Kelsie Nabben
December, 2022
This blog post is a special one. It syntheses some of my initial PhD findings on ‘resilience in decentralised technologies’, ahead of a talk at the conference ‘What’s Governing Web3?’ at the Centre of Excellence for Automated Decision-Making & Society in December, 2022.
TLDR:
Decentralised technologies are resilient when a phenomenon that I term ‘self-infrastructuring’ occurs. ‘Self-infrastructuring’ is when people are able to participate in designing, owning, operating, governing, and/or maintaining their own infrastructure for resilience. It is the boundaries people place around their own actions in relation to shared purposes or goals, that are then expressed in technical and institutional infrastructure. Self-infrastructuring emerges from the collective will to adapt and continue to self-govern according to one’s political preferences. Resilience in this practice is contextual, relating to a person’s purpose or goals, threat model, vulnerabilities, and opportunities. As such, self-infrastructuring is motivated by the political will to self-govern towards specific purposes and against specific threats through technological and social means. Decentralised technology communities’ self-infrastructure in a variety of contexts and through a variety of methods, including algorithmic and organisational structures (such as DAOs), full-stack self-infrastructuring (such as hardware), and institutional infrastructure. How well people can collectively articulate their political objectives, identify their threat model, and clearly express this in software code and coordination rules (on or off-chain), to organise and adapt in line with that objective, determines resilience. When self-infrastructuring does not occur and people cannot set boundaries around their own actions in relation to their goals, resilience breaks down.
Introduction
Decentralised technologies are often thought of as secure, empowering, and ‘resilient’. This multi-year research project set out to develop an in-depth understanding of the affordances of decentralised technologies in terms of resilience through an infrastructure studies lens and qualitative mixed methods of historical analysis and ethnographic observation and interviews with a variety of community-level developers, implementers, and users of decentralised technologies, and through identifying and tracing vulnerabilities. Here, I take resilience to mean “adaptability and transformability of a socio-technical system in response to threat or crisis” to foreground the dynamic, continuous adjustment, between humans and technology, as people pursue their goals, or flexibility under uncertainty, in people’s given context. This research has taken place at the disparate and ephemeral digital-first field site of ‘decentralised technologies’ to investigate and observe the concept of resilience and the infrastructural practices of decentralised technology communities across conception, design, development, implementation, and use. I show what resilience is in decentralised technologies, when and how it occurs, and under what conditions it is limited or challenged.
I show how resilience is conceived of, enacted using decentralised technologies, and experienced (or not experienced) by participants in its use to respond the overarching research question of ‘are decentralised technologies resilient?’. I find that resilience in decentralised technologies is understood and occurs through the practice of ‘self-infrastructuring’.
By self-infrastructuring, resilience in decentralised technologies is determined by the boundaries people place around their own actions in relation to a specific, shared purpose. It is in the study of infrastructuring in decentralised technology communities that leads to the discovery of resilience in decentralised technologies as ‘self-infrastructuring’. Good information infrastructures are those that are adaptive to changing circumstances and environments (Star & Bowker, 2010). Infrastructure is a ‘doing’ word and resilience is a dynamically emergent phenomenon, not a static end state or outcome. The verb “to infrastructure” denotes the activities, processes of integrated materials, tools, methods, and practices that make up and change an infrastructure (Star & Bowker, 2010). Thus, ‘infrastructuring’ is an ongoing process of doing, and these processes are incremental, iterative, and long-term (Karasti et al., 2010). Infrastructure emerges as a “relational property”, inseparable from its use, which exists or “is” when its purpose connects to its functional use by a person or community (Star & Ruhleder, 1996). The boundaries of action and interaction in decentralised technologies are infrastructured through both technical processes and social processes, resulting in both technical and institutional infrastructure. Self-infrastructuring enables voluntary participation in the rules of infrastructure, which allows for adaptivity and accountability in relation to shared goals. The agency that self-infrastructuring affords helps to facilitate adaptivity amidst internal or external crisis, without so much deviation from shared goals that infrastructure collapses.
Self-infrastructuring is why decentralised technology communities are adaptive and resilient in some cases — when people can collectively articulate their political objectives, identify their threat model, and clearly express this in software code and coordination rules (on or off-chain), to organise and adapt in line with that objective. This depends on how well people articulate their purposes and goals, identify their threat model and own vulnerabilities, and express this in the rules of infrastructure. Where there is a disconnect between the purpose of a group and these rules of their infrastructure, the ability to self-infrastructure is limited (in terms of the capacity to structure, own, operate, or maintain infrastructure) and resilience breaks down.
Self-infrastructuring is also why centralised digital platforms and infrastructures cannot be considered resilient by decentralised technology practitioners. As one research participant put it, “whoever controls infrastructure controls society”. What is implied is that this must be ‘us’.
