This extract is from Four Internets: Data, Geopolitics, and the Governance of Cyberspace, by Kieron O'Hara & Wendy Hall (published in the United Kingdom by Oxford University Press, August 2021), and draws from the preface and part II of the book. Aiming to foster the fruitful exchange of expertise and perspectives across fields to help us rise to this critical challenge, opinions expressed do not necessarily represent the views of the OECD.
The Internet appears a fixed part of modern reality. Its roots date back over 50 years, and by the first quarter of the 21st century has become the pre-eminent means of distributing information. To those in their early twenties, it will have been a permanent, if mutable, background to their lives. It seems unquestionable, like electricity and roads. Its design was intended to make it tolerant of disruption and faults, and it is pretty robust and resilient. And it is surely evident by now (our book was completed in lockdown during the 2020 COVID-19 pandemic) that it has become a piece of critical infrastructure, and needs to be protected as such.
Yet it is surprisingly delicate all the same, a gossamer arrangement of different types of hardware, protocols to describe how systems communicate with each other, and methods of social coordination, ranging from regulations to contracts to what amount to little more than gentlemen’s (and ladies’) agreements. The Internet is not a monolithic technological creation patented and run by a single company or government, but a congeries of systems, protocols, standards, hardware and organisations. Some of these organisations have national standing, some have global reach, and others have international standing. Some are public bodies, some private companies and some non-profit organisations. The Internet is one of the few institutions where the people round the table are as likely to be in corporate suits and Christian Louboutins, as black t-shirts with hoodies and Air Monarchs, or even (still, just about) tie-dyed denim and bare feet.
This system is furthermore truly sociotechnical – we cannot hive the technology off from the people who use it in their everyday lives. Every design decision reflects, and imposes (perhaps unconsciously), a balance of power, while cultural, economic and political tensions play out across collective-action problems. Neither computer science nor the social sciences are individually sufficient to encompass all the study required to understand the most complex piece of technology ever created, the structure of which is driven by the people who upload, download and link content. We have long argued that concentrated interdisciplinary research, encompassing social and technical studies, is required both to understand it and to engineer it.
The future of the Internet depends on answers to many questions. Most obviously, the COVID-19 pandemic has completely reshaped our relationships with the Internet, and with the wider world. What will the new normal be?
There are political questions, of which the elephant in the room during the time of writing was the outcome of the 2020 US Presidential Election. President Trump’s success or failure in 2020 would have massive influence over the Internet’s future, and that of the technology industry generally. President Biden will certainly govern in a different style, but there may well be aspects of his predecessor’s agenda that he preserves. Other unresolved issues concern the extent to which China will challenge the US for global leadership (it currently heads 4 of the UN’s 15 specialised agencies, including the International Telecommunication Union), and how the EU will work through Brexit, and the post-Merkel era.
The Internet is also co-created by its users, and every year it adds in excess of a hundred million new co-creators. These people will change the Internet in unpredictable ways. Our study of India shows some of the different directions these changes may take.
Finally, there is the technology. The Internet creates data, the fuel for Artificial Intelligence and smart cities, which many see as the future for humankind. Will they produce utopia, or dystopia, or are they merely hype? Other important technological trends include blockchain, automatic face recognition, virtual or augmented reality, undetectable deepfake technology, quantum computing, ubiquitous wireless broadband and medical wearables. Any one of these could have a massive effect on our privacy, public life and wellbeing.
Our research, as we describe in the first chapter of our book, is structured by three basic ideas about the Internet as a network, the Internet as a producer of data, and the Internet as the key technology for digital modernity.
- The different networks connected by Internet protocols are governed by subtly different principles, which may, over time, make it harder to get data across the Internet as a whole, and therefore may increase the pressures to fragment.
- These different principles have a strong connection with attitudes toward data, which is the chief source of power on the Internet.
- The visions and models of the Internet we discuss in our book are all, in their way, responses to the prime imperatives of digital modernity.
We hope to weave the narratives of networks, data and modernity into an understanding of how different cultures and ideologies see value in the Internet, and how it might be possible to keep it together, while respecting difference and diversity. The key, we claim, is how the Internet is governed.
Our book is divided into three parts. In the opening part, we introduce the question of Internet governance, and how it reflects values.
We set out the original Internet vision, the ultimate techie ambition of an open, unmanaged, free information space. This Silicon Valley Open Internet is an ideal, rather than a reality, but it has both political and engineering arguments in its favour. However, not everyone supports the political ideal and very many people have been concerned about the unintended consequences of the unimpeded flow of information. Opposition to the Silicon Valley Open Internet has crystallised around alternative views of how the Internet should function, in terms both of serving its users and of its effects on wider society.
