Bits and bytes of the decentralised social web
An ethnographic exploration of a contested infrastructure
Ia Croxton (11230614)
Supervisor: Yatun Sastramidjaja MSc Cultural and Social Anthropology Department of Anthropology Ia Croxton (11230614)
Abstract
In an age of concern about the social and political consequences of an increasingly
centralised internet, there are various movements around the world aiming to counteract this process under the guise of ‘re-decentralising the web’. This project ethnographically explores a subset of these projects – the decentralised social web – an infrastructure composed of various projects which act as alternatives to mainstream social networks. The writer argues that the decentralised social web is an infrastructure that acts as an “ontological experiment” (Jensen et al., 2015), whose political nature stems from its ontology as a site of complex interaction between various things.
Acknowledgements
I would like to thank everyone I met during my time in Amsterdam. To everyone who was willing to allow me into the world for a few months – thank you for teaching me so much. My course mates, friends, family, and Jonny – thank you for supporting me throughout this process. Most of all, thank you Yatun so for much help and guidance, and especially your patience with me throughout this process.
Table of Contents
Abstract ... 2
Acknowledgements ... 3
I would like to thank everyone I met during my time in Amsterdam. To everyone who was willing to allow me into the world for a few months – thank you for teaching me so much. My course mates, friends, family, and Jonny – thank you for supporting me throughout this process. Most of all, thank you Yatun so for much help and guidance, and especially your patience with me throughout this process. ... 3
Preamble ... 5
Introduction ... 5
Theoretical Framework ... 8
The materiality of the web ... 8
Agency, control and resistance... 9
Freedom technologists: the internet and activism ... 10
From theory to method ... 11
Method in practice ... 13
Research question ... 14
Chapter 1: An apophatic infrastructure – on what the decentralised social web isn’t ... 16
It’s not centralised ... 16
It’s not Facebook ... 20
It’s not capitalist ... 23
Conclusion ... 26
Chapter 2: Intentional infrastructure – on political intention ... 27
Transforming materials: squatting and hacking ... 27
Social justice on the web ... 29
Who is the ‘enemy’: Governments? Capitalists? ... 30
Conclusion ... 34
Chapter 3: Unintentional infrastructure – on the politics of things ... 35
It’s an experiment in anarchy ... 35
A battle of protocols ... 36
Inequality in networks ... 37
A social world of ones and zeros ... 37
It’s built to be an alternative ... 40
Conclusion ... 41
Conclusion ... 42
Preamble
“Alright, okay, you want to know why I'm really interested in user freedom stuff and
everything? And if I didn't sound enough like a space cadet, here is the real thing. One of my friends […] gives a talk about why she is interested in user freedom is because she has a pacemaker and she jokes that she is a cyborg lawyer. She is the creative director of the
software freedom conservancy. […] When she got her pacemaker implanted she wanted to be able to read the source code to the pacemaker and be able to have herself and her friends be able to make sure that it was safe and secure. She tried to go to the manufacturers who were like ‘why would you need to do that, you don't need to do that?’ Well it turns out there are massive vulnerabilities in pacemakers, many of which can be updated wirelessly. It turns out you can do an exploit where you can log in to somebody's heart and kill them over a WiFi network. […]
So the thing to me is, if you think about computers being integrated with our bodies – now, not everyone has a pacemaker – but how many of your thoughts are mediated through a computer? Your ability to express and your ability to think. Right? If you, these days, if I want to think about something I tend to use this network of computers to augment my thoughts. But if certain things to think about are not allowed, that's an impediment on my freedom to think and my freedom to express. So a lot of this stuff sounds like nerd-rights type things. Especially when you're talking about 'let's build a cyborg, let's build a cyberpunk future where everybody is playing text adventure games'. It's easy to believe, I think, when you start talking about this stuff too much to give people the impression that you've kind of lost connection with reality. But, to me, what used to be nerd rights are basically human rights today. If computers are mediating every part of our daily lives, then they're not just computing rights. They're human rights issues.”
- Chris Webber (interview conducted by the author, July 2018)
Introduction
The principle of decentralisation is a large part of why the pioneers of the internet envisaged it to be a new beginning for freedom of exchange. In its early days, the internet seemed to promise the deconstruction of the centralised powers that governed the world that came before it. “Governments of the Industrial World, you weary giants of flesh and steel, I come from Cyberspace,” wrote John Perry Barlow, one of the founding members of the digital rights group Electronic Frontier Foundation, in 1996 (Electronic Freedom Foundation, n.d.) “On behalf of the future, I ask you of the past to leave us alone. You are not welcome among us. You have no sovereignty where we gather”. The internet implied a network of individuals who could connect without telecommunications companies or postal hubs or immigrations teams deciding who you can and can’t communicate with. No advertising agencies, or journalists, or governments, or legal institutions deciding whether or not you are allowed a platform on which to share your information. The decentralised nature of the internet meant that if you have an internet connection, you could become a node on the network just as powerful as any other, giving the opportunity of a voice to every user.
In 2018, things look different. It has been declared that this dream of a decentralised internet is dead, as huge amounts of web traffic is governed by giants like Google, Facebook,
traffic through either Facebook and Google and 65% of data stored on the public cloud was stored by either Amazon, Google, or Microsoft (Moses, 2018; Synergy Research Group, 2018). In response to this professed crisis, various actors and groups, including the architect of the World Wide Web, Tim Berners-Lee, have claimed that there is an urgent need to ‘re-decentralise the web’ (Berners-Lee, 2014). When Amazon Web Services (AWS) had an outage in early 2017 it caused swathes of the internet to shut down for 5 hours, and this led to concerns about the technological implications of these single points of failure. But social, political, economic, and cultural concerns also arise from a centralised web for the fact that so much information lies in the hands of just a few organisations.
The landscape of the forces working towards internet and web decentralization is sprawling and heterogeneous. The efforts, philosophies and people working towards this goal are diverse. Some projects, for example, are attempting to replace elements of the web that have been in place since its conception, such as HTTP or DNS, with the aim of completely changing the way data is stored and sent across the web and eventually creating a fully ‘distributed’ or peer-to-peer web (such as the Inter-Planterary File System and Blockstack projects). This would mean every node in the network i.e. a computer, could communicate directly with another, without routing information through a centralised server. Such approaches have been garnering headlines of late due to the excitement about the potential use of a cryptographic solution called a blockchain, the technology Bitcoin is based on, to achieve this. Others, however, are contributing to decentralization in other ways, particularly through something called federation, which has been around since the very early days of the internet.
