Abstract
In this interview, José van Dijck distinguishes the concept of deplatformization from deplatforming and platformization. It describes the phenomena of the systematic pushing back of controversial platforms and their communities to the edge of the platform ecosystem, dominated by mainstream platforms (such as Google, Apple, Facebook, Amazon, and Microsoft). Deplatformization further demonstrates the hierarchical power relations within the global platform ecosystem and the complexity of platform governability. From the European perspective, van Dijck argues that public values and public institutions should play more active role in platform governance. The recent Russia–Ukraine war also indicates the vulnerability brought by such co- and inter-dependence on American platforms.
José van Dijck is a distinguished university professor in media and digital society at Utrecht University. She is a world-renowned scholar in media studies and the author (and co-author) of many influential books including The Culture of Connectivity, and The Platform Society. van Dijck is the former president of the Royal Netherlands Academy of Arts and Sciences and the recipient of Spinoza Prize in 2021. In recent years, van Dijck’s work has centered around the topics of platform economy, platformization, and its related challenges to the public society. In this interview, we start from the concept of deplatformization (originally developed in van Dijck et al. 2021) and discuss the urgency and complexity of platform governance in today’s world and its relations to contemporary geopolitics.
How would you distinguish these concepts: platformization, deplatforming, and deplatformization?
Yeah. The concepts may soundvan Dijck): confusing, but actually, once you start to analyze them, they’re very important not just as single concepts, but they’re all part of the same platform ecosystem as I call it, because that is what all the powers sort of relating to.
Now, let me start with deplatformization. Here is the core concept that I would like to just explain. But it can hardly be understood without referring to platformization. In earlier work, Thomas Poell, David Nieborg, and I describe platformization as the penetration of online infrastructures, economic processes, and governmental frameworks of online platforms in different societal sectors and spheres of life (Poell et al. 2019). But in practice, it refers to how a handful of tech companies actually operate an ecosystem of platforms in which these tech companies cooperate and compete at the same time. Of course, those big five companies, Google, Apple, Facebook, Amazon, Microsoft, coordinate that platform ecosystem in a way that gives them a lot of power.
Taken from that platformization concept, two other terms are important here: deplatforming and deplatformization, which are not the same. Richard Rogers (2020) has published a very important article on deplatforming, which means the removal of one’s account from social media for breaking platform rules. For instance, Donald Trump was banned from Facebook after the Capitol siege. That is a single user account banned from a single social network. Deplatforming may be the ultimate step in a platform’s moderation process. Usually, account holders are not de-platformed right away. If they break the rules, that is the terms of use that they have consented to, they get into a procedure—a platform usually first issues warnings, for instance, by flagging or removing pieces of content, then suspends someone’s account temporarily before eventually deciding on a permanent ban.
Beyond that one platform, there are also systematic purification attempts, which we call deplatformization. It goes much further than deplatforming. Deplatformization applies to tech companies’ efforts to reduce toxic content, mostly by pushing back controversial platforms and their communities to the edge of the ecosystem. They do this by denying them app access to basic infrastructural services: for instance, pulling certain platforms from the app store.
So deplatformization is the flipside of platformization. Platformization refers to the infrastructure on which the ecosystem is built. Deplatformization means the companies that operate platforms police the ecosystem to keep it clean from disturbances. These “hygienic” measures apply to a wider net than just individual users on single platforms, such as social media platforms. It also applies to app stores, cloud services, pay systems, and so on—much deeper down the stack.
Sounds like this is a systematic canceling of a small number of platforms from the side of those dominant players. If we think about this from the perspective of the public society, I can see the need for deplatformization. But if we take from the perspective of those dominant or mainstream platforms, why would they do this? Especially if you think about the relationship between these platforms, they’re codependent, which means they are financially beneficial to each other. So what are the motivations for deplatformization？
That’s a very good question. Why would they want to do that? After all, they monetize these platforms, they’re making money of them by extending their networks and network effects which means they grow bigger when they have more users.
