Try   HackMD

Chapter 2: The platformisation of social media

Missing elements

  • GUILLAUME: Enclosure as part of platform dynamics (walled gardens, filter bubbles, etc.) -> First or second step? [Bruns book on Polity Press against bubbles? Bruns, Axel. 2019. Are Filter Bubbles Real? Digital Futures. London: Polity Press.]
  • GUILLAUME: Moderation -> Step three
  • [DONE] Dissappearance of public infrastructures [@Plantin+Lagoze+Edwards+Sandvig2018]
  • APIs, technical means and strategies of enclosure (Helmond 2015; Helmond, Nieborg, and Vlist 2019) > open standards = public protocols = interoperability (and the prefix "inter" is important because it implies a symmetrical relationship between entities) whereas APIs = selective, granular, revocable compatibility, and an asymmetrical relationship) [Clarify notions: open standard, protocol, API] -> Second step (technical enclosure)

Word count: 5500 words

Table of contents

  • A multistep historical process
  • Early forms of social media
  • First step: The privatisation of chat venues
  • Second step: Technical enclosure
  • *Third step: Marketification
  • The social media market: current and future

A multistep historical process

In the previous chapter we tried to show what platforms are today and their significance in the current media landscape. However, the rise of platforms is a historical phenomenon that is the result of contingent processes. Therefore, in this chapter we take a diachronic approach in order to show how platforms came to be. We will refer to this as the process of platformisation.

Platformisation received increased attention in recent years. We identify two alternative ways of defining the term. The first one is to refer to platformisation as the process by which digital infrastructures (like the Web) are converted into platforms (Helmond 2015; Plantin et al. 2016). We call it a "meso" view of platformization as it focuses on specific socio-technical devices. In contrast, the second view, more broadly encompassing, considers platformisation as a deep, global process in which the whole of society is transformed by the pervasiveness and the logics of platforms (Poell et al. 2019; @Bratton2015). Our own conception of platformisation is closer to the "meso" view, insofar as we are interested in the specific steps by which open, public digital media got turned into, or absorbed by digital platforms.

The platformisation process did not happen overnight. There were gradual steps taken by the various commercial actors that would become the giants of the web as we know them today the "GAFAM" (Google, Apple, Facebook, Amazon, Microsoft). This phase, that could roughly be situated in the first decade of the millenium, can be called the proto-platformisation of the Internet in general, and of social media in particular. The first step is privatisation of the communicational space. The second one is its technical enclosure through the restriction of its interoperability with third party clients. The third step is marketification, that is turning the device into a marketplacepreferably multisided. Soon after will come the adoption of the platform model, from the examples of newcomers like Facebook and Twitter.

Using the history of chat devices as a school case, we put forward that what happened to chat devices is indicative of a general transformation of the whole digital media ecosystem in the last three decades. We divide the analysis to a prehistory, followed by three eras that stand for the three steps in the platformisation process. In each step, we will focus on a particular aspect of platformisation that characterised that historical moment.

At the end of the chapter, we review our improved understanding of the current landscape based on what we now know about its history and how it conditioned its present state. These arguments set the ground for the subsequent chapters, because they let us see the current outcomes as contingent, rather than predetermined and unchangeable. Historical analysis gives back the agency of current users to imagine that things can be otherwise, rather than accept the current state of affairs as necessary and unquestionable.

Early forms of social media

Nonwithstanding the critique we made of this term in the introduction of this book, if by 'social media' we mean a computer-mediated communication device aimed at many-to-many communication, promoting participation of a large number of users and supporting online communities formation, then we can trace down a number digital devices with "social affordances" [@Hogan+QuanHaase2010a] predating current platforms. Put simply, two large categories of social media can be distinguished according to their predominant temporal structure of interactions: synchronous and asynchronous social media.

Asynchronous devices are epitomized by the metaphor of the message board: some user "posts" a message that sticks around until some other users see it and reply to it, generally within a delay of several hours or days. Bulletin board systems (BBSes), federated or not within the FIDOnet network, are a good example of these devices. Another example is the USENET protocol supporting the Internet "newsgroups", which has been eventually semi-platformized as Google Groups. An even earlier device was "Notes", a structured conversation system hosted on the PLATO educational system in the 1970s. It was later commercialized as a key feature of Lotus Notes, a computer-supported collaborative tool suite that failed to become a platform. This generic idea of a structured conversation made of posts within topics, also called an "electronic forum", came in different forms and shapes, depending of the interactional affordances of the underlying infrastructure. Electronic mail was thus instrumentalized as an infrastructure for group discussion, using an automated distribution (remailer) program or 'list server'. The first implementation of this kind was ListServ on BITNET. The open source program became so successful that its name listserv became a generic term for automated email list management programs [@Grier+Campbell2000a].

