Research Summary: The Dissensus Protocol: Governing Differences in Online Peer Communities

TL;DR

  • The paper illustrates the complexities of governance and how they are often sidestepped in governance protocol design. The notion of dissensus, the opposite of consensus, is introduced as a remedy to simplistic ways of treating governance.
  • Based on case studies and a historical review of DAO governance, the authors construct a framework for situating the evolution of governance through three phases and an account of the usual strategies used for resolving tensions and disagreements in blockchain-based communities.
  • Purely algorithmic strategies to solving the issue of governance fail to do so. Every time the human factor is excised from the picture it is inevitably hurled back in because disagreement is an inevitable aspect of human interactions in democratic contexts. Dissensus is proposed as a notion that explains this phenomenon.

Core Research Question

Can governance be resolved purely through algorithmic protocol design?

Citation

Brekke, J. K., Beecroft, K., & Pick, F. (2021). The Dissensus Protocol: Governing Differences in Online Peer Communities. Frontiers in Human Dynamics, 22. Frontiers | The Dissensus Protocol: Governing Differences in Online Peer Communities

Background

  • Consensus: Consensus is a group decision-making process in which group members develop, and agree to support a particular decision (e.g. the one in the best interest of the whole or the most valid decision).
  • Consensus Protocol: A decentralized means of validating transactions between parties without presupposing trust or prior knowledge of each other, and without the need for a central authority.
  • Dissensus: Dissensus can be defined in opposition to consensus. It is the presence of divergence, disagreement and even conflict inside of or between groups in political settings. Dissensus has been characterized as a foundational and inevitable aspect of democracies.
  • Decentralized Autonomous Organization (DAO): A public and permissionless organization that is governed programmatically and is not under the control of a single entity.
  • Governance: The problem area for answering who can make decisions, how they are made, and who is accountable for a given system.
  • On-Chain Governance: On-chain governance refers to the system for managing the control and development of a blockchain-based network. This encompasses methods for collecting information, structuring incentives, collecting signals from ecosystem participants, and ultimately achieving consensus (usually voting) on decisions.
  • Off-Chain Governance: Off-chain governance refers to the set of possible ways in which collective decision-making happens outside of the blockchain network. This often involves forum discussions, chats, physical interactions and other kinds of informal decision making.
  • Austrian School Economics: Macro-economic school of thought that emerged in Vienna in the late 19th century. A central issue concerning adherents of this school is indirect coordination in free markets in the presence of ambiguity or the absence of specific information. Austrian theorists have emphasized the importance of economic signals such as price changes of goods which reflect deliberate actions of economic subjects, or interest rates that reflect the effect of time on price.
  • Stigmergy: Asynchronous coordination of multiple actors via the interactions of multiple agents within a common environment (e.g. leaving traces that act as signals for action) that often results in the emergence of complex structures without the need for any top-down means. The term originally derives from the study of insects, but it has found applicability in a variety of scientific domains outside of biology.
  • Backfeed: Backfeed was an experimental blockchain project for the construction of tools that scale open cooperation without the need of a central authority.
  • DAOstack: An open-source software stack and tooling library for building DAOs
  • Quadratic Voting: Quadratic voting is a collective decision-making process that allows citizens to vote based on how strongly they feel about an issue by allocating credits to votes.
  • Proof of Work (PoW): A consensus mechanism that validates transactions by requiring participants to prove that they expended a certain amount of computing power for solving complex cryptographic problems. This computationally intensive verification is distributed across many nodes so that it becomes nearly impossible for a small number of users to take control of a network.
  • Proof of Stake (PoS): A type of consensus mechanism used by blockchain networks to achieve distributed consensus where users stake their assets to become a validator in the network. Validators are responsible for the same thing as miners in proof-of-work: ordering transactions and creating new blocks so that all nodes can agree on the state of the network.

Summary

  • Consensus in discussions of digital governance in blockchain-based communities is considered a primary goal, but equally pernicious is dissensus which includes all these aspects of social interaction that lead away from consensus e.g. disagreement, conflict, dissatisfaction with the outcome of a decision, etc.
  • The authors define the notion of dissensus in contrast to consensus and set a framework for understanding behaviors of dissensus as an integral part of digital governance.
  • They examine two characteristic cases of dissensus through two case studies of organizations with diverging strategies in treating dissensus, Genesis DAO and Ouishare.
  • Treatment of the standard strategies that are used for resolving dissensus issues in on-chain governance is conducted, identifying four main consensus mechanisms: consensus protocols, voting, staking and forking.
  • The authors conclude that political issues and ideologies are inevitable even in a context where the solution proposed is purely technical. The fact that some blockchain communities prefer to approach the issue of governance as purely technical is an ideological stance in itself.