This insight informs the three key findings, that: resilience in decentralised technologies is about ‘self-infrastructuring’, self- infrastructuring is political, and that when self-infrastructuring does not occur, resilience breaks down.
1. Resilience in decentralised technologies is about ‘self-infrastructuring’
Resilience in decentralised technologies is about ‘self-infrastructuring’, meaning awareness of and participation in designing, owning, operating, governing, and/or maintaining one’s own infrastructure. It is the boundaries people place around their own actions in relation to shared purposes or goals, that are then expressed in technical and institutional infrastructure
In this context, resilience is enhanced by writing your own infrastructural rules and having agency to act according to these rules. This allows for people to take responsibility for their own infrastructure and self-governance. In this context, self-governance is being able to abide by infrastructural rules, be it technical or institutional — as the set of political, legal, and cultural institutions that form the backdrop for economic activity and governance and enable or constrain its operation or adapt them when needed.
Resilience in this practice is contextual, relating to a person’s purpose or goals, threat model, and vulnerabilities. This gives rise to a plethora of possible threats and vulnerabilities in and against decentralised technologies. In this respect, self-infrastructuring is motivated by ‘the will to infrastructure’. The will to infrastructure is the political, collective, strong driving purpose or desire to self-govern through infrastructural means, and against a specific threat, that is informed by a particular world view. The will to self-infrastructure is essential in motivating the practices of self-infrastructuring, despite the potential inconveniences, efficiency costs, and resources required in terms of time, effort, cost, knowledge acquisition, and/or risk. Even if the will to infrastructure is for a functional purpose, there is usually a political motive behind that.
Decentralised technology communities enact resilience by self-infrastructuring in a variety of contexts, against a variety of threats, and through a variety of methods. The self-infrastructuring practices of decentralised technology communities include algorithmic and organisational structures (such as DAOs), what I call “full-stack self-infrastructuring” (such as hardware), and institutional infrastructure (such as governance tools and processes to guide infrastructural practices (Nabben, 2022)).
Self-infrastructuring can both enhance resilience and create new vulnerabilities. These threats and vulnerabilities to resilience in decentralised technologies can be exogenous, attacking from outside of the community, or endogenous, emanating from internal community dynamics or forces. When self-infrastructuring does not occur and people cannot set their own boundaries around their actions in relation to their goals, resilience breaks down. This will be further explored in finding three. The second findings is that self-infrastructuring is political.
2. Self-infrastructuring is political
Self-infrastructuring is motivated by the political will to self-govern towards specific purposes and against specific threats through digital means. Political purposes are inscribed in the infrastructure of self-infrastructuring. Yet, in decentralised technologies, the politics of infrastructure dictate the requirements for resilience to produce new things. The political forms of rationality revealed by decentralised infrastructure is that people can self-govern. Decentralised technologies expressly aim to explicitly embed political rationalities by encoding the rules of governance actions and interactions. Political narratives shape the purpose of decentralised technologies and translate into the technical requirements of how an infrastructure functions.
From the outset, decentralised technologies are reflective of political aspirations, rather than the mathematical invention of public-key cryptography, or a cult of personality (for example, the identity of the Founder of Bitcoin, the pseudonymous Satoshi Nakamoto, remains a mystery to this day) (Nabben, 2021). Each case study reveals how the political motivates, influences, and is affected in each of these instantiations of self-infrastructuring from threat perception, to technological solutions, and institutional dynamics.
Through the political modes of self-infrastructuring new infrastructural dynamics emerge and new voices are inserted into political spaces (more on this coming soon in an accepted paper). In cases where self-infrastructuring occurs, novel dynamics are emerging where new stakeholders hold political influence. What occurs is that infrastructuring takes place through new organisational structures that include the materiality of algorithms, software, hardware, and people. It is not through the efforts of one project or group that political change occurs but through a decentralised movement where each module produces something that others can build on as composable infrastructure. Through these materialities, new things are enabled, from public blockchains, to distributed funding mechanisms, to censorship-resistant blogs, to opensource microchip distribution and integration.
Not only are these infrastructures inextricably linked to politics, from their design, to maintenance, to use but politics is a feature, not a bug, in shaping the threat models, design, and the resilience that infrastructure can afford. How well people can collectively articulate their political objectives, identify their threat model, and clearly express this in software code and coordination rules (on or off-chain), to organise and adapt in line with that objective, determines resilience. This means that resilience is largely determined by factors internal to the system (endogenously), for example, goals, rules, processes, software code, and community norms. Thus, resilience can be controlled for but only by acknowledging forces beyond the technical.
While these bespoke, decentralised infrastructures can be highly dynamic and adaptive through the practices of self-infrastructuring described, they don’t always work in practice, or as intended. This is explored in finding number three that follows.