Synchronous social media are epitomized by the practice of real-time written conversations called "chat". These devices have a multi-threaded genealogy (Latzko-Toth and Maxigas 2019 [FIXME]) intertwined with the history of online games. They often referred to predigital communication technology as suggestive metaphors (Latzko-Toth 2010 [FIXME]). For instance, the citizen's band (CB) radio technology used by truck drivers inspired Compuserve's CB Simulator chat service as well as ForumNet (also known as ICB or Internet Citizen's Band) and more broadly established the concept of conversation "channel" for public or closed group communication, while the telephone and portable short-distance radio devices or "talkie-walkies" inspired private chatting programs such as Talk, Phone and Party-Line in various early digital systems and networks.

The burgeoning of chat devices during the years 1980s-1990s seems to result from a desire by early users to appropriate real-time computer-mediated communication in order to develop social ties on networked computer systems (see Latzko-Toth and Maxigas 2019 [FIXME]). The trajectory of IRC as a digital device is interesting because it emerged among a quite large array of Internet-based chat services and eventually crystallised as the de facto standard infrastructure for online chat during in the 1990s, just like Relay of which it is inspired from came to be for the BITNET network established between IBM academic mainframes in the late 80s.

IRC is an Internet-based chat protocol created in Finland circa 1989. In the 1990s and early 2000s, it allowed millions of Internet users to have real-time, polyphonic, text-based conversations (Latzko-Toth 2014 [FIXME]). Officially released under the General Public License in 1990 and soon after described as a generic public protocol (notably in IETF's RFC 1459), it started as a modest program with limited features, and evolved into a large decentralised technical infrastructure consisting of myriad independent networks of servers. A handful of these networks (e.g. EFnet, IRCnet, Undernet, DALnet, freenode) attracted the majority of users, while many others were single-server chat devices running in private organizations, or as back-end chat engines for public websites called "webchats". In order to talk with each other, IRC users need not connect their clients to the same server, but the servers through which they connect must be part of the same network. Although users can have private conversations on IRC, they most usually engage in public conversations by joining topic-based channels.

Here we focus on synchronous social media, but a similar argument could be made about asynchronous ones by looking at the case of mailing lists or so-called "listservs". The mere fact that listserv as a group e-mail discussion managing program became a generic word in English to name this type of early (but still enduring) social media is indicative of its de facto standard nature. Although there are contenders (e.g., Mailman, ListProc), the key concepts of discussion list managers program remain the same and quite stable since the inception of listserv on BITNET. Even more importantly, all major list managers are free software.

This brief overview shows that social media can be traced back decades before the advent of the first socio-digital platforms such as MySpace, Facebook and YouTube. There was an age of computer-mediated communication were social media where thriving outside the platform paradigm. In the following sections, we take on a description of the chronological steps that led to the understanding of social media as a synonym of socio-digital platforms.

First step: The privatisation of chat venues

The first step involves an epochal shift from public services provided by different public or parapublic entities such as universities, to private services provided by private companies, which were called Internet Service Providers (ISPs). @Plantin+Lagoze+Edwards+Sandvig2018 explain this change in terms of structural transformations in society: in this case, the rising hegemony of neoliberalism, an ideology that favoured privatisation of public services, and the provision of utilities through public-private partnerships. This changed what a service is, in terms of access: now users had to "sign up", which does not simply mean registration, but also to accept and abide by the rules of the service provider. Literally, entering a "chat room" became the equivalent of accessing a private property. Eventually, this had an effect on the interface and how communication was mediated. In our account we emphasise the shift from public discussions around topics to private messages between individual persons with fixed accounts.

We can identify three aspects of privatisation. First, the shift of ownership from public to private. Second, the shift in access from logging in to signing up. Third, a subsequent reorganisation of the interface around user profiles and one-on-one interactions. The change in the political economical basis of chat resulted in changes to access, which lead to changes in the interfaces. Privatisation was not simply a change of ownership, but it lead to the profound reorganisation of chat as a socio-technical device, changing the sociality chat mediates.

