Did technology disrupt the truth? Probably not

One of the articles of my holiday-accumulated reading list was, given my current interest in the effects of digital media on cultural transmission and evolution (see here), Katherine Viner’s How technology disrupted the truth, a long read of the Guardian. The piece got extensive – and almost exclusively positive – attention (there are, when I write, 1,589 comments and around 64,000 shares). In fact, I found it quite hideous, and I believe it also embodies a widespread common-sense attitude towards digital technologies, and social media in particular, so I decided to write here some comments.


The article starts with a series of recent examples of spreadings of uncontrolled rumours, or plain falsehoods. A first important thing to notice is that none of these rumours/falsehoods originated in social media, or was linked to social media activity in any special way. David Cameron’s “obscene act with a dead pig’s head” (which, by the way, I never heard of, probably because never appeared in my social media stream), started from the Daily Mail, that took it from “a new biography of Cameron”. Farage’s claim that Brexit would have allowed £350m a week spare to spend on the NHS was widely reported by mainstream media. Michael Gove compared the experts warning about economical damages of Brexit to Nazis who smeared Albert Einstein during an interview with LBC Radio, and, again, this was reported by newspapers and televisions.

Vines rhetorically asks: “Does the truth matter any more?”, and of course her answer is negative. However, while it is fashionable to claim that we live in a “post-truth” era, have we ever been in a “truth” era? My first contacts with politics and media, as Italian, were with Silvio Berlusconi, that, in 1994, infamously promised “1 million new jobs” in his first campaign. Facebook was not even imagined at the moment, and in fact Berlusconi’s successes were ascribed to the control of traditional media, especially private televisions.


I imagine political campaigns have never been fair and moderate affairs (I can not resist to cite the not-very-relevant fact that some consider the death of Edgar Allan Poe due to over drinking because of cooping, i.e. the practice of forcing bystanders to vote, likely repeatedly, and possibly by offering alcoholic beverages), but my point is that, to claim that we live in a post-truth era, one can not simply present contemporary examples of rumours and falsehoods, but should show that there were less of them before, which is, in my opinion, not self-evident.

Moving to Cameron’s special interest for animals’ heads, it is also dubious whether these sorts of topics are more common nowadays than before. Like politicians’ lies, rumours involving sex or emotion – especially related to disgust – seem to have a long history (if you can combine the two then you hit the jackpot). In this case, there is a growing literature in cultural evolution on how cultural traits with particular contents tend to be more successful than others (in academic lingo these are called content biases). This is a recent – not yet published – paper from Joseph Stubbersfield, Jamie Tehrani and Emma Flynn, with a good review of the previous literature and a new experiment. While the details might be disputable (I have myself a submitted paper in which, together with Olivier Morin, we examine an apparent two-century long decrease of emotional expression in English written fiction), there is a clear consensus that some content, all else being equal, spreads better than other. Interestingly, this literature refers in general to oral transmission (in particular laboratory experiments in which participants are asked to listen to a story, and then evaluate it, or repeat it, or similar) or to mainstream media as books, newspaper articles, or collections of urban legends. Hence, quite the contrary of what is implied in the Guardian article, we know that this content is successful outside of social media, but we do not what differences social media generates. In fact, as I mention here, it would be interesting to investigate what are the effects of error-free, instantaneous, transmission allowed by digital media in respect to other forms of “traditional” transmission, but this is indeed an open problem.

So, it is at least questionable whether or not we live in a post-truth and sex-and-disgust-addicted era, but one might legitimately ask if social media contributes to worsen the situation, leading us in a more post-truth and sex-and-disgust-addicted era in respect to what was before. As I just mentioned, we do not know much on how different contents spread in different media, but we know, because it is repeated everywhere, that social media creates echo chambers, where information are amplified and reinforced inside an enclosed system, and different or competing views are censored (more or less defined like that in wikipedia).

Again, I think this is an interesting issue, but I also think that many commentators use social media as a convenient scapegoat. As I mumbled before, the important question is not whether social media creates or not echo chambers (yes, they do), but whether and how the effect is different from other media, or from daily life (how much diversity of opinions does one have in the workplace or in the local pub, and how much does this influence her/him?).  Studies on polarisation and echo chambers on social media, such as this one or this one, have, in my opinion, a quite limited interest, a part offering possibly valuable quantitative data. If one pre-select a community of Facebook users interested in scientific news and one of Facebook users interested in conspiracy theories is not very surprising that “users belonging to different communities tend not to interact and that they tend to be connected only with like-minded people” and “to focus their attention exclusively on one of the two types of information”. The authors of the paper claim to be careful when extrapolating general conclusions (they explicitly write “We emphasize that other kinds of data sets may not show the particular patterns that we observe here”), but they still do it, and, of course, media tend to be less cautious.

When social media polarisation is explicitly compared with polarisation in real life, the results are, to say the least, less clear. This study, from political scientist Pablo Barbera, for example, found that social media reduces polarisation. The logic is that:

social media platforms facilitate exposure to messages from those with whom individuals have weak ties, which are more likely to provide novel information to which individuals would not be exposed otherwise through offline interactions

which is what I myself suspected when I wrote about my politically schizophrenic twitter feed. In another example, the authors of this preprint conclude their abstract writing:

contrary to prediction, we find that the average account posts links to more politically moderate news sources than the ones they receive in their own feed. However, members of a tiny network core do exhibit cross-sectional evidence of polarization and are responsible for the majority of tweets received overall due to their popularity and activity, which could explain the widespread perception of polarization on social media.