The following example is a quick introduction to federation. Imagine trying to organise a video call with a friend. “How shall we do this?” asks your friend, “Skype?”. “Oh, I don’t have it installed. Can we do it on Facebook Messenger? “No, sorry”, says your friend, “Didn’t you delete the app? I have. What about Google Hangouts?”. “How do you even do that? Shall we just use WhatsApp?” This irritating situation is an illustration of why we tend to end up with particular platforms dominating the market for specific social networking use cases. It’s easier if all your friends agree on one system e.g. WhatsApp for messaging, Skype for video calls, Twitter for microblogging. Because of the network effect, it can feel
inevitable that we should end up with particular platforms monopolizing certain use cases.
Yet it isn’t the only way doing things. For example, if you were trying to send long letter-style messages over the web rather than a video call, you could use email, and you wouldn’t have to worry about what software your friend was using as long as you had each other’s unique identifier: an email address. You and your friend could send one another emails regardless of what client you are both using (e.g. Outlook) or what server your data is stored on (e.g. gmail’s servers). This is because email is federated, meaning both of your servers can communicate with one another using a shared protocol. If your friend wanted to change her client or server, in principle she could do so and it wouldn’t affect your ability to
communicate with her. For most other web communications uses – sending instant messages, updating your friends with a social media post, reacting to someone else’s post, commenting on a video, conducting voice-over-internet calls – the most popular pieces of software do not use protocols to federate, and users end up stuck in a closed system. When people claim that closed source software holds their users in “walled gardens”, this situation is what they are referring to.
However, there is no technical reason why other types of social networking can’t act like email. In fact, many of the web communications platforms that are popular today originally used protocols that allowed them to federate with other platforms but eventually removed support (usually after gaining a critical mass of users) (SOURCE ABOUT GCHAT AND SLACK). It is claimed that this dismantlement of federation is part of the reason for the increasing consolidation and centralization of the web today. In turn, many of the projects who claim to be working towards the goal of decentralizing the internet look to federation as a solution to the perceived over-centralisation of the web. In recent years, the decentralised or federated social web has seen a growth in interest. The fediverse, a portmanteau of ‘federated universe’ is a collection of free, open-source social media platforms which provide a non-commercial alternative to mainstream closed-source social networks, while also federating with one another.
The most popular platform in the fediverse is called Mastodon, a microblogging service similar to Twitter. To explain what makes it different to Twitter and to give the reader a better understanding of how federation could work outside of email, I should provide a brief explanation of the basic structure of the World Wide Web: the client-server model. In this model, information is sent between clients and servers which are connected to each other via the internet. The ‘server’ is software that stores and delivers information, and the ‘client’ is software that requests that information. If I want to access Twitter’s website, my web
browser (the client) sends a request to Twitter (the server) using a specific protocol to ask for the information they have stored, and that information is then delivered to me. Because Twitter is centralised, any information that passes through its system is stored on its own servers. When I post on Twitter, my client sends information to Twitter’s servers, where it is stored. When you read my post, your client also connects to Twitter’s servers, and requests that information. Mastodon, on the other hand, does not have any centralised servers. Users are required to either provide their own servers (called ‘hosting’) or choose another self-hoster’s servers to store their information on. Thus when I post to Mastodon, this information is stored on my own server, and if you want to read my post, your client must send a request to your server, which then sends a request to my server. Both our servers would communicate with one another using a shared protocol. This relationship is federation.
Due to this federated model, the fediverse is an ever-growing and evolving ecosystem. It is composed of thousands of different servers of various sizes. Due to the open source nature of the software and the federated model, people are not limited to using Mastodon to access the
information on the fediverse. If you had an account on another platform, for example Plemora (another microblogging platform), your server on Plemora could federate with my Mastodon server, and we would be able to communicate with each other despite having accounts on different platforms. This creates an inherently unstable network of servers which are constantly in flux as servers are born, die, expand and contract.
The decentralised social web doesn’t just exist, of course. It is the cumulative result of the actions of many people spread across the world with a range of motivations and desires. It is created with purpose, and, as such, is as much a political undertaking as it is a technical endeavour. It is simultaneously a social movement and a technological construction. In this thesis, I will explore the people, places, ideas and materials that form the decentralised social web.
I will attempt to show that its ontology is contested and multiple. Temporally, it is an infrastructure that simultaneously exists in the past, the present and the future. Tangibly, it exists as a material, a discourse, and as practice.
Theoretical Framework The materiality of the web
The idea that something material, like technology, can influence cultural practices has implications for ongoing debates relating to materiality. To say that technology, specifically internet technology, ‘has’ the ‘power’ to affect change in the world leads one to ask what it means for a seemingly material item to be infused with a seemingly immaterial trait like power. Does it have this power inherently? Is it ascribed that power by humans? If so, are those powers ascribed collectively in ‘cultures’ or something else, like individuals, or by the particular surroundings that an object finds itself in? And what is the internet: a concept, a thing, a physical infrastructure? While this scholarly arena remains sprawling and
controversial, the debates around which I will not have space to cover here, it is safe to say that interest in such questions has increased in recent decades. Whether approaching the material as agential entities in networks of human and non-human actors (Actor Network Theory [ANT] – Callon, 1991; Law, 1992); considering objects as ‘multiple’ (‘material semiotics’ – Mol, 2002); or even endorsing the phenomenological consideration of objects outside of their relationship to human consciousness (‘object oriented ontology’ – Harman, 2009); a number of scholars are moving away from the idea that there exists one physical nature upon which multiple human meanings are attached.
Henare, Holbraad, and Wastell (2007), for example, in Thinking Through Things argue that when anthropologists take the ostensibly pluralistic stance of cultural relativism, and assert that there is no one truth because the material world is simply interpreted differently by different cultures, they are contradicting their aim to take difference seriously while
paradoxically leaving space for so-called positivist science to ascertain the ‘real’ truth about the material world. They use the example of a powder called aché, something which the Cuban diviners who use it claim ‘is’ the physical manifestation of power. The authors claim that “the anthropological problem cannot be that of accounting for why [a diviner] might think that about power (explaining, interpreting, placing his statement into context)” but rather understanding that this powerful power exists in “a different world, in which what we take to be power is actually power, or, more to the point, a third element which will remain
ineffably paradoxical for as long as we insist on glossing it with our own default concepts” (Ibid: 14).