In October 2006, Microsoft shut down its MSN Chat service, closing thousands of chatrooms to the dismay of their users. The official reason put forward by Microsoft was the lack of profitability. The move started three years before, in September 2003, when Microsoft closed its chatrooms in Europe and Asia, officially to protect youth from sexual agressors. Chatrooms stayed open in countries where their access was subscription-based, like in the United States and Canada, "where MSN ha[d] billing relationships that enable[d] it to track abusers if necessary" (Microsoft 2004a, 43). Interestingly, Skype also put its "public chats" to an end with the introduction of Skype 4.0, in 2009. As a tech blogger Courtney (2009) observed at the time:

Skype for Windows Product Manager Mike Bartlett claimed yesterday, during an interview, that Skype was reviewing how to embark on "public conversations" in today's messaging world where services such as Twitter and Friend Feed also provide ongoing dialogues. However, Skype Public Chat has its own "space" in terms of user community; it needs to be brought back as soon as possible.

It was never brought back, in line with a trend that followed all the major proprietary chat device providers. Corporate owners had to take on the policing of chat devices as privately owned public space, akin to shopping malls that use security firms to maintain a legally required baseline of order on their premises. We argue that this was an important shift in the history of chat, because the previous generation of chat devices were understood and operated as public services, more akin to bus stops or the postal service, where policing is less of a concern.

As mentioned before, IRC networks have no corporate overlords, neither is there a central IRC authority that owns the totality of the infrastructure. Instead, each IRC network is made up of federated servers often run by otherwise unaffiliated actors, from individual geeks through academic institutions to IT companies or even criminal organisations. So much so, that upon logging in to a mainstream IRC network today, it is actually hard to find out who is sponsoring the resources behind the server. The model of Internet-wise, community-run, community-policed and community-developed communication resources may seem atavistic today, when even starry-eyed activists think that it is impossible to change the world without becoming entrepreneurs and finding a "sustainable" business model. However, running the infrastructure as a commons works for IRC, now just as well as in the 1990s. It allows users to understand and control the media they use to share and collaborate: an essential condition for nurturing technological sovereignty.

While IRC continued to run according to the values and practices established at the time of the early Internet, the dot-com bubble burst at the turn of the millenium. Following the bankrupcy of many start-up companies that experimented with innovative business models, market concentration set in within the online services sector, establishing social networks as the most lucrative business model. Only the companies that profited from user data collection through social networks survived, so that this business model based on user surveillance emerged as a de facto standard in the sector.

Surveillance as a business model warrants a closer look (Zuboff 2015). On the one hand, social media profits are turned through datamining and profiling of users, which enhances targeted advertising to the same audience (Fuchs 2015). On the other hand, social media monopolies establish themselves as obligatory passage points for data traffic to and service provision by other actors. They exploit this leverage for capital accumulation and market positioning. This process has been described as the "platformisation" of Internet services (Helmond 2015), which we described at length in the previous chapter.

Rather than a service in its own right, instant messaging after platformisation seen as a core component of social networking. While presence and contact management distinguished IM from other participation frameworks, the rise of social networking led to a hybridisation with chat devices. By now, they can scarcely be discussed as a distinct form of mediation, and therefore, sociality.

Today, users encounter chat integrated into social media platforms (such as Twitter, Skype and Facebook) or, increasingly, IM on mobile platforms (such as WhatsApp, Telegram, and Signal). While these are generally built with person-to-person communication in mind, over time group chat features have been added to their functions, which is exploited by users to form topical communities. Therefore, some IRC-like dynamics emerge against the script of these devices.

Second step: Technical enclosure

We showed that after a long and parallel history of chat devices, in the 1990s they consolidated into IRC. Then, the next generation of chat devices were Instant Messengers (Maxigas 2014). In the previous section, we considered the shifts in the political economics, interface design and sociality of chat that these next generation of chat devices brought. We emphasised the change of ownership, the switch from logging in to signing up, and the transition from topic-centered to profile-centric services. In this section, we consider the technical developments behind the scenes that accompanied, and in many cases followed up this paradigm shift.

The unreachable standard

In April 1997, a "bird-of-a-feather" on "Internet Relay Chat Update" was held in Memphis, Tennessee, in the context of the 38th Internet Engineering Task Force (IETF) Meeting. The main point at the agenda was that there was "no draft on the standards track addressing chat" at the time, as noted by the meeting chair, John Noerenberg, a Qualcomm representative (King 1997). The meeting ended on the decision to create an IRC Update Working Group (IRCUP) which would take the form of a mailing list gathering various actors interested in IRC standardization. One of the objectives of the process was to develop a perfectly scalable system, suitable both for private mini-networks (at the scale of an organization) and for vast public networks able to accomodate up to "one million users on 100,000 channels" (Ibid.), which was deemed to be a considerable number at the time.

