When Lavrentiy Beria, the former head of Stalin’s secret police, fell out of favour with the Soviet government and was executed, the authorities proceeded to expunge his achievements and very existence from the Great Soviet Encyclopedia. The publishers created four extra pages of the closest alphabetical entry – the Bering Sea – and sent them to every one of the encyclopedia’s subscribers so they could replace the entry for Beria. For people who lived in the Soviet Union, compliance with this kind of request wasn’t optional: being in possession of banned or unrevised texts implied criticism of the regime and carried with it a range of unpleasant consequences.
This story is told in David King’s The Commissar Vanishes, and I find it fascinating for a number of reasons. There is the dramatic irony of Beria, who had built a career inflicting this kind of treatment on others, receiving a dose of his own poison. There is the comical detail of him being replaced with extra information on the Bering Sea, making the episode less Orwellian and more like something coming out of Lewis Carroll. Above all, there is the image of readers being forced to carry out the censorship themselves, one by one, using scissors and glue. Nothing could be more distant from the present moment and everything we know about how digital information lives and dies.
This essay is about ownership of the digital means of producing, distributing and consuming critical literature. The observations apply primarily to Western liberal democracies in the present day – that is to say, those that are firmly in the age of the internet. But I would like you to keep Beria’s erasure in mind as you read, for it may yet provide us with a useful point of comparison.
Who owns the internet?
The internet belongs to nobody – in that it belongs to everyone – for at its most basic level it’s nothing more than a series of publicly administered protocols. The main governance organisation is the Internet Corporation for Assigned Names and Numbers (ICANN), a California-based non-profit. ICANN does receive funding from the US Department of Commerce, but is run at arm’s length from the US government (most other national organisations follow the same model). Various attempts to bring the internet under UN control, a move backed by authoritarian states such as Russia and Saudi Arabia, have been non-starters, leaving disgruntled states with few options – no nation that wishes to compete economically can afford to go off the grid. Thus the network of networks has achieved near absolute penetration, bringing with it an unprecedented capacity to disseminate information.
When the digital version of this essay goes online, it will be instantly available to a potential audience of 2.94 billion, including an estimated 800 million English speakers. There is no precedent in human history for this fact, and the implications for sharing political materials are far reaching. Consider the production costs of traditional publishing, as well as the logistical and financial burdens associated with disseminating texts across borders. Consider also the relative ease with which print texts can be banned, confiscated and/or destroyed, as well as how dangerous and difficult it once was to conceal compromising literature. Almost all of these problems have been eliminated, only to be replaced (even in relatively permissive countries) by new problems – often more subtle and intractable ones.
Take, for example, the current publication. Overland pays contributors, leases its own servers, and retains a high level of control over its print and digital content, all of which is eventually archived in a number of physical and electronic repositories. Its editors are publicly known and its authors are named, except in very rare circumstances when it’s necessary to protect their privacy. In most respects, then, Overland is a traditional publication, one that enjoys institutional backing and public financial support. These are ideal conditions – enviable, in fact – in which to produce political commentary and critique.
The reality for individuals and less established outlets is likely quite different, especially for those on the more radical end of the spectrum. There are, of course, a number of political publications – from Jacobin and Socialist Worker down to local newsletters – that maintain a physical presence. For most groups, though, online publishing is the cheapest and easiest way to share content. The entry costs for those looking to take out a domain name and have their writings privately hosted (using free publishing platforms such as WordPress) are very small. All of the material, just like this essay, can be made available instantly to an almost incomprehensibly vast audience.
In practice, however, the most significant distribution network for political texts is social media, chiefly Facebook and Twitter. These sites are increasingly acting as valuable publishing platforms in their own right (this is especially true of Facebook, which prolific authors, such as British Marxist commentator Kevin Ovenden, are using effectively as a blog). A vast amount of political debate takes place on these social media spaces – platforms that are, we shouldn’t forget, owned by US corporations and subject to US laws, even as the information travels on a series of public and state-controlled networks. Even Overland depends on them to a significant degree for promotion and delivery.