What I find particularly useful about this frame of thinking in the case of the decentralised social web is that by understanding reality as made up of multiple worlds we can dissolve the distinction between ‘concepts’ and ‘things’, categories which become murky when trying to conceptualise something like internet based technology. As Henare et al. state, following their line of reasoning, “alterity can quite properly be thought of as a property of things – things, that is, which are concepts as much as they appear to us as “material” or “physical” entities” (2007: 15). ‘The internet’ is multiple. It is a virtual network. It is a collection of cables running under the sea. It is an archive of information. It is a world that I can enter through the screen of my smartphone. It is an oppressive tool of the ruling class (Graeber, 2015). While such multiplicity could also be approached in interpretivist terms, the internet being a ‘floating signifier’ (Levi Strauss, 1950) that means many things in many contexts, I agree with the authors that thinking “through things because our informants do” (Henare et al., 2007: 25) allows us to take other people’s worlds seriously, and might better help to understand how something like the internet can be ‘centralised’ or ‘decentralised’, or how it relates to power. This is, of course, a philosophical standpoint with methodological
consequences which will be addressed later on.
Agency, control and resistance
There is an inherent claim in the idea of ‘decentralisation’ to the agency of things. If a centralised web can cause social and political harm, and the creation of alternative forms has the capacity to divert those social and political effects, then somewhere we must be able to find (at the very least an assertion of) agency in non-human actors. If we consider such a notion, we can see that it is similar to claims that art, a human artefact, can change the world. How can we make sense of such claims? On the one hand, it could be interpreted in symbolic terms: technology or art act as material symbols for social relations. For example, a religious icon as a symbol for belief in god may be discussed by its users in the language of power. On the other hand, one could look at the material in terms of the actual effect it has on a social situation. When that idol is in the room, how does that shape the social behaviour of the people who encounter it? They may dress differently, use different language with other people in the room, alter their body language. It is for this reason that Latour (2005) argues that it does not make sense to understand the world only in terms of human actors or that there is actually a social world distinct from the material world at all.
However, as a human-designed artefact, this leads to a debate about how much intentionality matters. Gell (1998), for example, argues that art can act as a social agent by being a vessel for the social intention of the artist. The effect on the viewer is thus a transferral of the artist’s agency. However, such an approach may not go far enough to make sense of human artefacts that were designed for perhaps a specific non-‘social’ cause, or had an entirely different intention, but still have social residue. Moreover, it has been argued that it draws too large of a distinction between artefacts and non-artefacts – does it make sense to distinguish
ontologically between objects that are designed by a human vs. by ‘nature’? Harman (2009), going even further in his critique of Latour and ANT, argues that defining networks as necessarily including only those objects that interact with humans is problematic. Can networks of actors exist between only ‘non-living’ material objects? What if there is a dog involved? What about a factory that has been designed so that all machines interact with one
another so perfectly that no human has to enter the site. Is this a social setting? Is the agency of the human designers of these machines present through their artefacts?
Media theorist Galloway (2004) explores the human decision-making that went into the design of the physical infrastructure of the Web, and the role that intention plays. He argues that the dominant software protocols that defined how information is transmitted over the web as well as governance protocols were intentionally constructed to give the techno-elite lasting control over the infrastructure of the Web. This is framed as a response to claims, such as the one by web inventor Berners-Lee that the initial intention of the Web was to promote freedom. By creating open source (non-proprietary) software protocols that people want to use and thus become dominant, Galloway argues, a decentralised-by-design system is still controlled by the actors who designed the protocol that distributes that power,
information, electricity, etc.
If there is control (and thus power relations) embedded in the materiality of the internet, whether by the (intentional or non-intentional) design of its original creators, by the
(intentional or non-intentional) behaviour of market forces, or by someone else such as state actors, this leads to one of the central puzzles of this research. If power is built into the structure of the internet itself, then might changing the way those protocols work be a type of resistance? And moreover, would the material creations of the technologists contain the essence of the designers’ agency, or be agential in their own right? Coleman (2012), in her ethnography of hacker culture, describes information security hackers’ fight against the criminalisation of hacking. They claim this is a form of censorship as “code is speech”. Could a methodology that dissolves categories like concept, thing, object, belief, etc. help us
understand how code can be speech, and thus how resistance relates to both the immaterial and the material in the practices of hackers and technology activists?
Freedom technologists: the internet and activism
The politics of the internet is a vast terrain around which activism takes many forms. It becomes apparent during altercations between traditional state actors and citizens; the role of social media by protestors for example, in the Arab Spring, is well documented, as well as the political manipulation of social media for electioneering that was exposed during the
Cambridge Analytica scandal (Guardian, 2018). However, analyses of such phenomena approach the internet as a tool used by traditional political actors like the state and civil society to further their aims. What is distinct about freedom technologists is that they are engaging in a politics of the internet itself. In the case of the constructors of the Decentralised Web, their political action takes place through the medium of technological practice.
The literature on hacker cultures is particularly pertinent in understanding the actions of internet activists. As Coleman and Golub (2008) identified, there are different “moral genres” of hackers. As well as the subversive information security hackers, many also fall into the camps of Free and Open Source (FOSS) or Cypherpunks, with different philosophies as to the use of information technology. While FOSS developers insist that software should be openly and freely available, Cypherphunks focus on developing software for individual privacy and are often referred as being influenced by libertarian values of individual freedoms (Ibid). These philosophies are influential in the design of the decentralised web with its focus on both FOSS software as well as an intrinsic design for privacy.
Media studies researcher Milan (2015) discusses “hacktivism” as a political practice, exploring its many elaborations and finding that it is an evolving form of political practice that takes place not only through exploiting technology in subversive ways, but can also involve changing that technology, or creating new technologies. For example, she quotes the IndyMedia (a publishing network) slogan of “don’t hate the media, become the media” (Milan, 2013: 127) as an example of creating technology for political change, rather than exploiting it. An example of how software development may be a type of “hacktivism” is discussed by communications researcher Gehl (2015) who did interviews with developers of Twitter alternatives, which tend to be decentralised. He found they were divided into
‘Platform Builders’ and ‘Politics Encoders’. The former see their work as purely an
expression of open source principles rather than political expression: “the only politics I see embedded in the work is that of software freedom” (Matt Lee, GNU Social founder in Gehl, 2015: para 60). The latter see their software development as a type of activism, all their actions and technological decision-making being a type of political practice: “As long as writing sentences is activism, so should the writing of code, the filing of bugs, the deployment of software, the dismantling and rewriting of software, software piracy, etc” (Wilkinson, rstat.us developer, in Gehl, 2015: para 73).
Also of note is the ways in which these actors negotiate the dystopian and techno-utopian narratives that frequently emerge in debates about technology and society are inherent in the idea of ‘decentralisation’. The very idea of actively creating technological solutions in light of concerns over power abuses of a ‘centralised’ future implies a
negotiation of these two opposing narratives. Barassi (2016), in her ethnographic study of internet technology use in political movements, found that utopian and dystopian projections of technology co-existed in the activists she studied, one in careful balance of the other. For example, social media is simultaneously conceived of as both a tool for authoritative
surveillance and a platform for grassroots democratic organisation. Another element of this study is to understand how these contradictory narratives are negotiated by the actors in question: how do fears around surveillance and lack of privacy exist simultaneously with hopes of internet freedoms? How do these actors negotiate these narratives when the source of fear is the very same source of potential liberation?