Method

  • The paper consists of a theoretical synthesis of reports and resources, producing a novel framework for approaching algorithmic governance. It engages in a critical analysis of online blockchain governance through the lens of notions and frameworks related to French political theory.
  • The authors include two case studies based on testimonials from members of each respective organization. Reports from individuals have been anonymized.
  • The authors also gather a diverse range of resources about the past of blockchain communities to synthesize an account of historical progression.

Results

  • The authors trace and construct a model of the historical evolution of governance ideas. Three stages have been identified, each with its distinct approach to what governance is and how blockchain technologies are related to norms and political life:

    • Materialist Governance, whose approach could be summarized by the motto “Code is law”, views governance as algorithms and algorithmic protocols as immutable.
    • For example, TheDAO was an attempt at operating a purely code-driven venture fund organization. It ended up in one of blockchain history’s most infamous hacks which led to well-funded attempts at solving the “governance problem” in the Ethereum community.
    • Design Governance on the other hand approaches protocols as a means to an end and focuses on how they steer incentives.
    • Influenced by the Austrian school, it posits that market dynamics and incentive design can replace the rigidity of imposed rules. Humans in such systems are treated as fully rational actors producing predictable behaviors.
  • This inspired “2nd wave DAOs" that utilized game-theoretic tools with the goal of aligning incentives, and marked the appearance of debates in the Ethereum community about on-chain vs off-chain governance. The former attempted to treat coordination problems as problems that can be fully resolved through blockchain protocols while the latter emphasized the inevitability of discussions.

  • Emergent Governance is based on notions about self-organization and emergence, which is governance as collective coordination.

  • This arose out of the increasingly prominent reports of government and market failures and introduces an element of adaptivity where specific protocols, code and governance mechanisms take different forms in different contexts.

  • For example, Backfeed, a project that applied the notion of stigmergy, analyzes economic signals such as collaborative ledgers that can function as signals for physical coordination and production.

  • The authors study dissensus in practice by exploring two diverging approaches to governance employed by two organizations with divergent values and governance strategies.

    • We are confronted with two case studies, one from Genesis DAO and one from Ouishare.
    • Both projects attempted to use network technologies to scale decentralized self-governance beyond regional boundaries but their approaches are pulling in two radically different directions.
    • In Genesis DAO, decisions were made exclusively through on-chain voting.
    • At one point tensions started arising that were difficult to resolve on-chain, which led to the formation of an ideology around governance where the algorithmic governance system was seen as an arbiter of truth for decisions.
    • The algorithm was approached as a solution to consensus where there was no need for debating and settling conflicts because the outcome of the algorithm was seen as the optimal way of scaling decentralized democratic processes.
    • In contrast, Ouishare is a non-profit distributed organization based on the organizational principles of direct action and local autonomy.
    • Ouishare had problems coordinating action outside of meetings and dialogues but maintained a highly adaptive character which led to its internal transformation.
    • The ability to embrace dissensus and ambiguity and work around them led to making difficult decisions as a group in contrast to splitting into different factions.
  • An attempt to map the territory of blockchain governance systems is made. Through a thorough review of the history of DAOs, the authors have identified four primary ways of resolving dissensus in blockchain communities: Consensus Algorithms, Voting, Staking, Forking.

  • Consensus Algorithms such as proof-of-work (PoW) and proof-of-stake (PoS), ensure that the network collectively agrees on contents stored and executed in the blockchain ledger. The above have been proposed as purely algorithmic solutions to trust, security and consensus by distributing control of the infrastructure over a network of nodes rather than delegating it to a central authority.
  • Voting is an economic mechanism to address voter apathy. In decentralized blockchain communities, democratic processes often take the form of on-chain “coin holder voting”. Reputation systems have been proposed in order to escape plutocracy by making active contributors more powerful in voting.
  • Organizations that have their own token often issue Governance Tokens for active contributors. Quadratic Voting also adds an element of weight to the vote so that users can express the degree of preference of one choice over another.
  • Staking is a way to add weight to a vote by locking a proportionate amount of funds to a smart contract. Staking also enables bets about predictions of voting outcomes which can reward winners. Hence, it enables various indirect ways of signaling preference to actors in a governance process (emergent governance).
  • Forks are copies of the repository of an open-source project that make important changes to the code. There exist two types: Soft forks are temporary and compatible with the rest of the network (e.g. add-ons), while Hard forks are permanent and the new code is incompatible with the original version (e.g. new project and token).

