All networks can be categorised as centralised, decentralised, or distributed. This depends on how they function: if the nodes branch from one single central point, the network is centralised; if there are multiple points, with nodes branching from them, the network is decentralised; if each node connects from one to the other, creating an infinite pattern of connection, it is a distributed network. In this blog, I will be focusing on centralised networks and the rise of censorship on social media, as well as discussing how decentralised networks are more user friendly.
Centralised networks are most popular because they are the most administrator-friendly network model. They were initially created “as a way to improve efficiency and take advantage of potential economies of scale.” (N-able, 2018) They are the most efficient when considering high-scale user database and quick growth (Goyal, 2015), perfect for social media.
For example, let’s look at Facebook. With Facebook, updates become accessible to all users at once, rather than administrators having to go through and update each and every device connected. (Oxen, 2019) They all link back to one central point, so when that central point puts out an update, all devices immediately connect to that update. However, this poses an issue when there is a server crash. If that single central point crashes, all users connected are unable to access the platform (Oxen, 2019). It is believed that this is the best option for such an inclusive site like Facebook, where anyone can join, since “Without a central node…it is less easy to hold someone responsible” (Stalder, 2005), meaning that it may be for everyone, but there is still monitoring involved to ensure safety. This can be seen as a blessing and a curse, because some censorship is important for the wellbeing of users, but it also allows those in power to abuse people’s rights online and can often backfire, targeting the same people these platforms swear to protect.
Social media and the wider internet has become more censored nowadays. The best example of this is Donald Trump being banned on a variety of platforms such as Twitter, Instagram, Reddit, and many more (Tannenbaum, 2021). Of course, with so many platforms against Trump, it could be said that his actions warrant this type of response. But it is this mentality that then paves a way for apps like Parler, which is a self-proclaimed “free speech” app, to do the same. “While mainstream platforms have become more stringent about their moderation in recent days, some claim that stricter rules ultimately push users to darker corners of the internet.” (Heilweil, 2021) While branding itself as a free speech app, Parler has banned a number of users (including yours truly) for presenting an opposing view to that of the conservative American, leaving it just as censored and biased as it’s opposition.
Centralised networks allow for this type of bias on social media, and it seems as though it is only escalating. An example of this is with TikTok and their stance on the Black Lives Matter movement. TikTok is outspoken in their support of Black Lives Matter. However, users find that when they are addressing issues related to this cause, their videos are being removed. The claims by TikTok is that there are certain phrases that are already blocked, and that “Hate speech and talking about hate speech can look very similar to an algorithm” (Ohlheiser, 2021). This then becomes counterproductive, because while TikTok is trying to create a safe space for black people by censoring words and phrases, they are actually silencing their voices.
This is common among other platforms, such as YouTube, which demonetises videos that use specific words or phrases that don’t align with the site’s values. If a victim of sexual assault were to come out with their story on YouTube, they wouldn’t even be allowed to use the words “sexual assault”. Users are forced to spread their stories across multiple platforms, each one for a different story. This interferes with the image of global connection and the inclusivity of the internet, as rather than people joining a single social media app, they are spreading themselves across a variety—all which ultimately function the same way. Centralised networks allow for this, because it’s easy to manage words when the censorship comes from a single hub.
While centralised networks and censorship seem to only be increasing, I still think social media users will benefit largely from decentralised platforms. They’re more reliable, having multiple centres, so if one server were to crash or get hacked it would not bring the whole system down, and users will still be able to access the platform (Oxen, 2019). And with multiple servers running the network, this would make it difficult for one person to corrupt all users’ privacy (Oxen, 2019). This would also interfere with censorship bias and allow users to access a platform without the increasing suspension of profiles or removal of content. Sure, it will be more work for those running the platforms, but it means censoring will be intentional.
References:
Goyal, S, 2015. ‘Centralized vs Decentralized vs Distributed’
Heilweil, R, 2021. ‘Parler, the “free speech” social network, explained’
N-able, 2018. ‘Centralized Networks vs Decentralized Networks’
Ohlheiser, A, 2021. ‘Welcome to TikTok’s endless cycle of censorship and mistakes’
Oxen, 2019. ‘Centralized vs Decentralized Networks’
Stalder, F, 2005. ‘Open Cultures and the Nature of Networks’
Tannenbaum, E, 2021. ‘Every Social Media Platform Donald Trump Is Banned From Using (So Far)’
Candace 