The Ethnography of a DAO
The following post offers an excerpt of a methodology has been developed in the course of my research to conduct ethnography in Decentralized Autonomous Organizations (DAOs). These questions, first identified in this post on ‘DAO Design Patterns’, have becoming a seminal starting point for engagement with, and analysis of DAOs. I hope you find them a constructive starting point for orienting yourself in the field.
Some prescient questions when analyzing a DAO include:
What is being decentralized?
Who or what is being made autonomous (both functionally and politically), and from who or what?
What is being automated?
What is being organized? (Nabben, 2022).
These questions will be detailed as follows.
What is being decentralized?
Decentralisation in crypto communities refers to the physical distribution of the people that run the computing architecture that runs the network and the political influence over the network (Buterin, 2017), as well as meaningfully empowering those individuals to participate in steering the collective (although this is not always the case). Thus, decentralization of blockchain-based protocols is composed of protocol, nodes, ecosystem, and digital tokens (Muzzy & Anderson. n.d.), requiring both technical and social consensus to function. Interactions between ‘nodes’ manifest across both digital and physical domains. It is for this reason that the ethnographic field of blockchain communities is described as “multi-sited” and “radically networked” (Rella, 2021).
The moving feast of crypto conferences shifts each month from city to city, forming a connected web of close interpersonal (despite often pseudonymous) relationships. Personal connections contribute to the flow of governance processes across chat applications, forums, Web application voting mechanisms like ‘Snapshot’, on-chain data from “multisignature” wallets, and end user addresses. In a regulatory context, a DAO can only be considered “sufficiently decentralized” when purchasers of a token asset “no longer reasonably expect a person or group to carry out essential managerial or entrepreneurial efforts” (Hinman, 2018). The qualitative dynamic of blockchain decentralisation is also pertinent to the analysis of political influence in a decentralized system, as the ability for an actor or group of colluding actors to influence or co-opt the system.
Who or what is being made autonomous?
By definition, DAOs include some notion of autonomy, but “what is required for a smart contract to rise to the level of a DAO is not exactly clear” (Wright, S.A. 2021). In practice, autonomy is relative. Ethnographers ask relevant questions here, including how do people think about their own autonomy, both individually and as a group, relative to the autonomy of others and other groups, what does it mean to engage in an autonomous system, and is the system autonomous from its own members? (Cefkin & Stayton, 2017).
The concept of autonomy in DAOs stems from the cybernetic aesthetic of computational organization. Dilger’s initial expression of a Decentralized Autonomous Organization is as a “self-defining and self-maintaining system” capable of “evolutionary” processes (Dilger, 1997). Here, autonomy refers to the emergence of meaning from within a system, comprised of individuals participating in a greater whole (Varela, et. al. 1974).
In blockchain communities, DAOs are conceived of as “blockchain-powered organization that can run on its own without any central authority or management hierarchy” (Wang, et. al., 2019). Blockchain experiments embody the politics and power structures which they want to enable in society (Husain, et. al. 2020). Here, autonomy refers to political autonomy, meaning individual freedom through self-governance, enabled by digital infrastructure and automation that removes the need for trusted third parties in economic and social interactions.
DAOs also aim for functional autonomy, at a collective level. Functional autonomy is the degree of flexibility an individual or group within an organization has, in order to respond to complexity or challenges as they see fit (Swann, 2020). Collective autonomy is the collectively determined capacity and scope an individual or group has to decide and act within the constraints set by collective organization. Thus, autonomous parts must make trade-offs to operate in conjunction with other autonomous parts, in order to function effectively within the intentions of the organization as a whole.
What is being automated?
Participation in a DAO requires engagement in an evolving entity that is comprised of people and automated components. Automated decision-making systems (ADMs) are defined as involving “procedures in which decisions are initially — partially or completely — delegated to another person or corporate entity, who then in turn use automatically executed decision-making models to perform an action” (Algorithm Watch, 2019). While DAOs seem like an extreme example, they are concerned with everyday societal structures of resource coordination and organizing, making them well suited for ethnographic inquiry.
Automation in DAOs refers to automatic execution of transactions, according to the rules of the system. On programmable blockchains, this occurs via “smart contracts”. Smart contracts are “computer programs stored on the blockchain that allows us to convert traditional contracts into digital parallels” (Szabo, 1994), and fundamental building blocks of blockchain-based applications. Smart contracts are programmed to define rules in advance like a regular contract but are automatically executed via software code when the contract conditions are realized.
Claims are made that “there is no need to wait for a human to execute the result” and “smart contracts remove the need for trust” (Ethereum Foundation, 2022). Yet, this is not entirely accurate as humans are needed to conduct the behaviour that triggers execution of the result. Often, an incentive is provided by the system to do so (such as miner rewards). In this sense, smart contracts are “passive” in the sense that members must make transactions and its only when a collection of individual transactions satisfy procedural criteria that the DAO affects an action at the organizational scale. Amart contracts follow predictable, procedural processes, and are not substitutive for human judgment or interpretation, similar to any algorithm that is programmed by directors and software engineers to follow the processes and procedures they are designed for (Burrell, 2016).
In practice, DAOs display varying degrees of decentralisation and automation (Tse, 2020) and ethnography is essential to garner a clearer sense of the design decisions, implementation, and social implications of these systems.
What is being organized?
The purpose of a DAO is not in what it claims or aspires to do but what it does (Beer, 2002). Rather than being ‘an organization’, DAOs are defined by a shared, functional purpose, and can be understood as in an ongoing, iterative social process of ‘organizing’. The parameters of DAO organizing are often institutionalised in some form of shared statement, whether that be a signed constitution, manifesto, or terms and conditions of engagement (Zargham & Nabben, 2022).
The referent object or resource being coordinated in DAOs can be diverse. The functional categories of DAOs purposes include protocol DAOs, where people come together to build and/or maintain a protocol that can function autonomously either with or without upgradeable smart contract parameters; Social Club DAOs, where people convene over shared common interests or hobbies, Service DAOs, where core contributors provide an ongoing service; and Investment Club DAOs, where a community invests in a particular asset class or for a particular purpose (Brummer & Seira, 2022). In some cases, these purpose are composed to provide multiple utility functions or operate at scale.
DAOs offer a unique industrial setting and cutting-edge domain where ethnography is being practiced to learn about human-machine organization and organizational resilience in digital domains.
Ethnographic contributions to DAOs illuminate its ‘aliveness’ by tracing the materialities of not just code and hardware but people, motivations, processes, relations, and politics. The ethnography of DAOs foregrounds not the technical but the social, human dynamics of these assemblages. ADM systems, technologies, and devices do not and cannot exist independently from human thought, embodiment and action. They are inextricably linked to humans, and entangled within social relationships, cultural contexts and human-made organizations and infrastructures (Lupton, 2019). Ethnography in this domain helps to “re-humanise automation” by accounting for how humans are involved at every stage of the design, development, implementation, and maintenance of automated systems (Pink, et. al., 2022). The study of DAOs then becomes a comparable field site of broader enquiries into where, why, and how people use automation in social institutions, how this may be navigated effectively, and the social benefits and drawbacks that this affords.
Remaining hard questions include how to explain or transfer ethnographic practices (such as reflexivity) and insights to a decentralized organization, how to incorporate ethnographic insights as feedback loops into design processes in sustainable ways for ongoing maintenance given the time it takes to communicate emerging results of ethnographic practice to design engineers, and how to measure the impact of ethnography on a DAO.
Acknowledgments: This analysis is an excerpt from a broader piece entitled “The Ethnography of a DAO”, co-written with Michael Zargham. Thank you to Ellie Rennie for review.