age | Molly Crockett

Summary

Dr. Molly Crockett, an assistant professor of psychology at Yale University, discusses the impact of social media on moral outrage. She explains that outrage is a response to moral violations and has evolved to promote cooperation within social groups. However, social media platforms, driven by engagement, may amplify outrage by promoting emotionally charged content. Crockett's research explores how online expressions of outrage have become easier and more habitual, potentially leading to negative consequences such as misinformation spread and political polarization. Her lab is developing an AI tool to measure online outrage and is investigating its societal effects.

Takeaways

• 🎓 Molly Crockett is an assistant professor of psychology at Yale University, focusing on the effects of social media on moral outrage.
• 📹 The Covington Catholic High School incident highlighted the rapid spread and potential misinformation of viral videos, leading to widespread outrage and subsequent corrections.
• 🗣️ Social media platforms' algorithms prioritize engaging content, which often includes emotionally charged material that can trigger moral outrage.
• 🔥 The expression of moral outrage online has become easier and more common, potentially leading to a perpetual state of outrage in the digital age.
• 🧠 Moral outrage is a mixture of anger and disgust at the violation of a moral standard, and it has evolved to promote cooperation within social groups.
• 🤝 Expressing moral outrage can boost an individual's social reputation, as it signals trustworthiness and willingness to uphold group norms.
• 🧩 Social media platforms may be altering the nature of moral outrage by changing the costs and benefits associated with its expression.
• 📈 The ease of expressing outrage online can lead to more frequent and intense reactions, potentially skewing the perception of moral violations.
• 🤖 Crockett's lab is developing an AI tool called the 'Digital Outrage Classifier' to measure expressions of outrage in social media posts.
• 🔮 The research suggests that social media may amplify personal benefits of expressing outrage while reducing some social benefits, leading to potential negative consequences for society.
• 🌐 The study's findings will be shared on the Digital Outrage Project website, highlighting the impact of social media on moral emotions and societal behavior.

Q & A

Who is Molly Crockett and what is her profession?

Molly Crockett is an assistant professor of psychology at Yale University.

What incident involving Covington Catholic High School students went viral in January?

A video was posted online showing students from Covington Catholic High School appearing to taunt a Native-American man while wearing 'Make America Great Again' red hats.

What was the public reaction to the initial video of the Covington Catholic High School incident?

The video clip went viral, and members of the liberal tribe expressed outrage on social media, with some calling for bans on teens wearing MAGA hats and others demanding the students be named and shamed.

What happened after a longer video of the Covington Catholic High School incident emerged?

The longer video suggested the incident was more complicated than originally thought, leading to some journalists expressing remorse for their initial reporting and a shift in the public discourse.

What is moral outrage and why do people express it?

Moral outrage is broadly considered a mixture of anger and disgust at the violation of a moral standard. People express it to shame and punish wrongdoers, especially when observed by others, which can discourage bad behavior and promote cooperation within social groups.

How does social media's business model influence the type of content users see?

Social media platforms like Facebook and Twitter make more money the longer they hold users' attention, so their algorithms prioritize content that draws the most engagement, which tends to be emotional and moral in nature.

What was the hypothesis Molly Crockett proposed regarding the impact of social media on moral outrage?

Molly Crockett hypothesized that social media might be changing the nature of moral outrage by altering the stimuli that trigger it, the ease of expressing it, and the consequences of such expressions, potentially leading to more frequent and intense experiences of outrage.

What is the 'Digital Outrage Classifier' and what is it used for?

The 'Digital Outrage Classifier' is an AI tool developed by Molly Crockett's lab at Yale to measure expressions of outrage online. It can analyze millions of Tweets and determine whether each one expresses outrage or not.

How does social media potentially affect the expression and consequences of moral outrage?

Social media can reduce the costs associated with expressing outrage, such as effort and risk of retaliation, and can amplify the personal benefits by extending the reach of moral signals to a wider audience, potentially leading to more habitual and mindless expressions of outrage.

What are some of the potential negative consequences of the way social media amplifies moral outrage?

Potential negative consequences include a worse ability to coordinate punishment, the spread of disinformation and propaganda, and the rise in political polarization as outrage expressions may push people towards extremes or create false beliefs about the level of outrage in society.