Then there is Google, which along with Facebook holds a monopoly on the advertising market – thus accruing an ever-growing influence over commercial media – and which, at least outside of China, serves as the de facto global directory of web content.
Search engines and social media platforms are in every way as significant as the internet protocol suite itself, for they are the true content delivery systems of the Web; it’s through these platforms that we are able to access and share information. By far the most popular systems of this kind are in wholly private hands.
Twitter for the people
The internet is a global system of computer networks. Its origins date back to the 1960s, when the US military supported efforts to create a networked communication system, but it wasn’t until the 1970s and 80s that it began to take today’s shape. What distinguishes the internet from other computer networks is the use of specific common protocols. The suite includes the Transmission Control Protocol (TCP) and the Internet Protocol (IP), which together enable the exchange of information by specifying how data is addressed, transmitted and received. Most laypeople will have also heard of the Hypertext Transfer Protocol (HTTP), which governs connections between documents on the Web, and the Simple Mail Transfer Protocol (SMTP), which is the standard for electronic mail.
Nowadays, the internet has become synonymous with the Web, which in the simplest of terms is the portion of the internet we access with our browsers, navigating from one hyperlink to the next or searching for documents through a search engine such as Google. Nowadays even email is largely accessed via web clients (think Gmail or Hotmail). But it wasn’t always so. When I first ventured online, in the mid to late 1990s, the dominant social network, Usenet, was part of the internet but not the Web.
Usenet was a discussion network first developed in 1979. Originally hosted at the University of North Carolina and Duke, the network quickly grew in popularity, spreading across the globe. It had its own communications protocol, akin to email but with an important difference: the origin of each message was known only to the server that first received it. In other words, the origin of the content was not known to the rest of the network. Usenet was, in fact, a wholly decentralised network, one that relied on nodes run by universities and private organisations and, later, internet service providers (ISPs). Discussions were divided by topic, such as alt.politics.socialism.trotsky (self-explanatory) or rec.arts.sf.written (written works of science fiction), and were mostly unmoderated. Users accessed the network through ad hoc clients, most of which were free and open source. Its culture developed alongside that of bulletin boards – a communications system that was supplanted by the internet in the 1990s – and its vernacular would be familiar to users of today’s social networks. It was on Usenet and bulletin boards that the first trolls dwelled, and where furious debate descended into ‘flame wars’. It’s to Usenet that we owe the word ‘spam’, as well as our methods for dealing with it – Usenet saw the first attempts to block malicious and unwanted messages via moderation at the server’s end and filtering at the client’s end.
It’s difficult to make a reliable estimate of how many people were using the network when the Web first appeared, but it was a significant public – relative to overall internet user numbers, it’s possibly comparable to the people who use social media today. But unlike social media, nobody made a dime from Usenet. By contrast, Facebook declared a turnover of US$17.93 billion in 2015.
Usenet did not require search engines: users browsed the directory of topics, or relied on other users to direct them to groups that matched their interests. When I began my PhD in 2002 – well and truly after Google had become established – I enlisted the help of the two literature groups I frequented to scope out my topic. Their help was invaluable.
As well as providing access to groups of like-minded – or, if desired, unlike-minded – individuals, Usenet allowed users to filter information in increasingly sophisticated ways. The ‘score files’popularised by clients such as the freeware newsreader Xnews, worked by assigning a positive or negative score to threads, topics or other users. In this way, one could curate a group’s often chaotic timeline by highlighting the contributions of certain participants, or by hiding others. Consider, by comparison, the nightmare of trying to get Facebook to display the latest status updates first – a setting that is no sooner entered by the user than it’s ignored by the system.
Flame wars and spam are often cited as the reason for Usenet’s decline, although others are more inclined to blame the rise of personal pages and blogs. Either way, its fall in popularity was so rapid that, by around 2010, it had been abandoned by most major American ISPs and by the universities that had originally supported it. For them, the cost had begun to outweigh the benefits. Its decline was accelerated by an anti-child pornography campaign aggressively pushed by in 2008 by then Attorney General Mario Cuomo in the state of New York. This led to some ISPs blocking the whole Usenet hierarchy instead of just to the alt.binaries section where such material could be found.
