The Regionalization of Cyberspace: Regional Internet Registries and their Impact on Internet Governance

My personal entry point to our ethnographic research at project B02 “Control/Space” was a meeting of the Réseaux IP Européens (RIPE, French for European IP Networks) in Berlin in May 2022. We went there to gain an initial insight into the current discourses on Internet governance. Around that time, our research group stumbled upon an interesting and thought-provoking map that seemed to be in clear contradiction to how internet maps usually appear (see Figure 1). Most internet maps usually represent the network character of the cyberspace and often consist of nodes and connections. However, this map shows neither cables nor networks nor stakeholders, but only these slightly unusual world regions, consisting of national territories. This map was basically the impetus for our interest in the Regional Internet Registries (RIRs). The specific and unique regional division of the maps compelled us to understand how these regions historically developed, how the lines between them were drawn, and what we can conclude from this regarding the spatiality of digital infrastructures. A comparison of the different regions could also tell us something about the varieties of Internet governance in terms of their multiple spatialities.

I will first discuss the general function of Internet registries, then, secondly, place them in the context of a multistakeholder governance discourse. Third, I will refer to the beginnings of transnational IP address allocation and use excerpts from interviews we conducted with the founding members of RIPE and Asia Pacific Network Information Center (APNIC) to illustrate the different logics of regionalization regarding the development of those two Internet registries.

The RIPE Network Coordination Center was founded in 1989, as the first of five Regional Internet Registries (RIRs) responsible for the support of the Internet in a specific service region. The most prominent responsibility of an RIR is to provide internet resources and related services (e.g., IPv4, IPv6, AS Number resources)to members in their service region. Resources and services are mostly distributed to internet providers, also known as Local Internet Registries (LIRs) (e.g., Telekom, AT&T or Orange). The RIRs play a fundamental role in the field of multistakeholder Internet governance, which covers a complex field of organizations, responsibilities, and technical expertise¹.

If we think about the general function of Internet registries in relation to space, we can imagine them as public registration and allocation entities of the nodes of a networked space, which are a fundamental basis of the Internet’s characteristics of connectivity. Every device connected to the Internet needs an IP address to make communication possible in the first place. Only by publishing IP addresses can data be routed through the networks and exchanged between different servers. So, historically speaking, the network space didn’t just spread through cables and protocol standards. Rather, network space was spread through the organizational work of certain social actors, who followed certain logics of network spatialization. In the case of the registries, network spatialization quite evidently led to a transnational regionalization of Internet governance.

Internet governance as multistakeholder discourse

To approach the specific question of regionalization, we must first talk about Internet governance in a broader sense as the RIRs are embedded into a so-called organizational ecosystem². In 2005, the United Nations General Assembly agreed to organize theWorld Summit on the Information Society from a multistakeholder perspective³. Since then, many international and multilateral organizations have publicly endorsed the multistakeholder approach as the main way to proceed with Internet governance. Organizations like the Internet Corporation for Assigned Names and Numbers (ICANN), the Internet Engineering Task Force (IETF), and the Internet Governance Forum (IGF) should be mentioned here, which bundle different tasks, responsibilities, and specific expertise in an organizational manner.

The discursive arena of Internet governance is further shaped by a plurality of annual conferences, which are organized by the different Internet governance organizations. Thus, these organizations not only function as transnational authorities with regard to certain responsibilities, but also as public and discursive arenas where the latest political and technical developments of the Internet are openly discussed. At these conferences, the different stakeholder groups communicate certain issues and problems, promote new standards and technical protocols, and debate the past and future developments of the Internet. For our subproject, these conferences were of great relevance to understand the dominant discoursein the field, and to identify relevant speaker positions, ideas, legitimizations, and narratives regarding internet infrastructure⁴. What became apparent during our ethnographic research were the specific political cultures of these organizations, which are heavily influenced by egalitarian ideasand a kind of “Habermasian” discourse ethicoriginally located in the academic lifeworlds of the USA and Western Europe.

Jon Postel and the beginnings of IANA

The specific regionalization of the Internet often begins with the naming of a single person: Jon Postel. As the “First Steward” of internet numbers and primary administrator of the Internet, Postel was an academic sitting in California in the 80s with a physical (!) notebook, distributing and administering the IP addresses of the whole world. The distribution of IP addresses was therefore organized as centrally as possible at the time: If you wanted to connect your computers to the Internet back then, you had to contact Postel personally. He then assigned you an address range and published it so that your devices could be reached. The question of why this has changed, and how Postel’s personal task has been transformed into a regionalized delivery of IP addresses, is not only connected to a global digitalization process in general, but also to the question of its political control. Internet governance at the time was primarily determined by players from the USA. The progressive commercialization of the Internet thus created a legitimate pressure to decentralize the distribution of IP addresses. The Internet and its corresponding network logic, so to speak, couldn’t expand and sustain itself without a change in the distribution of political power.