From theory to method
For me, the exploration of this phenomena was not just a study of a particular aspect of human experience, but in a way, also a methodological exploration. From the early stages of planning this research, I grappled with the puzzle of aligning the research design with some of the traditional ways anthropologists plan their research.
As many writers have noted, there are certain challenges that arise when dealing with the increased presence of the digital world in the daily life of people (Pink, Horst, Postil et al., 2016). The decentralised social web is composed of a range of people, all physically existing in different places, all interacting through and with the internet in a variety of ways. The very philosophy of decentralisation seemed to be the irrelevance of geographical separation – that dispersed networks can be formed of individual connections supposedly unhindered by cultural or geo-political spatial categories – therefore I did not believe it was possible to identify specific sites beforehand. Inspired by Gupta’s (1998) attempt to marry the
simultaneous translocality and localised nature of ‘the state’, I wanted to explore how I could achieve something similar when dealing with this phenomenon.
Gupta argues that ‘the state’ is constructed through both everyday practices of local
bureaucracies as well as through public discourse (1998: 212). However, what complicated things further in the case of the decentralised social web, was a sense that this phenomena does not just exist discursively or through practice, but also materially. Through my computer, I could conceivably see a user interface of every webpage and every piece of content that they currently contain. As the software is open source, I could look through its blueprints. The servers which compose the network and store the data, physically exist in the world and are visible and tangible. The users that make it up exist as – or, depending on one’s perspective, are represented by – avatars, accounts, email addresses, blog posts, code, etc. It seemed to be something that is constructed discursively, through practices, as well as materially.
In terms of my methodology, I needed a way to reconcile the treatment of human actors and their social relations – the narratives they tell, the stories they share, the other people they interact with – with a fair treatment of ‘things’ or the material (Henare et al., 2007). I felt the best way to do this was by framing my research around the idea of ‘infrastructures’.
Infrastructures have become a topic of increasing ethnographic concern as systems that connect people and things in a variety of ways (Larkin, 2013). Infrastructures of road
networks, water pipelines, electricity pylons, can be treated as the backbone of heterogeneous networks and thus allow ethnographers to study the political consequences of systems that connect spatially disparate phenomena. Infrastructures can be seen not just as material constructions – the tarmac on the ground or the pipes underneath it – but can include many other ‘things’ such as the various stakeholders involved in designing them or receiving their benefits, the educational infrastructures that go into managing the knowledge acquisition which allows the production of such a system or the movement of money that revolves around their implementation.
A benefit, says Larkin, of framing research around infrastructure is that it allows one to define what is in or out of the network. Edwards et al. (2009) debate how we can define infrastructure, particularly in the case of “e-infrastructures” or “cyberinfrastructures”. On a temporal level, must a system be “established” for a certain amount of time to qualify as an infrastructure? Spatially, must it be of a certain size in terms of users or space? The authors argue that it does not make sense to think in parameters like this, as infrastructures never go from a state of unbuilt to built, or unestablished to established, but are in a persistent state of negotiation. They suggest that: “perhaps e-infrastructure is emerging first on smaller scales of time, space, and service, as second-order systems built on top of and around the Internet and other existing information frameworks” (2009: 366).
I hoped that an exploration of the establishment of new “e-infrastructures” – in particular, the infrastructure of the decentralised social web – as the grounding anchor of my study would allow me to make sense of the heterogeneous actors involved in the infrastructures’ (ongoing) construction. Which humans are involved and how do they connect with each other through the infrastructure? How does the infrastructure itself engage with and position itself within wider infrastructures (the World Wide Web, for example)? How do the physical elements of the infrastructure connect to one another i.e. where are servers physical located, what is the role of human actors in connecting these? How do both human and machine negotiate who and what are accepted and removed from the infrastructure? Furthermore, I hoped to understand the relationship between the construction of these new or challenger
infrastructures and the political ideology of the actors involved. I wanted to explore not just the political ideology of the freedom technologists themselves, but how other forms of
ideology might be embedded in or emitted from the infrastructure itself. I wondered whether the creation of a new infrastructure which is granted a particular ideology by its creators, but embedded within a wider infrastructure which enacts a different ideology, might become a political battleground.
Jensen and Morita in their 2015 contribution to the debate on infrastructures argue that infrastructures are interesting to social scientists because they act as “ontological experiments”. The article is in part a contestation of Noortje Marre’s (2013) distinction between empirical ontology and experimental ontology in regards to infrastructures. Marres had argued that infrastructures become “experiments” when material artefacts are
purposefully imbued with politics, when politics is practised through material artefacts. Jensen and Morita, on the other hand, argue that this human intention should more properly be considered just a small part of the ontological experiments that infrastructures can be. They argue that infrastructures can still be a site of political action regardless of whether there is a human actor knowingly acting upon the material world with political intentions. Indeed, they argue, infrastructures are always ontological experiments for “they are sites where multiple agents meet, engage, and produce new worlds” (2015: 85). In this thesis, I will explore the role that political intention plays in the construction of the decentralised social web.
Method in practice
As an anthropologist who is planning research design, one is expected to define one’s field site, population and topic. But these always seemed to be nebulous concepts in the context of something as sprawling and difficult to define as the phenomenon of the decentralised social web. It would in fact have been inadequate to go ahead and pick a particular population, such as the users, the developers, or the advocates of this phenomenon. It would likewise be inadequate to decide on a particular geographically defined site. While the focus of the so-called ‘ontological turn’ in anthropology is around making a conscious effort to bridge the (according to Latour) artificially-created divide between nature and society and thus between constructivism and realism, in this case, I would imagine, it might actually feel more intuitive to most people to reject such a divide. Even while wearing the hat of an interpretivist social scientist, it would seem unintuitive to draw a line between what is ‘real’ and what is
‘constructed’ in the case of the web, the internet, or an online avatar, for example. Indeed, the phenomenon I was interested in seemed to be made up of a range of ‘things’ that do not fall neatly into the categories of nature or society (Latour, 1991): humans; material artefacts; ‘virtual’ artefacts; concepts; ideas; values. Therefore, my approach had to take a less
constrained form than pre-defining the population and selecting a particular geographic field site, or even multiple sites. My approach, instead, was to understand the decentralised web as an infrastructure, one composed of material and immaterial elements.