Discussion and Key Takeaways

  • Three-stage evolution of blockchain governance: The progression from materialist to design and finally to emergent types of governance reveals a tendency towards mechanisms that incorporate the human factor as an integral element of design rather than trying to solve it through purely technical means.
  • The “Dissensus” Protocol: By transporting the notion of dissensus from political theory to blockchain communities and applying it to governance mechanisms used by them, the authors come up with the Dissensus Protocol, a framework that takes into account the unpredictability of human behavior in blockchain governance.
  • Dissensus as a Driver of Hard Forks: They propose that hard forks should be approached as cases where dissensus is dominant and strong contentious disagreements arise that cannot be resolved and thus partition the online community into two. One of these two parts is always minoritarian and emerges out of an inability to influence decision-making through voting.
  • The Perils of Hard Forks: Hard forks don’t lead to sustainable projects, because they appropriate years of work merely by copying a repository. Thus the tokenomics of hard-forked projects often suffer and lead to losses.
  • Algorithmic Governance as Ideology: The authors identify that there is a peculiar tendency arising in blockchain governance communities that form strong ideologies around governance and approach algorithmic governance as an end in itself rather than as a means for better coordination towards common goals.
  • Governance Minimization: The authors argue that the history of blockchain communities shows that attempts at “solving” governance paradoxically led to governance maximization, because members are confronted with disagreement about the implementation of the same consensus protocols that purport to solve governance. For this reason, using blockchain-based communities has since shifted the focus from solving governance to “governance minimization”.

Implications and Follow-ups

  • This is a cautionary tale for organizations that focus too much on the algorithmic aspects of governance. Hard forks can often be detrimental and organizations should be aware of what causes them to prevent them.
  • People involved in DAOs should be more self-aware of their stance towards governance, because it may hide ideological commitments.
  • Forking is usually celebrated as a triumph of open-source code in web3 and DAOs tend to encourage them, but this tendency may hide the conflicts and tensions that exist in the communities involved.

Applicability

  • As DAOs and web3 native organizations are multiplying and because of their portrayal as the future of work, it is becoming all the more important to be clear of the complexities of blockchain governance. The paper provides a new language and vocabulary for DAOs to bring awareness to difficult and uncomfortable issues.
  • The Dissensus framework can inform attempts at creating hybrid forms of coordination systems where the algorithmic stack is complemented by community management, design and conflict resolution.
6 Likes

Congratulations @Hermes_Corp on this excellent contribution to the Governance section of the Forum.

The term “Dissensus Protocol” can seem somewhat confusing. It’s not a “network protocol” like TCP, SMTP or HTTP, but it clearly does point to “a way of doing things.” Can you say a little bit more about the choice of this term and its significance?

3 Likes

The term “protocol” in “Dissensus Protocol” may indeed confuse some people because it gives the expectation that the reader will be introduced to a technical solution. But this is nowhere to be found! However, it does indeed point to “a way of doing things”. So what does this peculiar and confusing phrase really mean?

Well, as it turns out, the Dissensus Protocol turns out to be an intentional word-play. The term is used in an ironic fashion in order to contrast itself to the Consensus Protocol, which is a technical solution. Consensus protocols like PoW and PoS exist to ensure verification of transactions in the network by making sure all nodes in the network agree on the current state of the blockchain. This is a purely algorithmic means of achieving consensus which some people view as a way of getting people “out of the equation”.

Given the overall criticism of the paper against the “code is law” stance, the use of the phrase gains a subtler layer of meaning. The phrase “Dissensus Protocol” in its ironic usage is not only aimed at presenting a better, improved version of algorithmic governance based on the inevitable fact of the existence of dissensus in online communities. Instead, it is used to point out that such a way of treating governance is dated, unrealistic, and even doomed to reproduce the problems which it itself brought.

The choice of words in the title is very intelligent because it seeks to specifically “bait” the people that follow the stance it aims to criticize.

6 Likes

Great summary @Hermes_Corp

Please, could you explain this?

1 Like