What is the Digital Outrage Project and how can one follow its progress?

The Digital Outrage Project is an initiative by Molly Crockett's team at Yale to study the impact of social media on moral outrage. More results will be released on their website, digitaloutrageproject.org, where interested individuals can sign up for updates.

Outlines

00:00

📚 The Impact of Social Media on Moral Outrage

Molly Crockett, an assistant professor of psychology at Yale University, discusses the role of social media in amplifying moral outrage. She begins with the Covington Catholic High School incident, illustrating how a viral video led to widespread outrage and calls for action against the students involved. The incident highlighted the rapid spread of information and the subsequent backlash on social media. Crockett emphasizes the need for empirical data on the effects of social media on moral outrage, given its potential impact on mental health, social relationships, and democracy. She introduces her lab's research into how social media might be altering our experience and expression of moral outrage, defining it as a response involving anger and disgust at moral violations. The talk suggests that outrage serves social functions, such as promoting cooperation and trust, and is neurologically linked to reward pathways in the brain.

05:02

🤖 The Digital Transformation of Moral Outrage

This paragraph delves into the neurobiology of moral outrage and how social media's business model and algorithms may be exacerbating it. Crockett hypothesizes that social media platforms, designed to maximize engagement, prioritize emotionally charged content, which in turn triggers moral outrage more frequently and intensely than traditional media or in-person interactions. She presents data from a study showing that people are more likely to encounter and react to immoral events online, experiencing heightened outrage compared to offline encounters. The paragraph explores how the ease of expressing outrage online, with reduced effort and risk, lowers the threshold for individuals to voice their indignation, potentially leading to both positive and negative consequences.

10:04

🔬 The Consequences of Social Media on Outrage Expression

Crockett discusses the potential implications of social media's influence on the expression of moral outrage. She suggests that while reduced expression costs can empower marginalized voices, it may also lead to disproportionate punishment. Furthermore, the unpredictable social rewards for expressing outrage online, likened to variable reinforcement in casinos, may foster habitual outrage responses without genuine emotional engagement. The speaker introduces the 'Digital Outrage Classifier,' an AI tool developed by her lab to measure expressions of outrage in social media data. Preliminary findings indicate that social media's reinforcement of outrage may be shaping user behavior, although the research is ongoing.

15:06

🌐 The Broader Implications of Digital Outrage

In the final paragraph, Crockett considers the broader societal implications of digital outrage. She raises concerns that social media's amplification of personal benefits from expressing outrage might overshadow its social benefits, complicating collective action and potentially aiding the spread of disinformation. Crockett also explores the role of outrage in political polarization, suggesting that social media's reinforcement of extreme expressions could push moderate individuals towards extremes. She concludes by emphasizing the importance of using existing data to understand how social media is transforming moral emotions and calls for public access to data from technology companies to assess the impact of their platforms on society.

Keywords

💡Moral Outrage

Moral outrage is defined as a strong emotional response to a perceived violation of moral standards, typically involving feelings of anger and disgust. In the video, moral outrage is central to the discussion on how social media can amplify and alter the expression of this emotion. The Covington Catholic incident exemplifies how moral outrage can go viral and lead to public shaming, illustrating the power and potential misuse of this emotional response in the digital age.

💡Social Media

Social media refers to digital platforms that enable the creation and sharing of content, as well as interaction among users. The video discusses how social media platforms, driven by algorithms prioritizing engagement, can exacerbate moral outrage by promoting emotionally charged content. The script cites examples such as the Covington Catholic incident and the Kavanaugh hearings, demonstrating how social media can influence the spread and intensity of moral outrage.

💡Tribes

In the context of the video, 'tribes' symbolize social or political groups with shared values and beliefs. The term is used to describe how people align themselves with like-minded individuals, leading to a polarization that can intensify moral outrage. The MAGA hats in the Covington Catholic incident serve as a visual symbol of a 'conservative tribe,' highlighting the tribal nature of moral outrage in social media discourse.