However, attempting something methodologically experimental was not without its difficulties. While it might have sounded tempting to “ethnographically explore a
heterogeneous infrastructure”, in practice, this was a persistent challenge in my research. From the beginning of the planning stage and throughout the fieldwork process I grappled with this question: what exactly am I researching? There were moments where I felt I had a strong grasp on my topic, and other moments, it seemed to be slipping through my fingers. In my mid-fieldwork meetings with my supervisor, the same questions kept emerging: What exactly am I studying here? What is it that ties all of these people and places and digital
experiences together? Which leads should I follow, how can I know what is relevant? What should I do now? I had this sense that I was researching a black hole.
Moreover, the creation of the decentralised social web is a social movement defined in technical terms, and attempts to understand it had to involve getting to grips with technical concepts that I was unfamiliar with. The whole world that I was entering was new and alien, and there was a lot of new language to learn. My original idea was to look at decentralisation movements more broadly, but it was only in the course of my fieldwork that I was able to realise quite how far the area extends. This allowed me to eventually narrow the topic down to the decentralised social web. Before I entered this world, I didn’t know the divisions that existed and didn’t have a conceptual map of how the decentralised social web differed from other decentralisation projects. I didn’t know how the different types of decentralisation were categorised by those involved (which seem to exist as social categorisations beyond simple technological distinctions). Thus comprehending this technical and social landscape was something that took place over the entire research period. The diversity of communities and the divisions within them was only something that I began to understand during the fieldwork process, and, of course, would never be able to achieve without much deeper and longer ethnographic immersion. What I came to realise was that engaging with a variety of
seemingly heterogeneous things, people, and places eventually allowed me to ‘approach’ the seemingly intangible phenomenon I wanted to study.
Consequently, my ethnographic material is diverse and scattered. It is comprised of
conversations, both formal and informal, both online and offline with people in Amsterdam and abroad. It includes my experience of being a part of online chatrooms, attending a hacker camp in Amsterdam, going on a bike ride with a new friend, browsing alternative social media platforms, attending a fortnightly hacker meetup, going to a talk at a design collective, and hosting a group discussion with 27 hackers. It comprises reading people’s personal websites, blogposts, thought pieces, and interviews conducted by other people and circulated on the fediverse. My material blurs the line between primary and secondary research, and it is heterogeneous, just like the subject matter.
Research question
The questions that guided my research and which I will explore in the following chapters are: 1. How is infrastructure negotiated in the making of the decentralised social web?
a) What role does human intention play in this negotiation? b) What role do ‘things’ play in this negotiation?
In the chapters that follow, I will argue that the decentralised social web is both an
assemblage (Deleuze and Guattari, 1987; De Landa, 2006) and an infrastructure that acts as an “ontological experiment” (Jensen and Morita, 2015), a site of dynamic interaction between different ‘things’ (i.e. concepts, people, materials, etc.). It has no clear boundaries, only existing as a reaction to outside forces. It is simultaneously constructed of concepts,
ideologies, people and material artefacts, all engaging in a complex and ever-evolving state of negotiation and re-negotiation. In my exploration of the decentralised social web, I observed an infrastructure with different layers of ‘political’. People and their ideologies are present and act as transformative agents in the infrastructure, but other elements like network topologies, protocols, and alternate infrastructures are also seen to play an essential role in producing the infrastructure of the decentralised social web.
In chapter one, I address the materiality of the infrastructure and contend that it is impossible to see the infrastructure as a material bounded entity. In fact, its borders can only be drawn based on what it is not, infrastructure it wishes to transmute (centralised, commercial social media). In chapters two and three, I look at how this apophatic ontology is performed, both intentionally and unintentionally, exploring the complex interplay between these different ‘things’. In chapter two, I look at the role that people and their political ideologies play in assembling the decentralised social web. In chapter three, I look at the role of other ‘things’ and their role.
Chapter 1: An apophatic infrastructure – on what the
decentralised social web isn’t
Apophasis is a rhetorical figure coming from apophatic theology. Stemming from the Ancient Greek for “speak other than”, this is a branch of religious thought that believes one can or should only speak of the Divine in negative terms, in terms of what it is not i.e. God is ‘infinite’ and ‘incomprehensible’ (William, 2014) It is the opposite of cataphasis – the use of affirmative statements to discuss a concept. Extending beyond theology, the dichotomy of apophatic and cataphatic is used by Alexander Galloway (2016) to categorise different theoretical attempts to comprehend the ‘nonhuman’ in cultural studies. When the nature of something is contested and in a state of constant change, as in the case of the nonhuman, but also I argue the decentralised social web, attempts to define it may veer down the path of definition by negation. One of my first observations about the decentralised web, when trying to approach what it could be, observed that it seemed to be an entity of negation. Among other things, it isn’t centralised, it isn’t Facebook, and it isn’t capitalist. In this chapter, I will explore how the infrastructure can be seen to exist as an apophatic ontology.
It’s not centralised
Something that comes up time and time again during my fieldwork is the term centralised or centralisation. After all, decentralised is a negation of this idea.
The issue of centralisation and decentralisation was the portal through which I originally entered the field. I was already peripherally aware of a phenomena happening whereby people were interested in ‘decentralising the internet’. I had heard about Bitcoin and blockchain, buzzwords used by people who claimed that there was a revolution on the horizon, where humanity would be ‘decentralising’ not just the internet but by extension wider economic, geopolitical and infrastructural configurations. When I began this project it was around the time of a boom in both the price of and interest in Bitcoin and the various alternative cryptocurrencies. The media I was consuming, in the news and on the social media platforms I was frequenting at the time –Twitter, Reddit, Facebook – focused on fantastical claims about their capabilities to disrupt the status-quo. The problem for me, however, was how deeply wrapped up in commercial narratives these claims seemed to be. Unfortunately, there was simultaneously what seemed to have been a speculative bubble surrounding these claims, so it felt very difficult to disentangle the lines between what was genuine excitement around the technology and what was “shilling”, advertising, or market manipulation.
When I began delving further into the term in my preliminary research of this project, I discovered a whole new realm of meaning to the term. Decentralisation as an idea in software existed far beyond the Blockchain scene. I read about how the term had a strong history in the hacker scene, how Steven Levy (2010) had noted in the 1980s that the drive to
“decentralise everything” was already one of the “codes” of hackerdom. I read about the exploding assortment of decentralised web services, and about the numerous projects that had been built in a decentralised way over the years from chat platforms to various web services, usually as tools that were popular with developers themselves. Projects focused on privacy, such as the Tor project – software that allows anonymous communication and famously used by government activists like Edward Snowden as well as users of the ‘Dark Web’ to evade
the detection of state actors – were decentralised. The decentralisation acted as a core feature of their efficacy.