Among the participants in the IRCUP working group were some key actors of the development of major IRC networks, including Microsoft. One of its spokespersons, Thomas Pfenning, had previously submitted a working paper proposing a series of protocol extensions under the name "IRCX" (the most recent version is Abraham 1998).

The IRCUP mailing list was active from January to July 1998, but its members failed to agree even on a common agenda, and it was eventually dissolved . Microsoft reiterated its proposal in April 2004. Its goal, it said, was to overvome the 'ill-fated' IRC standardization effort, and unify the archipelago of chat devices and IRC server code variants around a robust standard, offering a refined gradation of sociotechnical roles and adding several features to the protocol (2004b). Furthermore, Microsoft wanted to make IRCX the chat protocol of its "Microsoft Exchange Server" platform, which up-to-then used a proprietary protocol, Microsoft Internet Chat (MIC). Having failed to convince IETF and the community of IRC developers, Microsoft abandoned the project and focussed on a proprietary protocol, the "Messenger" protocol, which would underlie its instant messaging system, MSN Messenger, launched in 1999.

In a message dated February 1, 1998, Peter de Vries, one of the members of IRCUP, sums up the difficulties experienced in trying to agree on a standard for IRC:

The "dream" of IRC3 means different things to different people. Everyone would agree that IRC3 must "improve on the original concept and expand it into logical new areas to meet modern needs". But .. the agreement stops there. Sure, there are some issues that everyone agrees needs fixing (ie: scalability). But there are other concepts in IRC2 that some people think are "bugs" while other people think they are "features" []. Even ideologically [] there is disagreement. [] Some people believe that certain extensions would are "logical extensions of the paradigm".. that they would have existed already "if cpu or development time had allowed" or "if the problem existed then". Others take a look at the same concept and don't see it jiving with their "vision" of IRC at all. I think the "difference in vision" problem is a big one. IRC3 is a big problem many possible directions.. many possible (& divergent) solutions. This is no normal WG [working group]. [] We are still at the explosive "determine the paradigm" section [].

On the backend (Stalder 2013), IMs used a similarly heterogeneous range of proprietary protocols tied to centralised infrastructures, instead of the community driven protocols of IRC that was tied to it federated infrastructural model. This led to a balkanisation of the chat landscape, reminiscent of the early Internet, where different chat devices operated side-by-side without being compatible with each other. However, the difference was that this time the same users installed several alternative chat client software on their computers in order to stay connected with people who used one or the other IM. In the days of the early Internet, balkanisation had a different character, because users themselves were separated from each other according to their access to particular networks, whether the Internet or Compuserve, for instance.

The XMPP (Extensible Messaging and Presence Protocol) family of standards was proposed as a solution by open-source advocates in 1999. The open protocol aimed to make proprietary Instant Messengers compatible through a common language, addressing the problem of market fragmentation where messages cannot cross corporate silos; to allow anybody to provide compatible IM services, decreasing the grip of social media monopolies on users; and to support the seamless integration of messaging into the widest possible range of technologies, encouraging hybridization. Proprietary IM operators progressively added various levels of XMPP support (Skype 2003; Google Talk 2005; Facebook Chat 2008; AOL 2008). Ironically, XMPP itself soon fell victim to fragmentation due to its many extensions. Its popularity fell from ten million users in 2003 when it reportedly surpassed ICQ users to a level where its relevance itself is debated (XMPP Standards Foundation 2003; Quora 2015). Monopolies progressively dropped XMPP compatibility, yet incompatible implementations still form the core of many popular services such as WhatsApp, PlayStation, or Facebook. A similar contemporary attempt is the Matrix protocol which pitched platformisation as its ultimate goal from the very beginning, even though it is still unclear whether users really need IM interoperability [@Weinberger2014a].

Efforts targeting interoperability such as XMPP and Matrix are sometimes referred to as the vision of the Open Web, and associated with the emblematic figure of Aaron Swartz [@Knappenberger2014a]. The Open Web is a systematisation of early Internet practices, taken from the moment where the blogosphere constituted the contemporary frontier of digital culture, yet its territory was already threatened by colonisation from social media platforms.