At first sight, it sounds very irrational for big platforms to de-platformize other players in this ecosystem. But the efforts to apply deplatformization are indeed double hearted. On one hand, there is social and political pressure from users, society, and political actors, to remove alt-right platforms from mainstream services. For instance, we saw that after the Capitol Siege last year, with a lot of alt-right platforms involved like Gab, Parler, and BitChute; they were almost instantly disconnected, within 3 days, from mainstream platforms like YouTube, Twitter, Facebook, which were not just under the pressure of certain political groups in society. Both their own users and governments wanted these companies to behave responsibly and keep the social media networks clean from hate speech and refrain from inciting violence. More polarized friction may attract more traffic, which is good for business, but too much of that turns off users. So eventually, online hygiene means better (if not more) business, but it also means avoiding regulation from the state. On the other hand, the big tech companies do not want to entirely cut off alt-right users from their mainstream platform services because that is bad for business.
What we have observed in some of these bigger deplatformization moves is that some of the alt-right platforms are removed, but not entirely. They are pushed toward the fringe of the ecosystem. One example that we elaborated upon in the New Media & Society article is Gab. As an alt-right site, it was first removed from Apple and Google app store. It then took refuge to an in-browser extension. When that was disabled by Chrome and Firefox, Gab moved to Brave, which is an opensource web browser. At the same time, Gab was also banned from Amazon Cloud. Instead, it took refuge to an alternative cloud service called Epic, which was sympathetic to the kind of services that Gab offers. But then Epic was also cut off, but not entirely because it was somehow still served by Amazon cloud service.
So, what big tech companies usually do is that they try to marginalize them (alt-right/extreme platforms) by removing them from services but not entirely, because some services that they subsequently rely on are still serviced by the mainstream actors.
It is in fact a very murky business, with pros and cons. By pushing them toward the fringe, they are both satisfying that political and social pressure to block them from mainstream channels and not polluting those channels (ecosystems) that are very important for their business. But the rules by which they play this game of deplatformization are not at all transparent.
This just reminds me of something about civic values. Apparently, all these public values still have certain sway on the practices of these platforms. If we look at the discourses claimed by the alt-right or alt-extremist platforms, they are using similar narratives or discourses about decentralization, openness or free speech. It seems like that these public values have been appropriated by these platforms, but to undermine the same public values. This is like a paradox.
This was a very important observation that we made while doing the research. We realized that over the past years, tech communities that have paved the way for a more or less alternative ecosystem have usually done so from the left-wing ideological spectrum. One example of such alternative ecosystems is the Fediverse. It includes platforms such as PeerTube and Mastodon, which have been relying on “open” tech-concepts, like opensource, decentralization, community moderation, inter-operability, data portability, etc.
Now, what happened is indeed that the alt-rights have seen that you could build an alternative tech universe, and then appropriate it for your own purposes. What alt-rights have done is sort of appropriating these alt-tech discourses, to weaponize them as a disinformation tool, under the pretense of freedom of speech. But the discourses of decentralization and opensource are not the same as that of freedom of speech. They are totally different kinds of arguments. And what a decentralized alt-tech community wants is not that they leave speech unsupervised or totally free or unmoderated. On the contrary, the ideological discourse of the alt-tech is that communities need to moderate themselves in a way that conforms to principles of good governance. They want responsible moderation, which means transparency about the rules. And what we have been observing is that when alt-right platforms like Gab got into the alt-tech ecosystem, which was very community centered and built on open rules, they (alt-rights) start to abuse those open rules. They try to disrupt and appropriate the system by playing, apparently, according to those rules, but for a very different purpose to install and insert hate speech, misinformation, and so on.
Mastodon, for instance, was initially disrupted by Gab. How they responded was that they went back to the communities and informed them about what was happening. Afterwards, the communities on Mastodon decided to ostracize Gap from their system. But it took a huge community effort to not be disrupted by their own rules of opensource and decentralization, which of course means that all kinds of platforms can participate. After observing how that happened with Gab, we realized that there were many more attempts by the alt-rights to disrupt alternative tech communities.
And that should open one’s eyes to the mechanisms behind. Deplatformization does not only happen in the mainstream platform ecosystem. We are somehow forced to constantly play by the rules of that system, trying to keep our platforms, particularly social media platforms, hygienic. Those opensource, decentralization, and other alt-tech rules are also very invitational of another type of ideological discourse. It is not enough just to start an alt-tech community by designing and playing by other rules. You also have to be aware of how other players, especially not-so-friendly players, can undermine the alt-tech system that you have built.