What was different for me as a researcher about the idea of centralisation was that when I began my fieldwork I was already familiar with the concept. I had read about the idea not just in academia, but on the web: on forums, blogs, social media. I had read blogs about how important it is to decentralise the internet. I had read an interview with the founder of the World Wide Web, Tim Berners-Lee (2018) who claimed that “the web has evolved into an engine of inequity and division; swayed by powerful forces who use it for their own agendas” and how “powerful change for the better is possible — and necessary”. I saw people talking about how the “overly centralised” giants of the web – Google, Facebook, Apple et al. – were locking users into “walled gardens” of closed source software. I learnt how the software we use online should be “open, decentralised, federated and respectful towards freedom and privacy” (Disroot, n.d.) rather than “centralised” networks run by players in an overly
“centralised” market of web services. I learnt how centralisation is something to fear because one company having too many users allows them to create ‘lock-in’, where users become trapped into their services due to network effects. How the closed source centralised databases these centralised companies and platforms create are insecure and vulnerable to attacks as well as to being sold off to the highest bigger for advertising or surveillance purposes. I read how the increasing centralisation of the internet and the web are a danger to the freedom, democracy, and safety of people across the world.
What I didn’t realise while I was doing all of this reading was that I was not in fact just doing preliminary research. This was actually the beginning of my data collection. As I would come to learn, the aforementioned narratives that form the decentralised web are as much a part of its infrastructure as the code it is written in. As I engaged with my respondents over time, I heard the same ideas as the ones I had read over and over. In fact I found it to be a
universally accepted principle in the hacker scene and of course on the decentralised web. Centralisation is, to those involved, self-evidently not just dangerous but unethical. It is a universal, unchallenged enemy. However, what I came to learn from my informants is that the converse – decentralisation – does not have quite the same fixedness. It is a highly contested term. While centralisation seemed to be tangible and conceivable and visible in the world around us – decentralisation is not.
One of the formal interviews I conducted as part of this research was with Chris Webber, a name suggested to me by someone I had met in Amsterdam (a story I will tell later). He is one of the co-writers of activitypub, the protocol which Mastodon and other fediverse platforms use to federate, and (I was told) a well-known figure in this space. As he lives in the US, I got in touch with him online using a messaging protocol called XMPP and we arranged an online call using a platform called Mumble (a Voice over Internet Protocol tool). Below is an excerpt from my call with him.
Cwebber: The system we have at the moment with it being federated is not necessarily distributed. We have... it's decentralised, but not distributed. Have you seen that diagram that's got the centralised, decentralised and distributed diagram? It gets passed around at lunch.
Me: I think I might have... it's like equal lines between all the points as opposed to, like, clusters that are decentralised, isn't it?
CW: Right. So it comes from a drawing that was done by someone who was researching network topology in the 60s or 70s. Let me pull it up.
[…]
CW: Here's the link.
[20:06] (Channel) cwebber: https://www.researchgate.net/figure/Centralized- decentralized-and-distributed-network-models-by-Paul-Baran-1964-part-of-a_fig1_260480880
CW: We are basically at B, but we want to be at C. […] Well it turns out that activitypub was built in such a way that we left it open to being in a fully distributed system. But that's not how it's being currently built. But the protocol itself supports it. And I wrote another paper about this. [silence] […]
[20:11] (Channel) cwebber: https://github.com/cwebber/rebooting-the-web-of- trust-spring2018/blob/petnames/draft-documents/making-dids-invisible-with-petnames.md
[…]
CW: So it is actually possible to take activitypub as the spec is written and move it to a completely distributed-type system.
[…]
CW: However, one of the things that is happening right now is that a number of people who are implementing activitypub are like 'oh, well some of these things are too complicated, we'd like to move to a subset of it that we're calling whitepub that leaves some of these things that are currently left open and close them off so that we can more easily implement the systems that we have today’. So that's kind of a tension that's happening right now is: do we stay at B and make it as easy as possible, as activitypub can be implemented as B. It is currently implemented for B. Or do we say: ‘no it's really critical that we open up the possibility for C in that diagram’? For the completely distributed system. And I think that B is not sufficient enough. The federated model in the decentralised but not distributed model is not good enough. Especially for all the reasons on for example servers going down, or,
you know, censorship problems. I think we need to move to this other type of model, but I don’t have the buy in of everybody who is doing implementations today.
Chris, who co-authored activitypub, the W3C recommended protocol that has ballooned in popularity due to its use by developers of federated social network applications, explains that he is not satisfied with the kind of ‘decentralisation’ that the fediverse platforms use today. Using a popular image of different network topologies, he explains that the federated network model is not decentralised enough. To him, it is a sort of half-way house towards a desired future imaginary decentralisation whereby every user would be running and maintaining their own node, not relying on anyone else to represent them or to host their data. From his point of view, these federated systems are just following the status quo in web development – the technology is not quite ready to create these systems in a fully decentralised way. He tells me that there is a tension amongst those people using activitypub today because of those who want it to keep it as a federation tool, rather than keeping it open to create a fully distributed web. Something I found interesting was that Chris told me that most people who want a decentralised web want a fully distributed web and feel similarly to him in that federation is a sort of halfway house towards that goal. However, from other people, I got the impression that federation can mean more than that. For many, federation stands for something unique and provides a different sort of solution to centralisation than full distribution.
From some of these individuals, I was told federation represents a sort of ideal type for an “anarchist society”. Servers act as hubs that give people the freedom to associate as they wish and to change alliances as they please by disconnecting themselves from one server and reconnecting to another that better suits them. Those hosting the server have to create
reputations for themselves to create trust. All of this, in turn, promotes solidarity among those who decide to band together through their technological choices. It gives people a different type of freedom, a much more collectivist understanding of freedom, which perhaps differs from the more individualistic freedom discussed by Chris in the excerpt.
Others expressed things on similar lines telling me that part of what is exciting about the fediverse’s model is how it requires users to recognise what using the web involves. A moderator of a small Mastodon server, Herman, explained to me that part of the appeal of Mastodon is that it fosters a familiarity for users with the underlying technology of the web that is important to help spread awareness of the dangers of centralisation. Because servers cost money to run, moderators on Mastodon have to ask their users for donations. Because servers are run by moderators, users have to engage with them as a person to ask them if they want to change the rules, such as banning certain types of content or conduct on the server. Because users have to actively select a specific server, this helps remind users that data is always stored by someone, and by using web services they are making a choice to hand their data over to somebody, while that process is normally obscured.