In the Open Web, users and communities can develop their own very different communication devices, while maintaining the flow of information between devices through open standards and stardardised interfaces [@Helmond2015]. For instance, the RSS standard (advocated by Swartz, amongst many other things) allowed one to gather and read news from mainstream broadcast media like CNN, small personal websites, and niche publications, using the same software client and a consistent presentation of the downloaded content. Mash-ups have been a paradigmatic web making practice in this period, as they rely on remixing content from various publicly accessible sources in order to foster new experiences, use values and consumer practices. The IRC bots [FIXME: check this when we have the previous part] already mentioned in this book often rely on such practices to interface with public information and mediate access to services to chat users. Thus, "islands in the net" [@Stephenson1988a] could develop autonomously in the same territory, without becoming self-enclosed entities.

Technically, the design goal of interoperability is achieved through open standards and standardised interfaces such as the RSS (Really Simple Syndication) format described above. Websites or any other device linked to the Internet can provide their content in the RSS format, making it available for other devices to consume. Therefore, content is shared as part of the digital commons. Consumers of information published in RSS can make their own choices about how to use it, present it, filter it, etc. The political ideal of openness is achieved through providing symmetrical and standardised interfaces to access content on the Open Web in presentation-agnostic and programmatically readable formats.

Anne Helmond contrasts the technological imaginary of the Open Web to platformisation. She traces the technical process through which digital enclosures have been established in the digital commons. Even though the Open Web have only been realised to a limited degree, the comparison is crucial for the presentation of platformisation. The two involve the opposite kind of power dynamics, but they may appear to be very similar, or even undistinguishable on a cursory look. This is not a coincidence or an accident. This is the result of a consciously crafted "politics of platforms", as observed by @Gillespie2010a, where social media monopolies masquarade as carriers who democratise access to media and information, while trying to avoid the responsibility of publishers for policing content.

@Helmond+etal2019a show how platforms such as Facebook used APIs (Application Programming Interfaces) to colonise the World Wide Web. APIs are very similar to open standards and standardised interfaces in that they provide presentation-agnostic and programmatically readable access to content. However, APIs are designed to allow granularity and flexibility in how much access is permitted. They determine what is accessible and under what conditions. They also allow content publishers to change their policies at arbitrary moments, so that the shape and power of the public interfaces they provide can follow the twists and turns in their marketing strategies. Therefore, APIs are not open stardards designed to foster symmetric power relations in a digital commons, but strategic ways for major actors to advance their hegemony over what they may see as the digital frontier.

How social media monopolies colonise the web through technical means? The argument of Helmond and her collaborators is that this is a two-way process. The power differential is materially expressed in the fact that Facebook provides its content to third parties through arbitary APIs under the control of the corporation, while all other publishers on the Internet are supposed to use the rules set by the same corporation in order to make their content "Facebook ready" [@Helmond2015a 7-8]. The end user visible manifestation of this is the ubiquituos "Share" button visible on websites from mainstream news organisation like CNN to personal blogs and other social media sites. The Share button allows users to import content from other websites to Facebook, mobilising the standard markup on the web page and the Facebook API. While Facebook can change their API specifications and they markup standards at any time, it is expected that publishers make all their content always already "platform ready" in order to accomodate user expectations. This is how digital enclosure is technically implemented, and how the hegemony of platforms is asserted over other parts of the Internet.

Helmond's empirical data and theoretical argument of platformisation through APIs is convincing. However, it does not account for the emergence of APIs themselves as a technical measure in the service of a functional business model and a tool for articulating the sovereignty of platforms. What we add to her argument is the prehistory of APIs (proto-platformisation) through federation practices and protocol compatibility.

Federated networks and common protocols offer two alternative visions of "how things might have been otherwise", e.g. alternative technological trajectories, on which we will draw on Chapter 4 (What are the alternatives?). In this sense they are the basis for contemporary technological practices that challenge the hegemony of platforms. At the same time, they are also part of the story of platformisation, which we tell in this chapter.

It was already 1996 when Microsoft included an IRC client in the default installation of its popular Windows operating system, taking note of IRC's mainstream appeal (Kurlander, Skelly, and Salesin 1996). In the first major attempt to recuperate IRC, the software was developed by the company's Artificial Intelligence research unit, and the application connected automatically to the company's own IRC servers (Latzko-Toth 2010). Ironically, the Comic Chat IRC interface was never popular with users, and the only artifact that went down in history from the whole enterprise was the Comic Sans font, which is still the laughing stock of Internet users. Microsoft never figured out how to make money from the largest online chat phenomena of the time.