While Usenet can still be accessed and contributed to in the traditional way, it now survives largely on the Web, in the form of a service offered by – yes, you guessed it – Google.
Network and surveillance
The transition from Usenet to today’s dominant social networks is a move from a public, decentralised model to a corporate-owned, centralised one. It’s important to understand that architecture and ownership go hand in hand: it’s easier to monetise services delivered through a centralised model, where all the information passes through and is stored on an organisation’s servers, than through a distributed one, where no single entity controls the information, much less a peer-to-peer model that relies on the machines of individual users for both storage and distribution. It’s also far easier to surveil, censor and control a centralised distribution model. The enclosure of the digital commons therefore has material repercussions on the freedom to speak (safely) and, by extension, the right to dissent.
Suppose you live in Australia and run a blog that is critical of your government and its allies. You have purchased a domain name, and pay a small hosting fee to a domestic ISP. The very existence of the blog means it will be subject to indexing by all manner of directories and search engines, including Google. This is, after all, how you get a significant amount of traffic. If the blog is linked to a Twitter or Facebook account – another key driver of traffic – it will be subject to topic-based analyses to determine the interests of your readers. I hesitate to call these processes ‘surveillance’ because you have chosen to make this material public in your own name. Nonetheless, your output is being monitored by corporations and governments (both your own and others). The latter is sometimes referred to as ‘open-source intelligence’ – that is, intelligence gathered from publicly available sources.
The next time you travel to the United States, you’re detained at the airport. The customs officers don’t have to tell you why you’re being questioned. Maybe it is a random check, but you wonder if it has something to do with your little blog. As part of the questioning, you may be required to surrender the passwords to your social media accounts. Thanks to Edward Snowden’s revelations about the Five Eyes program, you know your private communications are subject to wholesale interception and analysis. This time we can just go ahead and call it surveillance, for now it’s private material that is being used to build a profile of you. As the public and the private sphere come together, you begin to wonder to what extent you have allowed yourself to become exposed. If you collaborate on political actions with other activists, you may have unwittingly made them vulnerable as well.
You write on your blog about a political meeting without disclosing its location. But you’re using your real name and it’s easy enough to link your identity to your mobile phone. Because you mentioned the time when the meeting took place, it’s possible to derive its location by analysing your phone data, and, based on data from the same location at the specified time, it’s possible to figure out who else attended the meeting. Australian authorities may not be interested in that information just yet, but the capability is there.
To be clear, this hypothetical scenario is not entirely due to communication tools being privately owned, but there is an important convergence. For instance, the ‘location services’ that collect the metadata so helpful to surveillance have the primary purpose of delivering highly profitable advertising. The identity profiles built about you for the purposes of marketing reside on servers in another jurisdiction with its own privacy laws, and you have no control over this information. So does the content you have posted on any of these services, even if you have since chosen to delete it. This lack of control extends to your ability to filter information you don’t want to see, as exemplified by the move away from chronological timelines on Facebook, Instagram, Twitter and other social media platforms. Instead, the presentation of information is dictated by the companies’ algorithms – a move designed to make the services more profitable.
The architecture and evolving nature of the Web are such that these risks are very difficult to evaluate, especially in the early stages. Users of a service have little sense of how information might be stored and accessed in future. In fact, one of the criticisms that Google came under when it built its Usenet archive – the first decade of which came in the form of a donation by a Canadian university – is that it allows access to material in a way not intended by the original users. It also opens up the possibility of building user profiles – which just so happens to be Google’s principal business.
Surveillance practices are not limited to governments. Prospective employers can analyse your social media profile and online publishing output in order to form a picture of your beliefs, interests and affiliations. There are businesses that offer employers this very service, just as there are businesses that investigate prospective tenants for landlords.
So now your formal published output – the things you made a conscious decision to put out there in your own name – combined with your far less tightly managed social media postings and the vast map of your personal contacts and connections, pieced together by myriad algorithms and sold for profit, can be used to determine whether you will find a job or a place to live.
What is a political safe space?
When it comes to surveillance, it’s control of the means of distribution, rather than the means of production, that matters. Recall the expunging of Beria from the Great Soviet Encyclopedia and how readers were required to actively collaborate in the act of political revision. Some would have done so willingly, to ensure the integrity of their copy of the text, while others would have done so out of fear of the consequences of non-compliance. Some didn’t do it at all, which is how copies of the original entry have survived.
Practically speaking, it’s much easier to revise an online encyclopedia, for one can just amend the single digital copy that everyone accesses. At the same time, it’s vastly harder, for doing so calls attention to the act of revision. This is what is known as the Streisand effect, named after Barbra Streisand’s attempt to have photographs of her home removed from the internet, a move that led to them being shared far more widely than they ever would have been otherwise. Thus, it’s not the tyranny of officially sanctioned texts we have to guard against, but the permanence of texts that can speak about us in ways we didn’t anticipate and that we have no power to restrict.
As remarkable a technology for communication as the internet is, it’s even more remarkable as a technology for control. This is what Evgeny Morozov might have meant when he quipped some years ago that ‘cyberspace politics is a zero-sum game’: the freedom the internet has given us to communicate, publish and organise is matched – if not overshadowed by – the unprecedented capacity it has given corporations and governments to monitor such activities.
When I discussed these issues with Andrew Cushen of InternetNZ, the equivalent local body to ICANN, he told me he believes this control is so pervasive as to make revolution impossible. It’s a claim that gave me pause for thought, all the more in light of InternetNZ’s commendable efforts to oppose mass surveillance (most notably through its critique of the current government’s failure to accurately define and limit its domestic intelligence powers). Perhaps Cushen is right. And this only reinforces my belief that the most urgent task for political groups and movements – revolutionary or otherwise – is to think strategically about what a safe political space is and what they can do to protect themselves (while continuing to be outward-looking and connected).
This conversation will resemble others that militants and activists have been having for generations, only within a new paradigm in which traditional environmental surveillance is no longer the primary threat. Some may choose to do very little, other than be mindful of not being as cavalier with the privacy of other members as they are with their own. Others might opt to build codified rules around personal use of social media versus other public interactions their members put their name to. Others still may choose to educate themselves about tools that enable anonymous, unmonitored browsing, such as Tor (the Onion Router), and to encrypt all email communications within the group.
All, I think, could benefit from carrying out the following thought experiment: imagine your country goes authoritarian tomorrow – not because you think it might happen, but because of what you might discover in doing so. International case studies abound – from Egypt and Ethiopia to China and Turkey – so it’s easy enough to build a fairly accurate technical picture of what a crackdown entails. Social networks that can be switched off by blocking the corresponding domain, such as Twitter or Facebook, will be the first to go, while the information they contain will be used to round up individuals.
Now ask: how would you communicate? What decisions will you make to protect the group? Will you wish you had been more careful in your use of information, or admonish yourself for not using decentralised communication tools along with the popular commercial ones? If the answer is yes, consider revising your current practices and tactics accordingly. This may mean advocating for some aspect of internet reform, adopting new distribution protocols or connecting with civil liberties groups active in this area.
There is, of course, no absolute safety, because there never was: the history of radical politics is also the history of its means of repression. It’s just a question of updating our thinking, perhaps dramatically. In so doing, we must be mindful that the medium is not just the message – it informs our habits of thought as well. Thus, educating ourselves about electronic surveillance can sometimes push us towards the politically catastrophic rocks of conspiracist thinking. Steer a course away from those, and keep safe.
This essay has benefited from conversations with Andrew Cushen, Thomas Beagle, Brenda Wallace and Dougal McNeill.
Artwork by Brent Stegeman.
If you liked this essay, buy the issue
Or subscribe and receive
four outstanding issues for a year