This stands in contrast to ideas that Chris, among others, have for what decentralisation might look like. Many projects focused on decentralising the internet look at ways to do it so that this new configuration is obscured from users. One of Chris’ ideas that he explained to me for example, is to employ something called ‘pet names’ to resolve a technical problem with achieving decentralisation, a trilemma called Ziko’s triangle. This axiom states that network addresses can only ever be two of the following three: decentralised, unique, and human-readable. Domain Name System (DNS), the registry system on which current web addresses are based, resolves this triangle by essentially re-centralising the web, which underneath it all is a decentralised network of IP addresses. The web addresses users are used to seeing,
URLs, have to be converted into the IP addresses to which they actually refer using a centralised DNS server. Decentralised networks, such as Tor.onion addresses, have had to sacrifice human read-ableness, and thus practicability, in order to be decentralised. Chris explains that a solution to the problem of web decentralisation could lie in resolving this issue of human readability by essentially obscuring a non-human readable address with a ‘pet name’. The inner workings of the web would be obscured in this case, and it could cause a decentralisation effort underneath the noses of web users.
Part of this contradiction is that we are talking about different things. Herman comes to the fediverse from a very different place than Chris. Herman is a Dutch artist, activist, and web developer. Chris, an American, comes from a much more academic place, he writes papers on decentralisation and is involved in working groups from the W3C, the global web standards agency. Their visions as to what decentralisation is and what it is for look very different. Herman’s interest comes from a place of intrigue, a desire to experiment, and to be part of something that subverts the status quo in the now. As a web developer, the web as it stands today is a world he is familiar with and a world that still contains that promise of decentralisation. “The web was the first big decentralised network”, he told me. Chris, on the other hand, is more focused on the potential for technologies that may not even exist yet. His desire to promote what he calls “user freedom” comes from a sense of urgency.
It is comments like this that makes me think that federation, for Herman as well as others, is admired in part due to nostalgia. It is something that has been employed by web developers since its early days. It is not just familiar but also contains those original hopes of what the web could have looked like. The idea of ‘re-decentralising’ the web has these tensions written into the name. It looks to the past, to a web that used to be, and also implies a need to find new solutions to achieve that goal. I think this could be one part of why there are mixed sentiments towards blockchain in some of the communities around the decentralised social web.
So while there may be an agreement in the decentralised social web community about why centralisation is ‘bad’, and why the ‘decentralised’ web is an alternative to it, there is no consensus as to what decentralisation actually is. Federation may be one type, but there are many other types as well, many of which don’t exist yet except as speculation. So while the decentralised social web may be called decentralised – all one can really say is that it is not centralised.
It’s not Facebook
I began my research in Amsterdam by going along to a ‘hacker meetup’, suggested by a peer on my course. It was at these meetups that I gained by entry to the field, after meeting several people involved in the hacker scene in Amsterdam. While different people would come each week, there were four consistent attendees who I would see every fortnight (Theodoor, Adam, Dom, Sarah). One individual in particular, Adam, was particularly engaging an brought me under his wing, eventually becoming my ‘key informant’. Over the summer, I spent a lot of time with Adam in particular, who provided a lot of insight into the history and relations of both the hacker scene in England, Amsterdam and the decentralised social web online. Early on, he invited me to come along to a few events including an open day at a ‘hackerspace’ in Amsterdam, and later a four-day ‘hacker camp’ that was being organised at ADM, a large squat in an abandoned ship building yard on the outskirts of Amsterdam. This camp became one of the key sites for my research. Spending four days on site with and
meeting a wide range of characters, I felt not just completely out of my depth, but like I had gained entry to another world. I met people who had travelled from around Europe to attend, from Berlin, Hamburg, Paris, Glasgow and across the Netherlands. This camp I learned was something people had been looking forward to all year and also represented its own form of rebellion against a more ‘corporate’ feel that larger hacker camps had apparently been developing. In line with the anarchist philosophy many of the people I met subscribed to, the camp was supposed to be organised in a ‘do it yourself’ fashion, with people collectively organising who would clean, cook, serve food, play music as well as host activities and talks at different times. I was encouraged by Adam and Theodoor to submit my own activity to run at the camp, so I decided to run a group discussion about my research – which ended up lasting about an hour long with 27 attendees of different types. Below is an excerpt from part of this discussion:
Figure 2: My submission for a talk at the camp
Adam finishes his talk about the history of the fediverse, and makes an announcement that I will be doing a follow-up discussion on his talk in five minutes, for those who want to stay. Once the main audience has spilled out of the room, those remaining start re-organising the chairs. Within what feels like seconds, and without anyone seeming to take charge, the meeting room has been mutated from the rows that face the central screen, into an imperfect loop of twenty-seven people facing me in the middle. I wasn’t expecting such a high turnout, and nervously introduce myself. I explain that, as Adam had discussed, I understand that people might be interested in federated decentralised social networks due to privacy, data security issues around centralised storage, wanting to have control over their data and their experience online. I ask: “but considering it is arguably more effort to make the switch to the fediverse, what do we think would motivate people to do so? Is it actually desirable or possible for the ‘masses’ to come on board? And, if so, what’s going to spark this?”
The conversation very quickly turns to Facebook. As the floor goes from speaker to speaker, example after example is given of how ‘trapped’ people are due to the way Facebook works. Despite people’s protests, or mine and others’ attempts to steer the conversation in a
different direction, it seems impossible to avoid the topic of Facebook. “The problem is many people stay on Facebook for the events – loads of listings are not advertised anywhere else – I don’t understand why none of the [fediverse] platforms are doing this”, one speaker notices. “It’s a real problem in academia”, says another, “students on courses have to be on Facebook because all their course groups will be sharing homework on there”. Filip interrupts: “why are we talking about Facebook? It’s outdated, no one under thirty uses it”. “Right”, someone else responds, “but maybe that will mean the publishers will be the ones to push people away from Facebook, since Facebook forces them to pay to share things”. As the conversation progresses, similar themes keep emerging: “What’s really annoying is how activists still insist on using Facebook to organise” […] “The journalists I work with find that 70% of their traffic comes through Facebook. That’s crazy. That’s way too centralised” […] “If there’s enough pressure, we could open up the protocol so that Facebook uses it too” […] “We shouldn’t want that at all. We want these private closed-source platforms to be gone. Everyone on Facebook is boring. The whole point of Mastodon is that the people I want to talk to are on there.” Adam is visually frustrated: “Seriously, can everyone stop talking about Facebook? Mastodon has nothing to do with Facebook. It’s not an alternative to Facebook.”
The second thing that the decentralised social web isn’t is mainstream social media. As the reader can see, the ‘alternative’ status of the fediverse is a pivotal part of its ontology. When trying to discuss the fediverse, the conversation continually turned to the perils of Facebook and its hegemony in certain arenas: student groups, as a news aggregator, for organising events – including by activists. The keenness to promote the existence of the fediverse as something fun and exciting in itself is met with the fact that it exists as an inversion of something else, something perceived as negative: the commercial, centralised social media that are popular today.