This also exposes the problem of centralization or hierarchy within the mainstream platform ecosystem, which raises the question of platform governability. Why is it so difficult to have an efficient, but also effective governance of platformization and platform economy? What is the central question here when we talk about platform governability?
Platform governability is a hugely important issue. Platforms, particularly the mainstream social media, are now moderating global conversations between individuals and communities without the intermediary governing ethics and rules of civil society actors, such as for instance, institutions. We used to have conversations that were moderated by institutions. Think about schools, think about libraries, think about public broadcasting systems. Having the moderation responsibilities over billions of people in the world without transparent, democratic governing structures, makes these platforms indeed very difficult entities to govern.
So I think the central question here is as follows: How can you govern (local) communities without having to rely completely on market mechanisms that mine and exploit our personal and collective data at scale, and which are the biggest monetizing assets of American big tech platforms? Or, on the other end of the geopolitical specter, in China, these assets are exploited by platforms that are under state control. In practice, platform governance is either controlled by states or by (globally operating) private companies. But we do not have “global states” to look after good governance of tech companies, and neither do we trust “global companies” with their interference in national or local platform governance.
So when looking at these platform ecosystems from the perspective of big five companies, for instance, it is very easy to see how their interests are mostly commercial. They are working in the interest of users, meaning not citizens, but consumers. My central concern here is to find another model. My question is: can there be another model of platform governance—one that accommodates neither states on one hand nor companies on the other hand, but a model that accommodates multi-stakeholder partnerships? Civic society actors, particularly open institutions, have been left out of the American or Chinese platform ecosystems; they can help align all stakeholders to democratic principles without being subject to either commercial or state-centered platform mechanisms that are deployed to commercialize or ideologize speech at a global scale.
So what I am looking for is, a model of platform governance, which I named the “European Model” (for lack of a better term), that is not entirely reliant on commercial companies and commercial mechanisms on one hand, or central governments and state control on the other hand, while giving more power to civic society actors such as citizens and public institutions, putting them in charge of their own data. I think they should be at the center of governance in terms of platform ecosystems.
The civil society, and the civic values, which underline the operation of non-government or non-commercial organizations, should definitely play a more substantial role in the whole process of governance. But maybe a very practical question. How can this be possible? Do you have any recent cases or examples that can be seen as early practices of such idea?
Yeah, it’s not easy to actually find examples that are already implemented and work well. I am currently involved in an experimental stage of a project called “PublicSpaces.” 1 The experiment started two years ago and is just contained to the Netherlands. But what it does is that it puts civic institutions at the heart of online space, which is the process I was just describing. There are 40 organizations, such as public libraries, museums, public broadcasters, involved in trying to create a more hygienic online environment. They want to translate their responsibilities as public organizations into online infrastructures that can carry the online public sphere without having to rely completely on either commercial systems. What typifies these organizations is that they’re independent, civil-society, community or public organizations. They may be funded by the state, but they are independent in operation.
With this platform, we are working on a couple of projects. One of the projects is called The PublicSpaces Digital Powerwash, 2 trying to make the online environment of these organizations more hygienic at different levels. For example, can we reduce the number and also the kind of apps that we use in this public environment? There are a lot of apps that are simply data absorbers. Can we stop them? Can we replace them by other kind of apps that are much more privacy friendly? Another PublicSpaces project I am involved in is called PubHubs. We are trying to design a privacy friendly online community build on the principles of Matrix, which is an opensource network. We are trying to sort of rebuild that interface to accommodate public organizations who want to communicate online, exchange content, and moderate their own discussion space. So they do not have to rely on Facebook, for instance, for their interactions with their audiences or YouTube for their content. We do not think this will happen overnight: development will take a number of years. But what is important for us is that these organizations, as very independent civil society organizations, can create their own environments that are moderated by themselves, and all moderation responsibility will remain with that community. It won’t be centrally organized or centrally controlled.
Yes, all these different cases are very interesting. From the design and operation of these experimental platforms you mentioned, it seems that they tend to frame themselves as alternative platforms against those mainstream players. Does this also follow the logic of deplatformization, in the sense that we are trying to create an alternative ecosystem, and trying to make our online space more “hygienic” and more civically engaged? How can we really compete against those mainstream commercial players, which have always been playing with the discourse of efficiency?