Federated networks and common protocols worked well in terms of engineering standards and served significant use cases of chat, but lacked the appropriate political economical foundations for capital accumulation, e.g. building profitable business models on them. This accounts for both their emergence and their demise. At this point there is a strong belief in the IT industry that can be summerised in the adage "Build it and they will come". Providing services which users adopt was seen as the primary imperative for capturing a market segment such as chat, while clarifying how to turn user activity into a revenue stream was secondary. Such an attitude does not amount to economic irrationality though, since the price points of publically traded companies reflected the same beliefs, so that more users translated to more money through the stock market. Financialisation meant that value came from capitalising on trust in the markets rather than varolirsation further down the value chain. Therefore, in the second step of platformisation the primary objective was integrating users into services that later evolved into platforms.

Now it is clear why Microsoft provided access to Windows users to IRC networks, and why, some time later, Google decided to make its Google Plus social network compatible with other services such as Facebook Chat through the XMPP protocol. We look at this two examples in more detail below in order to demonstrate the dynamics of the platformisation process. We could draw on many parallel developments in the media environment, but these two cases are enough to illuminate the overall logic of this historical moment.

Third step: Marketification

Capitalist competition is not about making competing products anymore, it's about developing competing markets. The monopolistic position goes to the platform's owner. This is the genuine invention that capital contributes to the process of platformisation. It is also that process that provides a specificity to digital platforms, beyond the notions of commercialisation of digital artifacts and the "corporatisation" of digital venues.

One aspect or form of recuperation is commodification. Commodification is when something at some point becomes a commodity to be brought and sold on the market. Commodification targets authentic things, which are often already perceived to be valuable for instance as a moral good but not yet recognised as an object of monetary exchange. The loss of authenticity through commodification produces anxiety in consumers, which can be diagnosed as the affective trace of capital's violence.

Later, as the World Wide Web took off, chat features were integrated into Web 2.0 social media platforms.

Eventually, surveillance came to be the key means for both maintaining social peace and deepening exploitation on social media platforms.[1] We mentioned this as the basis for market consolidation in Step One when we dealt with the privatisation of chat devices. However, we consider it telling that the change of ownership happened in separate step from the commodification and corporatisation of chat devices. First, public spaces were privatised, and it was only much later that a suitable business model has been found. This shows the crucial difference between economic rationality and political economy: the political impetus of colonisation was first, legitimised later by economic rationality, e.g. actually turning a profit on the services.

In the surveillance capitalist business model explicated by Zuboff (2019), everyday, informal, even intimate gestures are captured and stored, sorted and mined for the purposes of both targeted advertising and targeted repression. Such revenue is indispensable to the capital accumulation mechanisms of a growing section of capital, while the intelligence gained by authorities who share access to the information flows is essential to the maintenance of social order in both dictatorships and democracies. For instance, surveillance technically based on the analysis of log files accounted for 89% of Google's profit in 2014 (Griffith 2015).[2] All this hinges on successful platformisation: the ability of a vendor to install themselves as an obligatory passage point for generally mundane and often minuscule social interactions (Gillespie 2010). The kind of digital milieus where average Internet users chit-chat nowadays have been variously described by scholars as enclosures, walled gardens and social media monopolies (Lovink and Rasch 2013).

The social media market: current and future

There are several current developmens in the market of chat devices, which divert from the main stream of platformisation. We tentatively suggest that these local tendencies and turbulences around chat platforms are not anomalies, but they well be signs of the things to come.

Facebook seems to be emerging as a monopoly in the chat platform market segment, but it has not consolidated its position yet. To be precise, from the three steps of market consolidation, that is, (1) developing a relevant product, (2) buying up the competition, and (3) merging alternative offers into its own product line, Facebook is at the second step. A quick round-up of prominent players establishes this notion. WhatsApp was founded 2009 and Facebook owned since 2014; Facebook Messenger exists since 2011, started three years after WhatsApp; Telegram was founded in 2013 and Facebook bought it later that year; Signal is a non-profit, founded in 2018 by Brian Acton, who co-founded WhatsApp, and Moxie Marlinspike. Marlinspike (a prominent free software hacker and former head of cybersecurity at Twitter) worked on the chat products for WhatsApp, Facebook and Google to introduce the Signal protocol for end-to-end encryption and group chat between 2015 and 2016. In the next few paragraphs, we look at the state of play regarding encryption in more detail, and argue that it may be indicative of the next future step of platformisation.