A related, but different, problem (the two seem to be conflated in the Guardian article), is the fact that social media (in particular: Facebook) uses notoriously an algorithm to select the updates one will prominently see in the newsfeed. Facebook is not supposed to have a political agenda (however see here), but only tries to optimise its revenue, that is, the time one spends on the social media itself. It seems that, at the moment, the stories that appear more in the newsfeed are the stories posted by the contacts one interacts more with. So, if you like or click the posts of your friend Thomasine, the new posts of Thomasine will be more likely to appear in your newsfeed. This seems quite reasonable, but it reinforces the echo chamber phenomenon described above (“your News Feed will almost certainly become a place of even greater comfort and conformity” as stated, sarcastically, here). In addition, it has been noted (correctly) that “optimising for engagement” is dangerous, as stories that obtain many likes will tend to be attention-catching, or sensationalists, lacking in depth and moderate perspective, etc. (clickbait is the term).

But: is this related to social media? Is this not present in mainstream/traditional media? (I also want my share of rhetorical questions). As above, my impression is that this is an open problem. In addition, as social media only tries to optimise engagement, the clickbait effect (or the spreading of Cameron’s pig head story) depends, ultimately, on the behaviour of users. As I mentioned in a previous post, my Facebook newsfeed tend to be, mainly, a boring series of links to scientific papers and scientific news (you can try the thrill of my timeline, if you like…). Perhaps instead of juxtaposing digital media and traditional media one should understand why many people want to see and read what they want to see and read, instead of reading the measured analysis of the Guardian. Of course this does not mean that one should accept the situation like it is: informing users about the functioning of the social media (including, whatever possible, a basic understanding of the algorithms working under the hood), perhaps actively encouraging people to embrace diversity in their networks, and even thinking about creating/supporting different social media that, for example, will give more freedom on how information are presented, are all worthwhile and important projects which I would happily be involved with.

Finally, while the complaint about the state of the internet has recently been politically variegate (see for example Freakonomics – or even Matt Ridley that jumped on the bandwagon), the Guardian sees this as producing very specific consequences. Katherine Viner writes:

As the academic Zeynep Tufekci argued in an essay earlier this year, the rise of Trump “is actually a symptom of the mass media’s growing weakness, especially in controlling the limits of what it is acceptable to say”. (A similar case could be made for the Brexit campaign.)

In sum, Britain will be out of the EU because of Facebook. These are complicated matters, I reckon, but I would like to just present one piece of information. One of the clearest voting pattern of the British EU referendum was that people that voted “Leave” where in general older, and less educated, than people that voted “Remain” (details are complicated, of course: if I remember correctly,  young people were also the category with more non-voters). I believe – but I might be wrong – that a similar pattern could be found for Trump voters. As it happens, older and less educated people are also the people that use less social media and the internet. Saying that “social media usage favours Trump/Brexit opposition”   (admittedly, a silly sentence) seems to me slightly more correct than the common-sense opinion, embraced by Tufekci and the Guardian.


Some commentators gave what I believe is a charitable interpretation of Katherine Viner’s article (see for example Le Monde’s Les médias dans l’ère «de la politique post-vérité»), interpreting it mainly as a call to traditional press to assume a role of a trustworthy authoritative voice, by explaining, verifying, debunking – in case – rumours, drawing mainly on the last few paragraphs of the Guardian piece. All good, but a fairer summary would be “people spend time with social media (instead of spending it with traditional newspapers), this clouds their minds, and they end up to vote “yes” when asked if Britain should leave EU. Also, we are losing readers, so please help us!”

I believe that instead of longing for a (never-existed, and, if you ask me, not too palatable) world in which all citizens would read supposedly unbiased, but certainly centralised and top-down, media, and then take rational decisions, one should understand and accept technological and cultural evolution (as well as human psychology), and use this understanding to try and bend if for the better.

[If you arrived here: I wrote, after the post, a review paper on A cultural evolution approach to digital media]


Acerbi, A. (2016), A cultural evolution approach to digital mediaFrontiers in Human Neuroscience, 10, 636

Barberá, P. (2015), How Social Media Reduces Mass Political Polarization. Evidence from Germany, Spain, and the U.S., paper presented at the APSA 2015 Conference

Morin, O., Acerbi, A. (2016), Birth of the Cool. A two-centuries decline in emotional expression in Anglophone fictionCognition and Emotion, 1-13

Quattrociocchi, W., Scala, A., Sunstein, C.R. (2016), Echo chambers of FacebookSSRN.

Shore, J., Baek, J., Dellarocas, C. (2016), Network structure and patterns of information diversity on Twitter, arXiv, 1607.06795

Stubbersfield, J.M., Tehrani, J., Flynn, E.G. (2017), Chicken Tumours and a Fishy Revenge: Evidence for Emotional Content Bias in the Cultural Transmission of Urban Legends, Journal of Cognition and Culture, 17.


One thought on “Did technology disrupt the truth? Probably not

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s