What’s the recipe?
In his book, Alberto pointed out that information can now be passed on, and ultimately become cultural, without being memorized or even understood. But online information can also become cultural without catching people’s attention or even being shared at all. I will argue that the weight of these two parameters, and more broadly the type of content that will go viral online, hinges on: (i) the enunciation sphere (e.g. private or public), (ii) people’s use of the online platform (e.g. networking or reinforcing existing bonds), (iii) reputation management strategies (e.g. appearing warm or competent ), (iv) the platforms’ affordances (e.g. short messages or photos), and (v) what algorithms promote (e.g. shared or watched content).
Let’s (not) talk about sex
In “Cognitive Attraction and Online Misinformation”, Alberto convincingly argued that misinformation spreads not so much by mistake, but in virtue of its psychological appeal (Acerbi, 2019). He showed that fake news is full of psychological factors of attraction, such as information related to threat, disgust, or sex. Recently, with Manon Berriche, we looked at how well-established factors of attraction predicted user interactions (likes, comments, and shares) on an extremely successful Facebook page with the bad habit of sharing health misinformation from time to time (Berriche & Altay, 2020). One interesting finding for cultural evolutionists is that sex-related information did very poorly, with people avoiding interacting with sex-related information more than any other type of content. At first, I was puzzled, given that their attractiveness is manifest. But then I thought about how often I share sex-related content, and saw the elephant in the room: sex-related information is extremely attention grabbing but is inappropriate to share publicly. Thus, some content can be eye-catching without ever being visibly successful. As it was before the digital age, but we can now identify this kind of content with greater ease.
The social news gap
There is a well-known gap between what people read and what they share (Bright, 2016). Sex-related information and crime stories are guilty pleasures that people read a lot privately and yet do not advertise publicly, as it would negatively affect their reputation. Conversely, other content, such as science and technology news, “have levels of sharing that are disproportionately high compared to their readership” (Bright 2016, p 357). Science news is not very psychologically appealing to most people, but sharing it can signal one’s competence, as it suggests that one has the background knowledge and ability to understand technical content. In the same vein, phatic posts (i.e. statements with no practical information fulfilling a social function such as “I love my mom”) are viral on Facebook not due to the psychological attractiveness of their content, but because they allow users to reinforce bonds with their peers and signal how warm and loving they are (for more details see: Berriche & Altay, 2020). Thus, some very psychologically unappealing content can spread by virtue of its instrumental value (i.e. enabling people to convey desired impressions).
The instrumental value of a piece of information depends on one’s goal and audience. For instance, economic news are more shared on LinkedIn than on Facebook (Bright, 2016). Sharing business news on LinkedIn can help signal one’s competence to potential employers, whereas on Facebook, which people use to bond with peers and relatives, the instrumental value of business news is much lower. A piece of information can even have a negative instrumental value if shared in the wrong sphere, such as sharing phatic posts on LinkedIn instead of Facebook, or posting sex-related information on Facebook rather than on a private WhatsApp group chat.
People use different social media platforms to express different facet of their personality and fulfil distinct goals. Taking into account these parameters can help generate new hypotheses. For instance, on platforms that people use to appear competent, information should spread faster, as the primer for being the first to say something will be higher. And indeed, information spreads faster on LinkedIn and Twitter than on Facebook (Bright, 2016). As Hugo Mercier suggested, this dynamic could help understand why Twitter was one of the first to sound the alarm about the COVID-19 pandemic. Sharing threat-related information has a higher instrumental value for Twitter users since it allows them to display their competence (Boyer & Parren, 2015), whereas it doesn’t help to appear friendly or foster relationships (something with a high instrumental value for Facebook users).
The structure of online platforms and the possibilities for action that they offer (affordances) shape users’ utilization of these platforms. For example, Instagram is designed to facilitate picture editing, posting and sharing, while Twitter is news and text oriented. One will use Instagram to show off their summer body, and Twitter to share their brightest thoughts. From the platforms’ initial structure, user-based innovations will emerge and help users better satisfy their goals. For instance, to overcome Twitter’s initial 140 character limit, a small number of expert users started connecting series of tweet together, creating “tweetstorms”, that later became known as Twitter threads. This type of innovation widens the field of possibilities on these platforms and subsequently influences what will become culturally successful.
Apparently similar platform structures can hide disparities. On YouTube, Facebook, or Twitter, users can “like” content with a thumb up or a heart (if you’re a fancy scholar you can also call them paralinguistic digital affordances; Hayes et al., 2016). Yet, these likes don’t mean the same thing and are not used for the same reason across platforms. On Twitter a like is mostly a way to archive posts, share content with followers and signal that one enjoyed the post’s content (Hayes et al., 2016). On Facebook, likes have a strong phatic function, and are used to say “hi” to the poster or show one’s support (Hayes et al., 2016). On YouTube, where the like is private, it is a signal sent to the algorithm, either to have similar videos recommended, or to support the YouTube channel that posted the video. These subtleties need to be taken into account to predict information’s virality, as metrics of cultural success don’t mean the same thing everywhere on the web. Although in the end, a large share of the variance in online information’s cultural success might come down to what the algorithms promote.
The new black box
On Twitter and Facebook sharing is a necessary component for being culturally successful, while being attention grabbing is less rewarded, as the algorithm doesn’t promote content that people open but rather content that people publicly interact with (shares, likes and comments). On YouTube being attention grabbing is key, whereas being shared is secondary. Interestingly, YouTube’s algorithm doesn’t promote catchy videos that people open and close after a few seconds, but instead videos that captivate the audience’s attention until the end. This feature of the algorithm spurs innovation and favors techniques that incentivize viewers to watch the entire video, such as letting viewers know that relevant information will be revealed at the end, or creating short, eye-catching videos. Similarly, on Spotify, artists are incentivized to make short songs with very catchy beginnings given that they are remunerated every time one of their songs is played for more than 30 seconds.
A major obstacle for cultural evolutionists in the digital age is that these algorithms are mostly opaque, are being changed without notice, and promote very different types of content across platforms. In other words, the number of recipes to become cultural has risen sharply, and these recipes are being continuously edited whilst carefully hidden away. Luckily, the digital age also offers the opportunity to make finer-grained analyses of cultural transmission by incorporating media structures, communicative contexts, and reputation management strategies.
 The distinction between warmth and competence is synonymous to the one made between “qualities” and “tendencies” in the partner choice literature (Barclay, 2013).
Acerbi, A. (2019). Cognitive attraction and online misinformation. Palgrave Communications, 5(1), 15.
Barclay, P. (2013). Strategies for cooperation in biological markets, especially for humans. Evolution and Human Behavior, 34(3), 164–175. https://doi.org/10.1016/j.evolhumbehav.2013.02.002
Berriche, M., & Altay, S. (2020). Internet Users Engage More With Phatic Posts Than With Health Misinformation On Facebook. Palgrave Communications.
Boyer, P., & Parren, N. (2015). Threat-related information suggests competence: A possible factor in the spread of rumors. PloS One, 10(6), e0128421.
Bright, J. (2016). The social news gap: How news reading and news sharing diverge. Journal of Communication, 66(3), 343–365.
Hayes, R. A., Carr, C. T., & Wohn, D. Y. (2016). One click, many meanings: Interpreting paralinguistic digital affordances in social media. Journal of Broadcasting & Electronic Media, 60(1), 171–187.