That is a very important question. For one thing, we’re not trying to compete with the big commercial platforms. We’re also not trying to avoid them per se. The thing is that these big commercial platforms need to become aware of the public values that underpin this public space and respect that. Let’s take the offline situation as an example. We have public libraries that are not based on commercial values. We do not want a library search to be organized like a Google search: putting information next to advertising and personalizing that advertisement for each user. We do not want to apply purely commercial principles in a public space. That’s the only thing. If Google can provide an online environment—a cloud space, a search space—that is free of these commercial or proprietary mechanisms, they may provide those services to public institutions. With PublicSpaces, we’re not anti Big Five or against using their services. The thing is that we take issue with the conditions set by commercial providers of cloud services or the app stores. If these conditions are based on public values, conditions that are transparent and can be democratically controlled, we can align the needs of public institutions with the affordances of online platforms. That is basically the cat and mouse game that is being played now.
The European Union is working on the Digital Market Act and the Digital Services Act which will break some of the power of Big Tech platforms. But these expansive legal frameworks may not be enough to safeguard the public sectors in Europe. If we can manage to carve out a specific place that can guarantee public values based interaction, then we can consequently use common platform services. However, European public institutions cannot rely on the current moderation systems offered by global American platform providers, which do not subscribe to public values that are local and contextualized in civic public context.
So it’s not really about one replacing another, one canceling another. It is about the protection of public values and also the coexistence, especially the participation of civic actors in this whole public sphere, in which platforms are becoming increasing dominant.
Absolutely. This is about platform diversity: the co-existence of various types of online platforms and infrastructures, rather than sustaining a monoculture where all data power is centralized.
Another example that we have in the Netherlands concerns online infrastructurals services, such as identification. Online identification and login services are almost always commercially based or issued by states. In The Netherlands, there is now an alternative: an online identification system called IRMA, the acronym of “I reveal my attributes.” I coauthored an article 3 with Bart Jacobs (van Dijck and Jacobs 2020) who was the inventor of IRMA, which is based on very public principles, such as data privacy, inter-operability, non-profit, attribute-based, and opensource. IRMA was first developed experimentally and is currently very basically implemented in Dutch public systems, such as municipalities or health care systems. The alternative ID-tool has proven to work, so citizens no longer have to use Facebook-IDs or Google-IDs when they log in to public services.
But beyond that, what is much more important is that the European legal framework for identification systems called eIDAS, has been adjusted to align with the IRMA principles last year. Therefore, experimenting and developing so called alternative models shows how important it is to show that we can develop platforms according to the outside of mainstream online players, to show how they can be forced to play by different rules. And then subsequently, the European Union or individual national, legal frameworks can adapt those principles to set up condition and rules, for bigger platform companies to comply with. If the development of public or non-profit ID-services is picked up by tech companies, we do not mind as long as they comply with those rules. Yet, the availability of functional alternative services is very important.
I think we can come to the final question. We have talked a lot about platform governance mostly in the liberal democratic contexts. But if we look at societies like China, whose very own platform ecosystem is also ambitiously expanding in the global society, such as North America, Europe, and the global south. If we take that into account, it seems that the whole question becomes even more puzzling. What is the prospect of the Chinese platforms going global or becoming more dominant globally? Would this prospect even change our existing framework of platform governance?
If we look at the geopolitical situation, it’s clear that we have two central systems. Firstly, the Big Five American companies (GAFAM) exploit the American system, and I just explained how it is governed, how data are owned and exploited by these corporations. In that system, there’s a huge potential for corporate surveillance of online activities. Now that system is underpinned by an ideology of libertarian capitalism or surveillance capitalism, as is called.
If we come to China, we see a very important and powerful ecosystem. I call it the BAT system, an acronym for Baidu, Alibaba, and Tencent—tech companies that form the heart of the Chinese ecosystem which operates under state control, of course. Data are owned and controlled by the state but their collection and operation is executed by Chinese companies. This system is underpinned by an ideology of state surveillance and “state capitalism.” However, I think these two systems are much more aligned and interdependent than we tend to think. We see a lot of mutual stakeholderships between the two systems. Apple, for instance, receives 40% of its App store revenue coming from Chinese users. Tencent has stakes in, for instance, Spotify and a number of other companies.