This in an excerpt of RFC⁵ 1366, written by Elise Gerich⁶ and other early pioneers of the global registry system:

This discourse excerpt is illuminating to us for several reasons, as it indicates certain spatial and political problems of the IANA function at the time. First, the excerpt includes a clear call for a regional distributive organization due to the perception of a more global and diverse population. Reference is also made to linguistic differences and local customs that justify a decentralized organization. Finally, a selection process is mentioned, according to which certain existing organizations are to be entrusted with the task of distributing the IP addresses.

The RIPE NCC office was already located in Amsterdam at the time, close to Western European research networks and embedded in an English-speaking community. Here, a group of IP network operators held regular meetings and shared their experiences regarding the technical operation of the Internet. The first Internet Registry was thus to emerge from a pre-existing local network of technically qualified people, which then took over the distribution of IP addresses over the next few years and decisively shaped the political culture of Internet governance. To this day, Amsterdam and Western Europe form the core of this Internet region, both in terms of personnel and organizational power. And the RIPE network ultimately became a blueprint for future Internet registries.

Varieties of regionalization: RIPE and APNIC

The asynchronous historical development of the RIRs is very significant, as it also indicates certain hierarchies among RIRs regarding the cultural backgrounds of the people involved. The RIRs created earlier could develop certain organizational best practices, which the later registries could easily implement and take over, using pre-existing strategies and blueprints for community building.

APNIC was founded shortly after RIPE, in 1993, then followed by ARIN in 1997. For five years, between 1997 and 2002, the IANA function was organized across three regions. At that time, North Africa was part of the RIPE Region, the southern parts of Africa were part of the ARIN Region and Madagascar was part of the APNIC region (see Figure 2).

Figure 2: Regional Internet Registries before the creation of LACNIC and AFRINIC (Wikipedia, CC BY-SA 3.0)

During the interviews we conducted, we were able to speak to some of those involved in RIPE who had been there from the start. When asked about the specific assignments of the respective countries to the registrars, one of the people involved at the time replied:

Yes, that has also grown historically … the first ones were already Europeans, the European science networks, but then it was also personal contacts, so the first chair of RIPE…had many contacts in the Soviet Union at the time and was also very committed and also said we had to involve people and then it turned out that way. But I think that was mainly because the RIPE NCC already existed. They tended to focus more on the Middle East and so on, which is also connected to it and Central Asia, and that it was easier or that they trusted us more as a neutral organization than to agree on who should do it. And that’s why it turned out that way. In the beginning, we also divided up half of Africa, Africa was divided up [pause], so it was really a typical colonial thing.

RIPE as a region develops in the early days out of a specific local network of technically qualified people and certain infrastructural prerequisites. From there, the local network scales up to a transnational region in terms of political responsibilities. One key factor in legitimizing this upscaling of the local network into the region is its perceived neutrality, which allowed for the delegation of power to this regional center in Western Europe.

In contrast to RIPE and its specific regionalization, the beginnings of APNIC can be traced back to Japan and Australia. As countries with a pronounced digital development at an early stage, social networks of Internet governance also developed here early on, from which the regionalization of the registry could be organized. To this day, the offices of the APNIC organization are located in Brisbane, Australia.

A striking feature of the APNIC region are the seven National Internet Registries (NIRs) in the Asia-Pacific region, located in China, India, Japan, Indonesia, Taiwan, South Korea and Vietnam (see Figure 3). While neither RIPE, ARIN nor AFRINIC have National Internet Registries, within APNIC the distribution of IP addresses has been reassigned to national organizational units, which are still part of APNIC’s organizational structure.⁸

Figure 3: The National Internet Registries in the APNIC region (APNIC 2024)

We could interpret this as a kind of national territorialization of the registry function, as the APNIC region may seem to be too politically diverse to fit under the sole umbrella of APNIC. But although the NIRs refer to a nationalization of address allocation, they are still integrated into the transnational APNIC organization, and act in line with APNIC policies.

We were able to interview the Director General of APNIC last year, asking him about APNIC’s current challenges with regard to the political diversity in the region:

We really have not had…separatist kind of movements at down to a … subregional level. And maybe it’s the diversity of the region that helps that, for instance, if there was a move and there have been talks in the past about, say, a South Asian RIR or an Asian RIR, but if you look at South Asia, for instance, probably India might, you know, there might be folks in India who would think, ‘Yes, we need a subregional RIR’ and the natural place for that would be in India. But you might find…that no other country in South Asia really wanted to be under an RIR that was run by India.

Here again, the principle of neutrality is mentioned when it comes to the criteria for managing an Internet registry transnationally. As the central basis of legitimacy for all registries, neutrality fulfills the condition of managing the distribution of IP addresses from a regional center without causing mutual mistrust and territorial divisions. It is precisely the proclaimed self-image as a “technical” organization (in contrast to a political one) that gives both registries their legitimizing power and helps emphasize their neutrality. The NIRs can certainly be seen as a concession to the cultural, linguistic and political diversity of the region, while the respective cultural and territorial spaces are integrated into the open discourse format of APNIC.