💡Engagement

Engagement, in the video, refers to user interaction with online content, such as likes, shares, comments, and time spent viewing. The business model of social media platforms is shown to prioritize content that drives higher engagement, which often includes emotionally charged material that triggers moral outrage. This creates a feedback loop where outrage becomes more prevalent due to the social rewards it garners.

💡Reward Circuitry

The reward circuitry in the brain is associated with the experience of pleasure and the motivation to engage in activities that lead to pleasurable outcomes. The video explains that expressing moral outrage activates this circuitry, providing a neural reward that reinforces the behavior. This is likened to the craving for addictive substances like cocaine, suggesting that outrage can be addictive due to the brain's response.

💡Variable Reinforcement

Variable reinforcement is a psychological concept where rewards are delivered unpredictably, which is known to strengthen habits. The video suggests that social media uses this pattern to encourage habitual behaviors, such as expressing outrage, by providing likes and shares at random times. This can lead to a decoupling of genuine emotional response from the act of expressing outrage online.

💡Disinformation

Disinformation refers to false or misleading information that is spread, especially among groups or nations. The video raises the concern that the nature of moral outrage and its amplification on social media can facilitate the spread of disinformation. Outrageous content is more likely to be shared, which can be exploited by those spreading false news or propaganda, as seen in the Russian government's sponsored Facebook ads mentioned in the script.

💡Political Polarization

Political polarization is the phenomenon where differences in opinion between political groups become more pronounced and inflexible. The video discusses how social media might be contributing to this by amplifying moral outrage, particularly among more extreme users. This can create an environment where moderate voices feel pressured to conform to more polarized stances, leading to a divided society.

💡Digital Outrage Classifier

The Digital Outrage Classifier is an AI tool developed in the speaker's lab to measure expressions of outrage on social media. It classifies millions of tweets to determine whether they express outrage or not. This tool is used to analyze patterns and predict how social media rewards for expressing outrage may influence future behavior, as part of the research into the impact of social media on moral outrage.

💡Habit Formation

Habit formation is the process by which behaviors become automatic through repetition. In the video, it is suggested that the unpredictable social rewards for expressing outrage on social media can lead to the formation of habits. This means that individuals might express outrage without the genuine emotional response, similar to how a habitual snacker might eat without feeling hungry, leading to a mindless repetition of outrage expression.

Highlights

Molly Crockett, an assistant professor of psychology at Yale University, discusses the impact of social media on moral outrage.

A viral video involving Covington Catholic High School students and a Native-American man sparked widespread outrage and media reactions.

The incident highlighted how quickly social media can amplify stories, leading to public outrage and potential misinterpretations.

Crockett emphasizes the importance of gathering data on the effects of social media on moral outrage due to its societal implications.

Moral outrage is defined as a mixture of anger and disgust at the violation of a moral standard.

Expressing moral outrage can benefit individuals by boosting their social reputation and trust within a group.

The brain's reward circuitry is activated when people punish moral violations, similar to responses to other rewards like food or addiction.

Social media platforms' business models prioritize engaging content, which tends to be emotionally charged and can provoke outrage.

People are more likely to encounter and feel intense moral outrage from online content compared to offline sources.

Online platforms reduce the effort and risk associated with expressing outrage, potentially leading to more frequent and intense reactions.

Social media can amplify personal benefits of expressing outrage by extending the reach of moral signals to a broader audience.

Crockett's lab is developing an AI tool called the 'Digital Outrage Classifier' to measure expressions of outrage on social media.

The research suggests that social media may be transforming moral emotions from a collective good to a potential tool for self-destruction.

Outrage on social media could be linked to the spread of disinformation and propaganda due to the prioritization of outrageous content.

Social reinforcement of outrage online might explain the rise in political polarization observed in recent years.

Crockett calls for the use of existing data to understand how social media is changing moral emotions and their societal impact.

The Digital Outrage Project aims to release more research findings to the public to shed light on the effects of social media on moral outrage.

Transcripts

- Thank you very much, I'm Molly Crockett.

I am an assistant professor of psychology

at Yale University.

In January this year,

a video was posted online featuring a group of students

from Covington Catholic High School in Kentucky.

In the video, the students appear to be taunting

a Native-American man,

while wearing that iconic symbol of the conservative tribe:

red hats declaring, "Make America Great Again," or MAGA.