3. Resilience breaks down when self-infrastructuring does not occur
Resilience is compromised when the political and technical purpose of an infrastructure becomes misaligned. Investigating why resilience was experienced or not in the use of decentralised technologies highlights the inextricable and co-constructive interplay between social and technical components of infrastructure.
By definition, the interplay between social and technical dynamics of infrastructure in a sociotechnical system is inextricably interlinked, so that neither the social (such as people, relationships, and structures), nor the technology (such as hardware, software, and processes), can be considered in isolation from one another (Golden, 2013). In every case study, the importance and role of institutional infrastructure were foregrounded in generating and governing that adaptivity to enhance resilience, as the set of political, legal, and cultural institutions, that form the backdrop for economic activity and governance, which enable or constrain its operation (Hinings, et. al., 2017). Social coordination layers that surround the use of technology required to guide processes and practices of implementation around data ownership, governance, storage, and maintenance of data to bridge the “gateway moments” (Edwards, et. al., 2009) of linking infrastructure for scale, ongoing adaptivity to changes in environments, and to serve people’s needs in their local context (see Nabben, 2022 for the draft working paper. The full one has been accepted by IEEE and is coming soon TM.). Based on this insight, I argue that separating the technical operation of infrastructure and the social processes of structuring and governing (institutional) infrastructure undermines resilience.
Self-infrastructuring as a doing word, means that decentralised technologies are not resilient independent of the humans that use them. Resilience is not about perfecting the technical but that the infrastructure is only relevant when used, and when used, will break down. This means that there is a difference between the narrative of decentralised technologies as resilient, and how resilience is afforded in practice through both technical and social means. In technologically dominant or technologically deterministic engineering cultures such as public blockchain protocols, it is often thought that the technical layer of infrastructure is superior to the weak, corruptible, social dynamics and human layer (Filippi, & Loveluck, 2016).
In the techno-deterministic view of some decentralised technology community members, the social dynamics of a system are only vulnerabilities that are bound to break down, amidst a narrative of “governance minimization” through “governance automation” to remove social dynamics through algorithmic governance (Ehrsam, 2020). Yet, vulnerabilities can occur across the social, technical, economic, and even legal and or environmental dimensions of infrastructure, and only a multi-dimensional approach to resilience can account for the adaptivity required to identify and address these concerns. Indeed, social dynamics can and do break down resilience, but they are also required to make it. This depends on the severity of the consequences of the threat model at hand. For example, one engineer asked me about my findings on decentralised storage about the role of institutional infrastructure in facilitating resilience: “so, it’s the social layer that breaks down, not the technical?”. This misses the point, that resilience does not emerge from infrastructure, it emerges from infrastructures that support human capacities. How people engage to shape, use, or re-purpose an infrastructure is an active process of adaptation, which influences and determines resilience.
In practice, it is not possible for the technical layer to be resilient while the social layer is not, as the social and technical dynamics of infrastructure are co-constitutive. It is through use that resilience is situated in context, dependent on social dynamics, and adapting (or not) to meet the challenges of both external threats and internal vulnerabilities in relation to the purpose of the technology and community in question. When people imagine decentralised technologies as resilient, they often do not take into account the complexity of the lived experience of self-infrastructuring towards one’s goals. This is also true of centralised technologies. We think they are resilient because of the narratives that companies propagate. In this case, people do self-infrastructuring with them too (in that they use them for purposes other than what they were originally intended and may even adapt them in the process). How resilience plays out differently in decentralised technologies relates to agency in relation to rules.
Both the social and technical can be co-constructive or undermining of resilience, as demonstrated in the interplay between sybil attack bots and machine learning defence in GitcoinDAO. Decentralised technologies, being peer-to-peer or permissionlessly participatory infrastructures, by nature do not exist without “embeddedness” in the “organizational arrangements” (Star, 1999) of a community. Infrastructure is “part of human organisation” (Star, 1999). Through the social layer, decentralised infrastructure can also be organised and re-organised (adapted) to improve resilience. The social layer of institutional infrastructure affords adaptivity to deliberate and change the rules of the system in the face of changing circumstances.
Another aspect that delineates self-infrastructuring in decentralised technologies from centralised infrastructure is that institutional dimensions are not imposed on the system from the outside but arise endogenously through the phenomenon of ‘self-governance’. Infrastructure is fundamentally relational; it emerges for people in practice, connected to activities and structures. It consists of both static and dynamic elements, each equally important to ensure a functional system. Infrastructuring accounts for the social and organizational dimensions of infrastructure. This understanding requires adopting a long-term rather than immediate time frame and thinking about infrastructure not only in terms of human versus technological components but in terms of a set of interrelated social, organizational, and technical components or systems (Bowker, et. al., 2010). Resilience, in this context, is process-oriented, as a continual, dynamic, adaptive phenomenon, rather than outcome-oriented.