Counter-intuitively, the bleeding edge trend seems to point away from platformisation (at first sight). Encryption is gaining a foothold in the feature set of chat devices. Encrypted group chat have been the holy grail thought after by political free software developers and hacktivists in the last decade. The last years produced production ready solutions as a result of mostly grassroots and academic research. These are valorised by capital in recent years, being introduced to commercial chat platforms.

There are three reasons for the recent availablility of encrypted group chat in commercial chat platforms. First, a change in the conditions of possibility, or in more simple words, the availability of technical solutions. However, new technologies are only socialised if capital is forced to integrate them in the mode of production, so that this is a necessary but not sufficient reason. Second, pressure from civil society, based on the historical currency of encryption, driven by the eroding trust between users and platform providers. As work moved online due to the COVID-19 pandemic, questions about the privacy of platforms moved to the centre of public attention. Third, pressure from the state, whose sovereignty is challenged by platforms (Bratton 2015), so they increase demands for platforms to take responsibility for policing and moderating user generated content. However, in case user generated content is not accessible to the platform provider, then it is unreasonable to make it responsible for policing. Thus, end-to-end encryption solves multiple problems for platforms.

A logical counter-argument is that end-to-end encryption endangers the surveillance capitalist mode of production. However, surveillance studies have shown that metadata is more valuable than the actual content of messages. Therefore, as long as the metadata continues to be available, the surveillance capitalist continues to be viable. Furthermore, platform sovereignty mainly depends on the ability of the vendor for censorship (e.g. closing offending accounts). Of course, such developments can be understood as just another iteration of the eternal struggle between transgressing users and civilising administrators, which we described earlier.

Figure FIXME shows the prominence of group chat encryption technologies in the programme of the Chaos Communication Congress, the seminal annual European hacker convention. The adoption of end-to-end encryption by commercial chat providers is marked on the timeline for comparison. We can see that the social world of peer production influences industrial practices.

History proceeds driven by contradictions to be resolved on a higher level. Here, the contradiction between the discoursive self-representation of platforms (as neutral carriers) and their perceived functioning (as biased publishers) is resolved through implementing the long term demand for privacy by civil society, but in a way that exploits privacy features to protect capital against the state, not users. The point is that the demand is implemented in a way that simultaneously undermines it.

References

Abraham, Dalen. 1998. "Extensions to the Internet Relay Chat Protocol (IRCX)." Internet-Draft. https://tools.ietf.org/id/draft-pfenning-irc-extensions-04.txt.

Bratton, Benjamin. 2015. The Stack: On Software and Sovereignty. Boston, MA: MIT Press.

Courtney, Jim. 2009. "Skype for Windows 4.0 Goes Gold; Improved UI, Audio and Video Performance Standard." Voice on the Web blog post. http://voiceontheweb.biz/skype-world/skype-markets-skype-world/skype-for-personal/skype-40-goes-gold/.

Fuchs, Christian. 2012. "Google Capitalism." TripleC: Cognition, Communication, Co-Operation 10 (1): 4248.

-. 2015. Culture and Economy in the Age of Social Media. New York: Routledge.

Gillespie, Tarleton. 2010. "The Politics of 'Platforms'." New Media & Society 12 (3): 347364. doi:10.1177/1461444809342738.

Grier, David Alan, and Mary Campbell. 2000. "A Social History of Bitnet and Listserv 1985-1991." IEEE Annals of the History of Computing 22 (2): 3241. https://ieeexplore.ieee.org/iel5/85/18190/00841135.pdf.

Griffith, Erin. 2015. "Bad News for Google Parent Alphabet: The 'G' Will Still Foot the Bill." Article in Forbes Magazine. http://fortune.com/2015/08/10/google-ads-money/.

Helmond, Anne. 2015. "The Platformization of the Web: Making Web Data Platform Ready." Social Media + Society 1 (2).

Helmond, Anne, David B. Nieborg, and Fernando N. van der Vlist. 2019. "Facebook's Evolution: Development of a Platform-as-Infrastructure." Internet Histories 3 (2): 123146. doi:10.1080/24701475.2019.1593667.

