Are Echo Chambers a Threat to Intellectual Freedom?
Do echo chambers exist? What drives them?
Posted March 30, 2019
A feature of digital communication that has provoked concern is the presumed existence of echo chambers. An echo chamber, in this context, is taken to be an environment where people experience only similar opinions to their own, and where contradictory or alternative opinions are not considered. Their existence has been taken to be a threat to debate and intellectual development by many: “The echo chamber may be comforting, but ultimately it locks us into perpetual tribalism, and does tangible damage to our understanding.”1 That is to say, the existence of the echo chamber is taken to limit our freedom to learn and to tolerate differences of opinion.
Several recent articles have begun to analyse the degree to which such presumed echo chamber effects actually occur in digital technologies.2,3 Some articles, including several from the BBC, have even doubted their existence and/or importance:4-6 “But this claim - that our digital media habits are an exercise in the lazy endorsement of our prejudices - is itself lazy and full of prejudice.”6 In the light of the concerns about echo chambers promoting tribalism, this polarisation in views about the existence or ubiquity of echo chambers may seem somewhat ironic.
Which of these positions appears closest to the truth, and what does the evidence from studies about behaviour and cognition tell us about these effects? It turns out that the situation is not straightforward, and depends as much on the psychology of the user, as it does on the structure of the digital environment. Ultimately, principles of reinforcement developed in the pre-digital age may have much to say about the nature of the echo chamber.
One line of work that has led to a re-evaluation of the harms that echo chambers may perpetrate is based on analyses of the types of exposure people have to different political information sources. A recent survey of about 2,000 internet users3 noted that those who rated themselves as being more interested in politics, and those who rated themselves as having contact with a greater range of media, also rated themselves as avoiding echo chambers. The authors concluded that: "This work challenges the impact of echo chambers and tempers fears of partisan segregation since only a small segment of the population are likely to find themselves in an echo chamber.”3
This work suggests that those with an interest in, and perhaps understanding of, their chosen area to begin with, and those who have a greater mastery of the digital world, will be more effective at sampling a wider range of opinions, and more able to avoid echo chambers, than those who do not have those abilities. These results fit with the view that the digital world is most risky for those who do not really know what they are looking for, or doing. When a person lacks the skills to navigate an environment, or, indeed, lacks a full understanding of that environment, itself, then they will experience a good deal of cognitive stress. They will have to accomplish two tasks at once: searching for information and learning how to search for information. From a range of studies, we know that this is one of the very situations that will lead to the development of tunnel vision in what is searched for, perhaps in order to cope with this complexity and strain7.
These conclusions regarding the importance of echo chambers3 are based on self-reports from individuals using digital technology. Such self-reports may well be informative, but they are also subject to bias on the part of the informant – either conscious or unconscious – and may not reflect the reality of the situation. As a consequence, arguments, based on these types of data, suggesting that digital communication networks may be safer from echo chamber effects than previously thought5,6, should be treated cautiously.
Perhaps a more telling line of evidence regarding the nature and influence of echo chambers comes from the analyses of actual social media behaviours.2,4 One recent study analysed the entire Twitter network over two weeks, and explored how news was shared across users.4 This study found that users accessed and shared a wider range of opinions than predicted from a view that echo chambers were the standard mode of operating on social media. The majority of Twitter users were not polarised in their views, and shared a greater amount of "politically-moderate" content than they sampled. However, it was also noted that the most active 1 percent of users did display highly polarised views – posting and sharing more extreme content than they read – and these users were followed more.
A similar study also explored the impact of polarisation, and found this was an important factor in the way in which information was spread2. This study examined the online behaviours of 250,000 Twitter users. In contrast to the above study, the users were about evenly divided in terms of whether they were polarised or not. The reasons for this discrepancy are unclear but probably reflect different criteria being applied to define this concept. The existence of echo chambers was pronounced among these users, in that there was a: “…strong correlation between the polarization of a user and the average polarization of both his/her nearest neighbors and the received tweets.”2 Moreover, those with high levels of polarisation tended to spread information in support of their views much more than those with less polarised views – giving them greater influence.
Why does this effect occur? There are many potential reasons. It may have to do with the algorithms used by social media to signal content of potential interest, or receiving opinions that are similar to one’s own is reinforcing – it supports and generates such behaviour.
A study conducted prior to the development of digital communication networks is relevant in understanding this phenomenon. In 1955, Verplanck conducted an experiment8 and noted that: “…if…one agrees with opinions expressed by a speaker, the speaker will give still more opinions, and that returning the speaker's words in paraphrase has the same effect.” That is to say, agreeing with somebody, and restating or reiterating their views, will increase the frequency with which opinions, and especially opinions of that sort, are expressed – the echo amplifies the original signal!
There are individual differences in the degree to which this reinforcement mechanism will be effective. Verplanck noted that those who express lots of opinions, and who tend to make these opinions short and pithy, receive more reinforcement, and consequently increase their rate of opining (they become more polarised). Those who emit opinions less often, and express longer and more nuanced views, will not get this level of reinforcement. As the internet favours the former approach, then its structures may be well suited to promote the existence of echo chambers. When people did not get an "echo" of agreement in Verplanck’s study, they would either simply restate their position, change the topic until they got an agreement, or leave the situation altogether.
The available evidence does suggest that echo chambers exist, although not all digital communication occurs within them, but the research suggests that they have a disproportionate influence on the spread of opinions, and may well be a product of the type of reinforcement experienced and delivered digitally. They will generate a reduced range of more polarised views, and they cannot be ignored – even if that is an uncomfortable fact for media companies keen to promote themselves as bastions of free speech.
1. The Guardian (4.12.17). Echo chambers are dangerous – we must try to break free of our online bubbles. https://www.theguardian.com/science/blog/2017/dec/04/echo-chambers-are-…
2. Cota, W., Ferreira, S.C., Pastor-Satorras, R., & Starnini, M. (2019). Quantifying echo chamber effects in information spreading over political communication networks. arXiv preprint arXiv:1901.03688.
3. Dubois, E., & Blank, G. (2018). The echo chamber is overstated: the moderating effect of political interest and diverse media. Information, Communication & Society, 21(5), 729-745.
4. Shore, J., Baek, J., & Dellarocas, C. (2018) Twitter is not the echo chamber we think it is. FRONTIERS. MIT Press. https://sloanreview.mit.edu/article/twitter-is-not-the-echo-chamber-we-…
5. BBC (17.4.18). The myth of the online echo chamber. http://www.bbc.com/future/story/20180416-the-myth-of-the-online-echo-ch…
6. BBC (4.3.19). Do digital echo chambers exist? https://www.bbc.co.uk/news/entertainment-arts-47447633
7. Reed, P. (2019). Does gaining information from digital media require freedom? Psychology Today. https://www.psychologytoday.com/intl/blog/digital-world-real-world/2019…
8. Verplanck, W.S. (1955). The control of the content of conversation: reinforcement of statements of opinion. The Journal of Abnormal and Social Psychology, 51(3), 668.