Over the next few days, the video clip went viral,

as members of the liberal tribe

expressed outrage on social media.

A congressman called for a ban on teens wearing MAGA hats.

Celebrities and journalists

demanded the kids be named and shamed.

Some even wished them physical harm.

Then, in the midst of the uproar,

a longer video emerged,

suggesting the incident was more complicated

than originally thought.

Here's the New York Times:

"Fuller Picture Emerges of Viral Video."

And CNN:

"A new video shows a different side of the encounter."

Some journalists expressed remorse

for how they reported the story.

This one's from the Atlantic:

"I failed the Covington Catholic test.

"Next time there's a viral story,

"I'll wait for more facts to emerge."

These mea culpas were met with more outrage,

this time from conservatives.

And, some weeks later,

one of the students sued the Washington Post for defamation,

to the tune of 250-million dollars.

Now, this story, just like every story, has unique features.

But, a lot about this story

is starting to feel more and more

like just another typical day in America.

It seems like we can hardly go five minutes

without a new episode of tribal outrage

erupting in the public sphere.

And if you spend any time on social media,

you might feel like we're locked

in a state of perpetual outrage.

And there seems to be a growing sense

that social discourse around moral

and political issues in this country

has become quite unproductive,

and that social media

might be at least partly to blame.

Now, there are many opinions out there

about moral outrage in the digital age.

In the wake of the Covington Catholic incident,

Twitter was declared the crystal meth of newsrooms.

(audience laughing)

Some journalists advocated leaving Twitter altogether.

And, pundits have even suggested

that outrage itself is addictive.

But the reality is that there is very little data

that bares on these sweeping claims

about morality and social media.

And gathering this data is imperative, because right now,

our government is deciding whether

and how to regulate tech companies

on the basis of what their products

might be doing to our mental health,

our social relationships,

and our democracy.

Just yesterday, representatives from Facebook,

Google, and Twitter

testified before Congress on harmful online content.

And legislative decisions on this topic

in the next few years will change the course of our society,

for better or for worse.

In my lab at Yale, we've been studying how social media

might be changing the way we experience

and express moral outrage.

Now, what is moral outrage?

Broadly, we can consider it

to be a mixture of anger and disgust

at the violation of a moral standard.

Now, people don't just want to sit alone with their outrage.

They want to express it.

They're motivated to shame and punish the wrongdoer,

especially when other people are watching.

And, expressing outrage is effective in the sense

that being the target of shaming and punishment

makes people less likely to transgress in the future.

Now the fact that outrage discourages bad behavior

suggests one way that outrage could have evolved:

by promoting cooperation within social groups.

Those groups, with more people willing to shame

and punish wrongdoers,

would be more successful than less-moralistic groups.

But, in addition to benefiting groups,

outrage also benefits those individuals who express it.

People who express outrage are more likely to be trusted

and chosen as social partners than people who don't.

So, expressing outrage benefits individuals

by boosting their social reputation.

And these benefits of outrage and moralistic punishment

are encoded in the brain's reward circuitry,

which becomes more activated

when people punish moral violations,

compared to when they turn the other cheek.

Notably, the same brain circuitry

is activated when people are craving other kinds of rewards

like delicious foods,

or, for those people who are addicted, cocaine.

These psychological and neural mechanisms for outrage

evolved in the context of small, face-to-face interactions

in groups of hunter-gatherers many thousands of years ago,

like we heard about earlier.

But, much of our social discourse now

takes place online,

in much larger networks than ever before in human history.

I've been studying the neurobiology of moral outrage

for more than a decade, but in 2016,

in the wake of Brexit and Trump,

I started thinking about how

the modern context of social media

might be changing the nature of moral outrage,

with potentially far-reaching consequences for social life.

My hypothesis is built around a very simple theory

about how we learn from rewards, and adapt our behavior.

Moral outrage is triggered by stimuli

that call attention to moral-norm violations.

People express outrage with a range of responses

that vary in their costs,

and expressing outrage leads to a variety

of personal and social outcomes,

which then reinforce the responses,

making them more likely to occur in the future.

This simple framework can help us to organize

the many ways in which social media might be changing

the costs and benefits of moral outrage.