Hogan, Bernie, and Anabel Quan-Haase. 2010. "Persistence and Change in Social Media." Bulletin of Science, Technology & Society 4: 309315.

Ippolita. 2015. The Facebook Aquarium: The Resistible Rise of Anarcho-Capitalism. Revised and updated English edition. 15. Amsterdam: Institute for Network Cultures. http://networkcultures.org/blog/publication/no-15-in-the-facebook-aquarium-the-resistible-rise-of-anarcho-capitalism-ippolita/.

King, Ian. 1997. "Minutes of the Ircup (Internet Relay Chat Update) Bof, April 9, 1997, 38th Ietf Meeting." Notes taken by Ian King from Microsoft. Retrieved from the IETF FTP server.

Kurlander, David, Tim Skelly, and David Salesin. 1996. "Comic Chat." In SIGGRAPH '96: Proceedings of the 23rd Annual Conference on Computer Graphics and Interactive Techniques, ed by. John Fujii, 225236. New York: ACM; ACM. doi:10.1145/237170.237260. https://sci-hub.se/10.1145/237170.237260.

Latzko-Toth, Guillaume. 2010. "Metaphors of Synchrony: Emergence Differentiation of Online Chat Devices." Bulletin of Science, Technology & Society 30 (5): 362374. doi:10.1177/0270467610380005. http://bst.sagepub.com/content/30/5/362.short.

-. 2014. "Users as Co-Designers of Software-Based Media: The Co-Construction of Internet Relay Chat." Canadian Journal of Communication 39 (4): 577596. http://www.cjc-online.ca/index.php/journal/article/view/2783/2493.

Latzko-Toth, Guillaume, and Maxigas. 2019. "An Obscure Object of Communicational Desire: The Untold Story of Online Chat." In Second International Handbook of Internet Research, ed by. Jeremy Hunsinger, Matthew M. Allen, and Lisbeth Klastrup, 381394. Dordrecht: Springer. doi:10.1007/978-94-024-1555-1_8.

Lovink, Geert, and Miriam Rasch, eds. 2013. Unlike Us Reader: Social Media Monopolies and Their Alternatives. Amsterdam: Institute of Network Cultures.

Maxigas. 2014. "History of Real Time Chat Protocols." http://relay70.metatron.ai/history-of-real-time-chat-protocols.html.

Microsoft. 2004a. "Protecting Children Online." In Global Citizenship Report 2004, 4243. Annual report for the fiscal year 2004. http://download.microsoft.com/download/7/5/4/7548d2a2-9d22-4d2d-8c13-2521f6a1c148/citizenship2004.pdf.

-. 2004b. "Exchange Chat Features/IRCX." DOCS previous versions documentation. https://docs.microsoft.com/en-us/previous-versions//cc767140(v=technet.10).

Plantin, Jean-Chrisophe, Carl Lagoze, Paul N. Edwards, and Christian Sandvig. 2018. "Infrastructure Studies Meet Platform Studies in the Age of Google and Facebook." New Media & Society 20 (1): 293310. doi:10.1177/1461444816661553.

Stalder, Felix. 2013. "Between Democracy and Spectacle: The Front and the Back of the Social Web." In Unlike Us Reader: Social Media Monopolies and Their Alternatives, ed by. Geert Lovink and Miriam Rasch. Amsterdam: Institute of Network Cultures. http://felix.openflows.com/node/223.

Weinberger, Matt. 2014. "Matrix Wants to Smash the Walled Gardens of Messaging: Introducing Matrix, a New Messaging Standard That Wants to Make Im, Voip, and Video as Interoperable and Open as Email." Article in IT World from IDG. https://www.itworld.com/article/2694500/unified-communications/matrix-wants-to-smash-the-walled-gardens-of-messaging.html.

Zuboff, Shoshana. 2015. "Big Other: Surveillance Capitalism and the Prospects of an Information Civilization." Journal of Information Technology 30 (1). doi:10.1057/jit.2015.5.

-. 2019. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. New York: Public Affairs.


  1. "The legacy of the 20th century has accustomed us to think that social control pertains only to the political, but it has long since become primarily an economic question of commercial implications. It is no coincidence that the NSA has made use of the collaboration with Microsoft, Yahoo, Google, Facebook, Apple and so on, to obtain data for the surveillance program PRISM." (Ippolita 2015, 7) ↩︎

  2. "Google is a profit-oriented, advertising-financed moneymaking machine that turns users and their data into a commodity." (Fuchs 2012, 47) ↩︎