Most of the platforms on the fediverse act as an alternative to a specific mainstream social media platform. For example, Mastodon uses the Twitter format, while Friendica and Diaspora act like Facebook. Peertube fills the role of Youtube. These platforms do not necessarily try to deny this status – Mastodon developer Eugen promoted his platform as a “Twitter without the Nazis” (Glaser and Oremus, 2018), but the desire for them to be something more or different than also exists, creating a tension. While having their own unique features, they are often modelled on and work very similarly to the social media platforms which are more well known.
This state of being an inversion of mainstream social media is an interesting manifestation of the negotiation of techno-utopian and techno-dystopian narratives with relation to
technological change. By being an anti-Facebook (and anti-Twitter and anti-Youtube) alternative platform, the decentralised social web contains both the concerns felt towards social media and the fantasies of what it could be. While mainstream social media websites are seen as enabling data-mining, surveillance and even thought-control, giving power to large corporations and elites at the expense of the sovereignty of everyday people, the decentralised social web is, by extension, the opposite of that. By giving users the power to manage their own data, the power balance is shifted away from the powerful. By enabling users to tinker with the infrastructure as they please and to have a say in how these platforms are moderated, users become empowered rather than exploited. Contained in this alternative status is the idea that if Facebook is destroying the world as we know it and the original potential of the internet, then an anti-Facebook can save the internet and the world.
Part of this alternative nature is manifest in the types of communities that have migrated to the fediverse. As well as free software advocates, the fediverse is full of communities who for whatever reason feel alienated from mainstream social media. While I will discuss this
further in Chapter 2, it is important to note that many of the communities that are currently on the fediverse are ‘alternative’ groups: followers of extreme far left and far right political movements, vegans, sex workers, or those with otherwise controversial hobbies or interests. Often these are groups that are not just controversial in mainstream society – but are
alternative in that they have been actively barred from the mainstream social media platforms. Thus integral to their usage is a need for something different, to be part of
challenging a status quo that is not built for them, and instead building something better in its place.
At one point during my fieldwork, I encountered an advert online to an event in Rotterdam, advertised as a “general introduction to Mastodon and the fediverse”. This was run by an artists’ collective who were interested in the social application of technical ideas and involved a few talks by people who hosted servers on Mastodon, as well as others who had never heard of it before. I got chatting to a stranger while having a drink at the end who told me that his reason for coming along to the talk was that “I just quit all social media, so now I’m finding out what else there is”. In a world where social media is seen to be a threat, these alternative social media act as a promise to reassemble those building blocks, to transform something dystopic into something utopic. Because while the decentralised social web is many things, it is not Facebook.
It’s not capitalist
An important part of most fediverse projects is that they are based on “free (libre) and open source software” (FLOSS). The FLOSS community has a long and well documented history going back to the early days of computers and hackers (Kelty, 2008). It is often said that ‘free’ in this context implies freedom, not free of charge or gratis. Richard Stallman (2017), creator of the GNU operating system, famously describes this as “free as in speech, not beer”. There is a whole world of debate among what exactly free and open source software is (and whether it should be called free, or libre, or open source) with various different schools of thoughts and thought leaders involved. According to Stallman, who advocates for ‘free’ rather than ‘open source’, the difference is that ‘open source’ projects often use licenses that do not give users access to all of what he calls the four freedoms of free software:
The freedom to run the program as you wish, for any purpose (freedom 0). The freedom to study how the program works, and change it so it does your computing as you wish (freedom 1). Access to the source code is a precondition for this. The freedom to redistribute copies so you can help others (freedom 2). The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this. (Stallman, 2017)
For example, the source code may be available to see, but not to distribute modified versions to others. The projects on the fediverse tend to designate themselves as ‘free and open source’ rather than just open source. While free may not mean “free as in beer”, debates about capitalism are still ever-present in the construction of the decentralised social web.
Firstly, arguments that source code should be left open often hinge on the idea that source code is a form of intellectual property. Software is thus understood as a form of information, which it is suggested should be shared, rather than something that can be owned and sold. When source code is open, anyone can replicate it and so using a simple transactional model that is employed with scarce products does not work. In turn, open source projects usually require alternative forms of monetization. For example, they may provide the software for free but ask for payment for services.
In the case of the fediverse, these open source projects have to deal with the inevitable problems of funding in open source projects. Competing with the web services that people are used to receiving for free means that it is difficult to convince users to pay for the
software in a transactional payment model. However, these projects also require resources to keep them going, such as contributors’ time and keeping the servers running with storage space and electricity – creating a tension. Presently, many Mastodon servers are dealing with this issue through crowdfunding. The moderators of the servers may have different ways of dealing with the problem, but on the instances I observed, I noticed people tend to send out the occasional reminder that they are doing it for free and that any contributions are
appreciated. In other cases, they may run them on a pay-as-you-can model. The difference being an expectation to pay if you can afford it, as opposed to a donation. In other cases, they may be running a server for other reasons and not feel they need to ask for compensation for their running costs because it’s “not enough to worry about”. As shown in following quote by the main Mastodon developer, crowdfunding is considered to be a viable way of covering the running costs of the projects.
It is funded by the community. Crowdfunding is big in 2018 and it’s proven that it works. I mean many years back crowdfunding was not a thing and websites or software that relied on donations was not doing very well at all. But in 2018 people finally understand the value of being a patron to something that they value that does not necessarily have commercial value. And we now have tools to redistribute small amounts of money to creators and people who do something that you like so most mastodon servers are crowdfunded by their users and my work on mastodon is likewise funded by my patrons on Patreon. (Eugen Rochko/gargron, Mastodon developer, in Glaser and Oremus (2018) [live interview over podcast, transcribed by me])
The second way that the fediverse’s open source nature is entangled in debates about capitalism is that the way in which open source projects are built can be argued to be a fundamentally different economic model. Eric Raymond is a software developer and anthropologist who wrote the seminal book The Cathedral and the Bazaar, who played a pivotal role in the widespread adoption of the term ‘open source’ instead of ‘free’ software. In this book, which was lent to me by Adam, he argues that the open source approach to software development “is most effectively understood not in conventional
exchange-economy terms but as what anthropologists call a gift culture in which members compete for status by giving things away” (2001: 115-116). By working collaboratively, he argues, open source software is built not due to the greed of individual contributors – dreams of future profit for entrepreneurs, the salaries of workers, shareholder profit expectations for
businesses – but due to the desire of individuals to build reputations, and for the expectation that contributing to someone’s project today means that someone else will contribute to your project tomorrow. It can thus be argued that the open source development model is a