It is not as clear-cut as you think. They are not like two distinct ideological systems in the world. Yes, the Chinese system is becoming more dominant vis-à-vis the American system. And that raises a lot of geopolitical questions, as we see in Russia, which is a very special case right now. But as I just explained, vis-à-vis both systems, I raise these questions of independent institutions and the importance of let us call it a third European way, whose ecosystem will evolve from the principles of decentralized, interoperable, data portable, users in control, opensource that will be designed and structured by independent institutions in control of moderation and governance.
Now, if you have those three systems, the Chinese model, the American model, and the European model, you can see how that European model is hopefully sort of defending its own ideological values in the context of online platform ecosystems that they do not own. Because one major problem is that Europe owns very few of these big infrastructural platforms. I just explored with you this theory of how we can change the existing ecosystem to serve that third model. As long as platform providers are abiding to these principles, and conforming to European laws, we do not mind if the Europeans are sort of dependent on the infrastructures (either Chinese or American).
However, it is not comportable for nation-states to be entirely dependent on foreign companies or other superpowers for all your online traffic and transactions to take place. That is what the Russian situation makes very clear. You can be de-platformized from all online infrastructures in 3 days if you do not abide by certain principles (of the dominant platforms). That is why it can make you vulnerable if you do not own and operate your own infrastructural services. Deplatformization, in other words, is a geopolitical instrument, a political-economic weapon. Several months ago, when Russia started the war against Ukraine, a lot of American platforms had to withdraw as a result of the western economic sanctions against Russia, but Russia also blocked the use of these western platforms, like Twitter, Facebook, and Instagram to “protect their citizens from being exposed to western rhetoric.”
Interestingly, at that moment, the owner of Gab, Andrew Torba, responded to the deplatformization of Russian online community by western platforms, saying “Hey, Russia is now getting the Gab treatment.” He is basically comparing how Russia, as a state, was ostracized, was pushed to the margins or completely cut off from the Internet. And that indeed is what deplatformization means. It becomes a global geopolitical weapon, more or less to protect and preserve or to push and control a platform ecosystem in all its ideological weaponry.
So that is why it is so important to look upon platformization as a mechanism that involves not just the power over one single platform, but the power of running an ecosystem, an online infrastructure as such. So that, in fact, is what the Russia situation teaches us. If you are very dependent on western American platforms, you can be disconnected from the mainstream core of that ecosystem very rapidly. These are all indications that platform ecosystems are not simply commercial mechanisms or state-controlled entities, but they are becoming more and more part of a global geopolitical contest, in which states (governments) and companies are interdependent and interacting to keep control of that infrastructure. It is much like the oil fields in the past and still in the present. We see the same thing happening online in terms of infrastructural resources.
Like you mentioned, Europe does not own any of these infrastructural platforms. Would the approach you mentioned earlier about building these alternative platforms and allowing more civic society actors and institutions to participate be a good solution? Or do you think the European society should also try to nurture its own platform economy to have more dominant players to compete against United States or China?
I do think it has become urgent now that Europe at least takes control over its own infrastructure. Being dependent on Russian gas is simply not such a good idea these days, right? Being dependent on American tech firms for online communication, transaction and data distribution is also not such a good idea. What we’re seeing is indeed there are some European initiatives. They’re not just state initiatives, but also commercial initiatives, for instance in terms of providing cloud services based on European public values and located in data centers based in Europe.
For instance, there is this big project GAIA-X, which tries to provide cloud services that are bound to Europe. It is a public–private partnership for delivering cloud services on the basis of public values, acknowledging that European generated data should remain in Europe and comply with European laws. If you consider data to be an important resource like oil or gas, then it is very important to keep that at your own disposal, on the conditions that you set yourself. So, defining the rules of this infrastructure is very important. Europe is currently taking some measures, especially as a result of these DSA and DMA acts. It is important that we take action and reconsider our dependence on certain big systems. This is a moment of awareness. Of course, we were getting so used to everything being interconnected and everything being dependent on each other. This inter-dependency and co-dependence are now becoming a geopolitical stake. You know, it has always been a geopolitical asset, but now it is becoming a geopolitical liability in many ways—not only in terms of being dependent on data and platforms as infrastructure, similar to oil and gas. The geopolitical situation for the past years, especially since Russia’s invasion of Ukraine, has completely changed. This will be a big issue over the next couple of years, I think.