Let's start by talking about how social media

alters the stimuli that trigger outrage in us.

One feature of social media that's relevant here

is its business model.

So, platforms like Facebook and Twitter

make more money the longer they hold our attention,

so their newsfeed algorithms prioritize content

that draws the most engagement.

And, research as shown that the most engaging content

is emotional content, especially moral emotions.

Bill Brady and colleagues recently showed

that for a variety of political issues,

Tweets containing moral-emotion words

like hate, attack, destroy, are more likely to go viral.

So, every word like this in a Tweet

increases its likelihood of being re-Tweeted by 20%.

The implication of this is that if newsfeed algorithms

are selecting for engagement,

this creates an information ecosystem

that's promoting the most outrageous content,

which might be artificially inflating

our everyday experiences of outrage.

To test this idea, I analyzed some data from a study

conducted by Will Hoffman and colleagues,

and in this study, about 1200 participants

were messaged on their smartphones at random times,

five times a day, over three days,

and they were asked to report their recent experiences,

including whether they had learned about any immoral acts

in the past hour,

and if so, where they learned about it,

and how they felt about it.

By analyzing this data,

I was able to determine where people are most likely

to learn about immoral events, or in other words,

the kinds of events that typically trigger moral outrage.

And what I found is that people are encountering

outrage-triggering events more often online

than in person, or through traditional forms of media

like print, TV, or radio.

And by analyzing how people felt

when they learn about these immoral events,

I found that people experienced more intense outrage

in response to events they encountered online,

than what they encountered in person,

or through traditional forms of media.

So, people are experiencing outrage more often,

and more intensely, in response to online,

than offline content.

Next, let's take a look at how social media

transforms the expression of moral outrage.

If we want to express outrage offline,

we have a few options.

We can gossip about the wrongdoer,

or we can directly confront them,

either verbally or physically,

and these latter methods, of course,

require more effort,

and they also carry potential physical risks.

But online, we can express outrage with a few keystrokes.

We can post angry comments either directly to the target,

or to a broader audience.

We can do this from the comfort of our bedrooms.

And it's clear that all of these responses

require far less effort than expressing outrage offline.

And the risks of retaliation are of course reduced,

because you're behind a screen, or hiding in a large crowd.

So, if we think of outrage expression

as a kind of cost-benefit decision,

the hypothesis is that social media

reduces some of the costs of expressing outrage.

And what this means is that for any given person,

their threshold for expressing outrage

is likely to be lower online than offline.

Now, I really want to emphasize,

this in itself isn't necessarily a good thing,

or a bad thing; it's complicated.

So, reduced costs

might mean than marginalized voices

find it easier to express outrage at injustice,

while facing fewer physical threats.

I think #MeToo and #BlackLivesMatter

are good examples of this.

But, on the other hand, reduced cost of expression

might also in some cases lead

to disproportionate levels of punishment.

Finally, the way that outcomes are delivered on social media

might make our expressions of outrage

more mindless or habitual.

So, responding to moral violations with outrage

is the best way to get likes and shares,

and these social rewards

are delivered at unpredictable times.

A pattern of reward delivery that's known

as variable reinforcement,

and this is well-known to promote the formation of habits.

Casinos use this kind of reward delivery pattern

to keep gamblers hooked on slot machines.

So habits are behaviors that are expressed

without concern for their consequences.

So just as a habitual snacker

reaches for the bag of chips

without actually feeling hungry,

a habitual online shamer might express outrage

without actually feeling very outraged.

In my lab, we've been exploring all these questions.

And to do this, we needed to build a new tool

for measuring outrage on social media.

We've been collecting tens of millions of Tweets

from a variety of episodes of viral public outrage,

including the Covington Catholic example I opened with,

as well as the Brett Kavanaugh Supreme Court hearings,

Jussie Smollett's faked hate-crime by Trump supporters,

Trump's decision to ban transgender individuals

from the military,

and United Airlines violently dragging a passenger

off an overbooked flight.

With this data, we've been building an AI

that can measure expressions of outrage online.

We're calling it the "Digital Outrage Classifier,"

and we can feed it millions of Tweets, and it will determine

whether each Tweet expresses outrage, or not.

Here are a couple of examples of Tweets

categorized by our classifier

from the Kavanaugh Supreme Court hearings last year.

This one, it says, is outrage.

And this one, it says, is not.

It does pretty well.

Using this classifier, we've been testing a hypothesis

that the more your social network

rewards you for expressing outrage,

the more likes and shares you get,

the more likely you then become

to express outrage in the future.

So in other words, if you express outrage today in a Tweet,

and you get a lot of likes and shares for it,

does that make you more likely to express outrage tomorrow?

Now, this work is still ongoing, it's in progress,

so I can't share just any more with you today,

but so far, the data are very consistent

with our hypotheses.

And, so for the rest of my time here, I'm gonna zoom out,

and talk a little bit about some of the implications

should our predictions continue to be born out in our data.

Now, recall that outrage has both social

and personal benefits.

We think that social media might be amplifying

some of the personal benefits,

while at the same time potentially reducing

some of the social benefits.

It amplifies the personal benefits

by dramatically extending the reach of moral signals.

So, for example, if you hear somebody make a sexist comment,

and you tell them off for it in person,

you only get credit for that

from whoever happens to be watching at the moment,

which is probably not that many people.

But, if you express outrage

about that same sexist remark on social media,

you broadcast your moral character

to your entire social network.

That might be thousands or even millions of people.

And you can expect a big pile of likes

and re-Tweets as other people want to chime in

and signal that they agree with you.

And, we think this is especially true in a political context

where people want to signal

that they are good tribe members to the rest of the tribe.

But, personal benefits of outrage aside,

it remains an open question whether digital outrage

is good for society.

One potential drawback is a worse ability

to coordinate our punishment.

If one function of outrage is to point out

those actions or people who are deserving of punishment,

lowering the threshold for expressing outrage

could muddy that signal,

making it more difficult for people to align their behavior.

If everything is worthy of outrage, effectively, nothing is.

Another potential bad consequence is that outrage

might be facilitating the spread

of disinformation and propaganda online.

Bad actors who want to spread this kind of content

seem to know either intuitively,

or explicitly, that newsfeed algorithms

are prioritizing outrageous content.

So here are a few examples of Facebook ads

sponsored by the Russian government in 2016.

You can see how much moral-emotional language,

and outrage-provoking language are in these ads.

We're currently testing in my lab whether outrage

is indeed more prevalent

in false news, compared to real news.

And if this is true, outrage expressions

could serve as a potential behavioral marker

for disinformation and propaganda

that's at risk of going viral.

Finally, we're looking at whether this social reinforcement

of outrage online

might explain the rise in political polarization

that we've seen in recent years.

And there are a couple of different ways

that this might happen.

First of all, if more politically-extreme users

are expressing more outrage

than politically-moderate users,

and these expressions get amplified

by the newsfeed algorithms,

then more moderate users might feel pressured to chime in.

And then of course, they'll get rewarded

by their social network.

If they internalize these rewards, then over time,

this could push people in the middle towards the extremes.

But people don't even necessarily need to feel more outrage

to appear polarized to their peers.

If social media decouples

the expression of outrage from its experience,

people might falsely believe

that others are more outraged than they actually are,

and be less-willing to engage in bipartisan discussions

on the basis of these false beliefs.

Our data so far suggests that if moral outrage is a fire,

social media is like gasoline.

Technology companies have been arguing that their products

provide neutral platforms for social behaviors,

but don't fundamentally change those behaviors.

Of course, this is an empirical question,

and the data that we need to answer it already exists.

But not all of it is publicly available.

These data can and should be used

to understand how new technologies like social media

might be transforming moral emotions

from a force for collective good,

into a tool for collective self-destruction.

We'll be releasing more results soon

at digitaloutrageproject.org.

You can sign up for our mailing list

if you want to follow along.

I want to thank our team:

Billy Brady, a post-doc in my lab who's leading this work.

Killian McLoughlin, our research technician.

Tuan Doan, a Yale undergraduate thesis student.

Our funders, NSF, Democracy Fund,

Social Science One,

and our volunteers and colleagues

who have given us feedback on our work,

and of course you for your attention.

Thank you. (audience applauding)