Thread by @M_B_Petersen

The final keynote of #bystander22 is by @jayvanbavel on "Morality in the Anthropocene".

We evolved in small-scale groups but is now connected in vast, social media network. What are the consequences? 1/17

As EO Wilson noted: "We have paleolithic emotions, medieval institutions and godlike technology". How can we manage our ancient emotions in the modern technology environment? Contrast with Zuckerberg's notion of "moving fast and breaking things". 2/17

57 % of all humans use social media. The average user uses it 2.2 hours per day. On average, 5.7 years of your life will be used on social media. All this has emerged fast. 3/17

Social media is an attention economy. What drives engagement - and what are the effects of this?

We don't know. Science is lagging behind the technological innovation. Data is lacking. knowledge is developed in segregated communities. This is a true interdisciplinary task. 4/17

But we know that the structure of social networks and their info flows are engineered to maximize profitability. And we know (from whistleblowers) that CEOs do not react on signals of problems. The focus is on profit. 5/17

Crockett (2017) showed that people mainly learned about moral and immoral acts from online sources. More so than from in-person sources and TV and print media. What we especially learn from online sources is about *immoral* acts. 6/17

Online sources trigger more moral outrage than legacy media & in-person sources. This is not always bad. Examples are #metoo & #blm. Another early example is Arab Spring, which was coordinated on social media. Also, dissent in Russia today is a current example. 7/17

But there are anti-democratic examples too, e.g., the Jan 6 insurrection.

The power of social media is that info can travel rapidly via loose networks. These collective processes can mobilize individuals. 8/17

Moojiman et al. show that moral rhetoric on social media can predict eruption of offline violence. 9/17

Why is this messaging spreading so fast?

The process is this: You read something which triggers you. You can quickly re-share, which triggers others. This create cascades.

These cascades are problematic when the info is false or harassing. 10/17

The three important factors according to The MAD Model are:

Motivation (e.g., group identity) - Attention (creating engagement) - Design (e.g., algorithms that amplifies certain content) 11/17

These features interact to shape what people share.

For example, when sharing an article, particular sentences are selected to maximizes attention, even if they are not representative. These may create a competition for attention by sharing more extreme content. 12/17

In a 2017-study, it was found that moral words in online words increased sharing dramatically. Emotions alone doesn't do this. It is the combination of morality and emotionality that drives online sharing. A later meta-analysis provides further support. 13/17

Why are moral emotions being spread? Using the "attentional blink task" in a 2020-paper, it is found that these words capture attention more. In a follow-up study, it was found that words that capture more attention in the lab are more likely to be retweeted on Twitter. 14/17

Importantly, the larger sharing of moral-emotional words only reflects increased sharing among like-minded people, i.e., the ingroup. This "echo chamber effet" isn't found for other types of language. This is both confirmed on Twitter and using experiments in the lab. 15/17

In fact, when people use more moral emotional words, people from the other group reacts negatively. 16/17

Alot of the research has been done in the US. We need a more global sense of what happens. 17/17