Self-Infrastructuring Futures
Having previously explored the origins of the concept of resilience in decentralised technologies in my historical research, I turn my attention to the ethnographic futures of decentralised technologies to ‘unfix’ and ‘unsettle’ the concept of self-infrastructuring for resilience in the remainder of this piece.
I hope to share more with you on soon.
Acknowledgements:
Thank you to my supervisors, Professor Ellie Rennie and Associate Professor Chris Berg, as well as colleagues at the RMIT Blockchain Innovation Hub, and collaborators (especially Michael Zargham and the team at BlockScience), without which, this would have been much less fun or interesting). There are also numerous contacts, communities, and research participants who have so kindly engaged in interviews and/or observations, that contribute to these findings.
Thank you also to those that support and encourage my work by subscribing to my blog on Substack or NFTing my work on Mirror.xyz.
References:
De Filippi, P, and Loveluck, B. 2016. “The Invisible Politics of Bitcoin: Governance Crisis of a Decentralised Infrastructure” Internet Policy Review, 5, 3. doi: 10.14763/2016.
Edwards, P, Bowker, G, Jackson, S, and Williams, R. “Introduction: An Agenda for Infrastructure Studies.” Journal of the Association for Information Systems, 10 (5, 2009). doi: 10.17705/1jais.00200.
Ehrsam, F. 2020. “Governance Minimization”. Paradigm. Available online: https://www.paradigm.xyz/2020/10/870#what-governance-minimization-is-not . Accessed December, 2022.
Folke, C., S. Carpenter, B. Walker, M. Scheffer, T. Chapin, and J. Rockström. 2010. “Resilience Thinking: Integrating Resilience, Adaptability and Transformability.” Ecology and Society 15 (4). https://doi.org/10.5751/ES-03610-150420.
Golden, TD. 2013. “Sociotechnical Theory, in: Encyclopedia of Management Theory”. SAGE Publications, Ltd., Thousand Oaks. 752–755. https://doi.org/10.4135/9781452276090
Hinings, C.R., D. M. Logue, and C. Zietsma. “Fields, institutional infrastructure and governance.” The Sage handbook of organisational institutionalism. 2017.
Karasti, H., Baker, K. S., & Millerand, F. (2010). Infrastructure time: long-term matters in collaborative development. Computer Supported Cooperative Work (JCSCW). doi:10.1007/s10606–010–9113-z.
Nabben, K. (2021). “Resilience as ‘Political Decentralization’: An Alternate History of the Cypherpunks Origins of Decentralised Technology”. Available at SSRN: http://dx.doi.org/10.2139/ssrn.3938626
Nabben, K. (2022). “Decentralized Technology in Practice: An analysis of socio-technical resilience in IPFS”. Available at SSRN: http://dx.doi.org/10.2139/ssrn.4082016
Ranciere J. 2006. The Politics of Aesthetics: The Distribution of the Sensible. Transl. G Rockhill. London: Continuum
Star, S. L. The Ethnography of Infrastructure. American Behavioral Scientist 43, no. 3 (1999): 377–391. doi:10.1177/00027649921955326.
Star, S. L., and G. C. Bowker. (2010). “How to Infrastructure.” In Handbook of New Media: Social Shaping and Social Consequences of ICTs, Updated Student Edition, 230–45. London: SAGE Publications Ltd, 2010. https://dx.doi.org/10.4135/9781446211304.n13.
Star, S. L., and Ruhleder, K. “Steps Toward an Ecology of Infrastructure: Design and Access for Large Information Spaces.” Information Systems Research 7 (1996): 111
Hyperlinks:
Kelsie Nabben on Substack: https://substack.com/profile/1619235-kelsie-nabben
Nabben, K. 2021. “DAO Vulnerabilities”. BlockScience Medium. Available online: https://medium.com/block-science/dao-vulnerabilities-509ff074a296.
Nabben, K. (2022). “Decentralized Technology in Practice: An analysis of socio-technical resilience in IPFS”. Available at SSRN: http://dx.doi.org/10.2139/ssrn.4082016.
Nabben, K. (2021). Is a “Decentralized Autonomous Organization” a Panopticon? Algorithmic governance as creating and mitigating vulnerabilities in DAOs. In Proceedings of the Interdisciplinary Workshop on (de) Centralization in the Internet (IWCI’21). Association for Computing Machinery, New York, NY, USA, 18–25. https://doi/10.1145/3488663.3493791
Zargham, M. and Nabben, K. (2022). “Aligning ‘Decentralized Autonomous Organization’ to Precedents in Cybernetics” Available at SSRN: http://dx.doi.org/10.2139/ssrn.4077358