In the second part, we focus on three alternatives that have become highly influential; these, plus the original open vision, give us the four Internets of the book’s title. These aren’t the only visions that are available (hence our title is Four Internets, not The Four Internets – a definite article would be misleading). Other Internets could be imagined, if the appropriate ethical vision found a technological realisation and sufficiently powerful institutional backing.
We also look at another ideological model of how the Internet can be used as an instrument of policy – not one that comes with its own positive vision, but a spoiler that might undermine the integrity of the Internet and thereby (perhaps accidentally) contribute to fragmentation. It is not a vision for the Internet, but a model for its exploitation.
Figure: Four visions for the Internet and a spoiler
Each of the models is named after a geopolitical entity which has devoted the most effort and resource to implement the model. The Silicon Valley Open Internet is an example – it references an ideal, the openness of the Internet, and associates it with a place, Silicon Valley, where it happened to originate and where much of its support resides. The other models will also be associated with places – Brussels, Washington DC, Beijing and Moscow – where the most powerful geopolitical support for the model can be found.
These geopolitical labels are suggestive enough to give an immediate sense of how the model is intended to operate. This does not mean that that we subscribe to any kind of cultural or technological determinism, so that we think the future of the Internet, or of any corner of it, can be predicted with any degree of accuracy from cultural, geopolitical or technological facts (or vice versa). In fact, we think the exact opposite. The key desideratum of governance policy and strategy is to be prepared, and we believe that our particular narrative is helpful in thinking of squalls ahead.
Read more on the Forum Network: "Technology, Privacy and Market Power: What policy makers must do next" by Anthony Gooch, Director, Public Affairs & Communications Directorate, OECD
Neither do we want to say that governments dictate everything. Internet governance is also affected by companies, supranational organisations, and so on. Not every aspect of the Chinese Internet is dictated by its government – we also need to look to Tencent, Ant Financial, Huawei and, not least, the Chinese people. Indeed, not everything in the Chinese Internet conforms to our ‘Beijing’ model, and conversely other nations, including democracies as will become clear, are attracted to the Beijing model in at least some contexts, and will employ it in their own tailored way. And even when we do focus on government, we should not make the elementary mistake of assuming that governments are single actors with an indivisible corporate view. Within any particular government, the finance, health, defence, interior, telecommunications and education ministries may all adopt wildly differing views about how the Internet should best be governed and how it might serve their own purposes. As Internet policy experts Ian Brown and Christopher Marsden argue, “It is not possible to map Internet regulation as a patchwork of national networks where international regulatory discussion centers on areas with overlapping jurisdictions or unclear jurisdiction. This comparison of Internet regulation with the Law of the Sea or medieval mercantile law … is untenable in practice.”
In short, when reading Four Internets, don’t focus on the geopolitical exemplar, but on the adjective: open, bourgeois, commercial, paternal. These adjectival aims can be adopted by anyone, at any time, permanently or temporarily, for principled reasons or opportunistic ones.
The labels are labels, suggestive but not destiny. Other models are possible, and indeed exist, and other narratives than ours could no doubt be generated. We have attempted to construct abstract, hypothetical visions, stressing some aspects of ideological engagement with the Internet, with the aim of interpreting actions and policies and making sense of debates. Think of them as Weberian ideal types (perhaps even World Wide Weberian ideal types), placing (an) order on the complexity of experience. We believe that these Four Internets are a useful clustering of attitudes towards the Internet, indicative of likely future directions, controversies, tensions and risks.
The third part of our book sets out some of the implications of the ideological centrifugal force. This isn’t a Jeremiad. There are radically different views of how communications should be managed, and these need to be reflected and respected in governance structures – even if engineering imperatives are compromised as a result. But it seems clear that the more closely the diversity of Internet governance tracks the diversity of the world it spans, the easier it will be to contain this force.
Find out more about Four Internets: Data, Geopolitics, and the Governance of Cyberspace, by Kieron O'Hara & Wendy Hall (published in the United Kingdom by Oxford University Press, August 2021)
 Brown & Marsden 2013, 4-5.
 Berners-Lee et al. 2006, O’Hara et al. 2013 for the interdisciplinary study of web-like technology-mediated networks from an engineering perspective, which has been called Web Science. Similar important and valuable research from a sociological perspective has gone under the umbrella title of Internet Studies: Dutton 2013, Graham & Dutton 2014.
And check out also the OECD Going Digital Toolkit which helps countries assess their state of digital development and formulate policies in response: