Skip to content
anders-jilden-150790-unsplash Cropped
08 December 2013

How the infrastructure of the internet disappeared and why we ought to go look for it

Edward Snowden’s revelations indicate a massive breach of trust in internet infrastructure. However, there is one positive aspect to his revelations: they lead to a greater public awareness for questions regarding the net’s infrastructure. Mainstream media have started to question the geographical location of data centres, address encryption practices or to inquire internet exchange points.

We should be thankful for this. And we ought to identify elements that make internet infrastructure visible, so as to conceptualise points of entry for understanding it before it disappears into oblivion again.

Processes of disappearance

Infrastructures can easily disappear from general perception. The screenshot below, taken from Google Trends, shows how the volume of searches for “internet infrastructure” has decreased since the beginning of the recordings in 2004. While this chart has serious flaws from a scientific perspective – no scales, no variables for comparison, undifferentiated anglophone clipping – it inspires further thinking about visibility and loss of interest in infrastructure.

Looking at the physical aspects of internet infrastructure, it is possible to trace a process of absorption.[1] Internet infrastructure has a materiality that ranges from data centres over cables up to transmission facilities. But in order to use the internet today, non-professional users hardly ever have to deal with this materiality – except at the end points of the network. But even there, the infrastructure increasingly deprives users from perceptibility.

In the mid 1990s, the early days of the European wire-based internet, the endpoints could be associated with modem noise, various cables and a fixed location. Today, visibility is reduced to a socket in the wall, few cables and the WiFi router. The latter dissolves the internet into our personal sphere, weakens the restriction of use that comes with locality and lowers the sensory perceptibility of the infrastructure.

When wireless (mobile) access infrastructure developed, invisibility clearly was identified as an advantage from the outset. From a user’s perspective, there was no materiality to the internet infrastructure, unless one lived close to a transmission tower. Wireless internet reached us as a feature in mobile phones. To date, we refer to smartphones, and not to internet devices with phone functionality – as if telephony absorbed the internet and not vice versa.

Wireless internet infrastructure also contributes to a blurring of the endpoints themselves. In metro-zones internet access has become so ubiquitously available that it has become part of the environment: There is no need for an act of connecting. Being connected is the status quo. Even when there is no intention to actively access the internet, apps ensure that resources of the internet infrastructure are constantly being tapped into.

What is left to be observed of the internet infrastructure are apps, says internet researcher Jeanette Hofmann. They settle on the interface to the tangible, the physical. Apps are the cars of the internet. Through applications, users align the infrastructure with purposes. They help mobilise resources to pursue other activities. “[Infrastructure is] part of the background for other kinds of work,” writes the ethnologist Susan Leigh Star (Star 1999, p . 380). She points out that invisibility is a hallmark of effective infrastructures. However, effective only means that the infrastructure works for someone, not necessarily for everyone. “One persons’ infrastructure is anothers’ topic, or difficulty” denotes Star on the relational character of infrastructure (Star 1999, p . 380).

“Visible upon breakdown”

Casually speaking: The one who has got a problem with infrastructure will notice. In Star’s words: “Infrastructure becomes visible upon breakdown” (Star 1999, p 382). As an example for a breakdown in an information infrastructure, Star mentions a server failure.

However, looking at areas where internet infrastructure has developed in a similar manner to Germany, this example does not appear very instructional anymore: Since for most users perceptibility of internet infrastructure is reduced to the interfaces of apps, any breakdown experience is limited to error messages within apps. In other words: What is visible at breakdown is a breakdown, no more; the failure of a delegated action. In infrastructure terms, when dealing with apps, users face empty signifiers.

Unlike when encountering a pothole in a road, internet users can usually neither determine the point of failure of their internet experience nor can they identify who may be held responsible. Error sources may range from the device over the software to processing units on other continents. The infrastructure of the internet remains unrecognisable to those who do not deal with it professionally, such as network engineers.

The task: to make internet infrastructure visible

Science can contribute to making the infrastructure of the internet more visible and show, i.e., “Hidden Levers of Internet Control” as Laura DeNardis titled her paper on governance through infrastructure (DeNardis 2012). But more systemization can be achieved. Broadly varying uses of the term “internet infrastructure”, of which I only list some here, show that. Internet infrastructure can be approached from different angles:

  • physical components from cables to the spectrum: “physical layer” (Benkler 2000, p. 3);
  • entities that are involved in providing or controlling it: “organizations and institutions” (Kritische Infrastrukturen, n.d.) and actors “designers, developers, users, administrators” (Bowker, Baker, Millerand & Ribes, 2010, p. 98);
  • technologies that are being used: “software systems like the Domain Name System (DNS)” (Critical Internet Infrastructure 2013);
  • technical architectures and related design principles that underly the technologies: “logical layer”, “software and standards” (Benkler 2000, p. 3), ” Protocol Politics” (DeNardis 2009), layering and best-effort principle (see VanSchewick 2010);
  • the purpose or group of users, “for public use” (Frischman 2007, p. 3), “regardless of the type of use or identity of the user” (Frischman 2007, p. 1), “enables knowledge work” (Bowker et al ., 2010, p. 98), “common facility” that “must be free of entry barriers to effective communication” (Benkler 2000, pp. 10 & 14 on “core infrastructure”);
  • the resources that it allows to activate:  “content layer” (Benkler 2000, p 3); “pervasive enabling resources in network form” (Bowker et al., 2010, p. 98); “supply systems of our society” (Kritische Infrastrukturen, n.d.)
  • a mixed form: “interrelated social, organisational, and technical components or systems” (Bowker et al, 2010, p. 99).

Analysing the many approaches and tracing their origins is beyond the scope of this article. The list is by no means complete. It does, however, substantiates two things: 1. Internet infrastructure is probably referred to more often than the term is explicitly put to use. 2. It suggests that wherever only a singular aspect of infrastructure as being addressed, any critical assessment should be accompanied by questioning the context and inquiring about possible interdependencies. This is what makes thinking about infrastructure interesting.

Detect interplays

This awareness ought to be rooted in net policy: that, for example, networking standards may fundamentally influence options of use. That the sociality organised around an internet exchange point may contribute to where peerings between network actors come about and thus, whose data has to travel long or short routes. That the suggestion to introduce something like a “Schengen” or “German routing” may lever influence for new players, or at least whet the appetite. Where national boundaries are introduced into the internet for one purpose, they may as well be activated for other purposes. In conclusion, interventions that target one aspect of infrastructure can have side effects elsewhere.

Due to the revelations by Edward Snowden, we have not experienced a breakdown of internet infrastructure in the way Susan Leigh Star has described it. The internet experience of the vast user base has apparently not directly been altered by the spying, storage and analysis activities at the junction points of the internet infrastructure. On top of that, breakdowns cannot reasonably be attributed to sources in an internet that is not only controlled in a distributed way but also ever more perceived through apps. However, such activities increase the likelihood of breakdowns in other spheres of our democracies, for example, by provoking chilling effects that prevent us from exercising the right to free expression. We ought to oppose this. A prerequisite is to pay more attention to the infrastructure of the internet and its many facets.

1. I am referring to Germany here, but the same may hold true for other areas where internet technology has been adopted in similar phases. Another story may have to be told for areas where internet access has developed differently, like in Sub-Sahara Africa where mobile internet always has been more important for the general public than fixed broadband.


This post is part of a weekly series of articles by doctoral candidates of the Alexander von Humboldt Institute for Internet and Society. 

This post represents the view of the author and does not necessarily represent the view of the institute itself. For more information about the topics of these articles and associated research projects, please contact

Uta Meier-Hahn

Former Associated Researcher: The evolving digital society

Sign up for HIIG's Monthly Digest

and receive our latest blog articles.

Man sieht in Leuchtschrift das Wort "Ethical"

Digital Ethics

Whether civil society, politics or science – everyone seems to agree that the New Twenties will be characterised by digitalisation. But what about the tension of digital ethics? How do we create a digital transformation involving society as a whole, including people who either do not have the financial means or the necessary know-how to benefit from digitalisation?  And what do these comprehensive changes in our actions mean for democracy? In this dossier we want to address these questions and offer food for thought on how we can use digitalisation for the common good.

Discover all 12 articles

Further articles

Man sieht mehrer Spiegel, die in unterschiedlichen Formen angeordnet sind und verschiedene Oberflächen, wie den Himmel, eine Hauswand und so weiter widerspiegeln. Das Bild steht sinnbildlich für die vielen verschiedenen Bedeutungen von autonomen Systemen in unserer Gesellschaft. You see several mirrors arranged in different shapes reflecting different surfaces, such as the sky, a house wall and so on. The image is emblematic of the many different meanings of autonomous machines in our society.

The age of machine autonomy?

Can machines be autonomous – or is it a human prerogative? This categorical question dominates many discussions on our relationship to purportedly intelligent machines. A human vs. machine rhetoric, however,...

remote work is moving towards the city

The Anywhere Jobs are nowhere near – How remote work is moving towards the city

Remote working allows us to work from "anywhere". So why are cities, of all places, becoming the new mega-hubs for digital work? What does this change bring to rural regions...

Content moderation on digital platforms: A more intensive horizontal effect of freedom of expression?

To what extent should private companies be bound by freedom of expression? Investigating a more intensive horizontal effect of fundamental rights.