Networked Regionalization between Centrality and Fragmentation

A look at the interpretation patterns of the discourses at RIPE and APNIC shows that the Internet registries have emerged from a general problematization of political centrality and territoriality. Regionalization can thus be seen as a spatial solution or as a strategic “middle ground” between centralized political control and the ongoing threat of internet fragmentation. Political neutrality and a community-driven, bottom-up consensus are the recurring patterns of organizational legitimization at APNIC and RIPE.

What also becomes apparent is the historical significance of certain sub-regions, which spread from Western Europe in the case of RIPE and from Japan and Australia in the case of APNICs to the rest of the region. As mentioned in the interviews: culture, language, and political diversity were central factors regarding the specific regionalization of APNIC and the formation of National Internet Registries within the organization. However, both registries began with a small social network of technically qualified people who shaped the IANA function from specific regional centers and disseminated it through a specific open culture of decision making and discourse. For this, the conferences of these organizations played a fundamental role in the expansion of that specific form of governance and the constitution of those regions, helping to socially stabilize the networks of people over time and convey this specific communicative culture of Internet governance across national borders.

Within geography, the concept of the region has so far been thought of primarily in relation to places and territories. Among other things, regions have been thought of as local responses to capitalist processes, as a set of cultural relationships that enable shared identification, or as a medium of social interaction⁹. This blog post can, in contrast, be understood as an attempt to consider regionalization in relation to the network space and network logic of the Internet. Its governance emerges as a specific figuration in a historic and conflicting interplay with distinct logics of territorial control. Thus, the regionalization of the Internet can be interpreted as a result of conflicting spatial logics, as a solution to the problem of networked governance that counters particular and territorial logics and the impending fragmentation of the Internet with a regionally organized social network and open political participation opportunities.

The map of Internet regions from the initial example precisely illustrates the tension between a network space and a territorial space¹⁰ as an imagined reification of a regionalized figuration of control. It remains to be seen whether this specific structure of multistakeholder Internet governance will prevail in the coming years, or whether there will be further nationalization or even questioning of the regional organization of the core functions of the Internet.

Author biography

Sezgin Sönmez is a sociologist and postdoctoral researcher at Technische Universität Berlin. He conducts research in the CRC subproject B02 “Control/Space”, which analyzes the spatiality of digital infrastructures in contextures, maps, and discourses.

---

1 Sönmez & Knoblauch 2023.

2 The term is interesting here, as it seems to combine the social actors of Internet governance with the complex technical materialities of the Internet.

3 Hofmann 2016.

4 The discourse was examined using a Sociology of Knowledge Approach to Discourse (SKAD) by Reiner Keller (2007).

5 RFC stands for “Request for Comments” and is an open exchange format that has existed since the end of the 1960s and is still used today by the Internet community to publicly communicate suggestions regarding technical and organizational optimizations.

6 Gerich 1992.

7 The emphasis was added by the author.

8 The only other world region that has NIRs is LACNIC, with two registries in Brazil and Mexico.

9 see Gilbert 1988; Gregory et al. 2011; Massey 1995; Paasi 2002.

10 Löw & Knoblauch 2021.

---

References

APNIC (2024): https://www.apnic.net/about-apnic/organization/apnic-region/national-internet-registries/

Gerich, E.P. (1992). Guidelines for Management of IP Address Space. Request for Comments. RFC 1366. Internet Engineering Task Force. doi: 10.17487/RFC1366.

Gilbert, A. (1988). The new regional geography in English and French-speaking countries. Progress in human geography, 12(2), 208-228.

Gregory, D., Johnston, R., Pratt, G., Watts, M., & Whatmore, S. (Eds.). (2011). The dictionary of human geography. John Wiley & Sons.

Hofmann, J. (2016). Multi-stakeholderism in Internet governance: Putting a fiction into practice. Journal of Cyber Policy, 1(1), 29–49.

Keller. (2007). Wissenssoziologische Diskursanalyse: Grundlegung eines Forschungsprogramms. Springer.

Lambach, D. (2020). The territorialization of cyberspace. International Studies Review, 22(3), 482-506.

Löw, M., & Knoblauch, H. (2021). Raumfiguren, Raumkulturen und die Refiguration von Räumen. Am Ende der Globalisierung. Über die Refiguration von Räumen, 25-58.

Massey, D. (1995). Spatial divisions of labour: social structures and the geography of production. Bloomsbury Publishing.

Paasi, A. (2002). Place and region: regional worlds and words. Progress in human geography, 26(6), 802-811.

RIPE NCC (2024): https://www.ripe.net/membership/internet-governance/internet-technical-community/the-rir-system/

Sönmez, S. & Knoblauch, H. (2023). Grenzen im Internet? Die RIPE-Debatte um das Internet anlässlich des Ukrainekrieges — Working Paper No. 13. Berlin: TU Berlin.