DAO Constitution Can Never Be Fully Reduced To an Algorithmic Representation

From P2P Foundation
Jump to navigation Jump to search


Joshua Tan:

"Institutions, organizations, constitutions, and computational constitutions give us a way of talking about the design of DAOs in a holistic way.

Visualizing the need for DAO instances to have institutional and constitutional patterns, which may optionally be aided by DAO frameworks (like Aragon or DAOStack) or computational constitutional patterns (like the Commons Stack toolkit).

After all, a constitution (even a non-computational one) is more than a document; it’s a description of the values and principles that govern an institution’s decision-making processes. It defines roles and associated rights, the protocols for exercising those rights, and the conditions under which they can be amended (e.g. voting). But a constitution also provides a basis for a shared identity; it extols shared values and may provide vision or direction regarding future decision-making under those values without providing specific details as to what types of decisions might need to be made. In this way, a constitution can never be fully reduced to an algorithmic representation. But not all constitutions are created equal — clearly, some constitutions help communities prosper and others lead to conflict and dissolution. In one famous set of institutions studied by Ostrom, namely the International Forestry Resources and Institutions (IFRI), different configurations of roles, rights, rules, and protocols led to widely different outcomes in different contexts. There was no single, “optimal” constitution or institution that worked across all forests and all communities. However, certain arrangements were repeated across many examples, such as monitoring systems and community consultation. We refer to such arrangements as constitutional patterns: each pattern refers to a family of related constitutions that may be applied to a local context.

In this sense, DAOs, as they exist today, tend to resort to a relatively small family of constitutional patterns related primarily to making and voting on proposals to direct the expenditure of funds. Raising funds, proposing expenditures for specific work, voting on those expenditures and verifying the completion of work, as well as lodging and resolving disputes, are necessary activities for a wide range of organizations. The institutions that underlie these activities are being implemented by organizations like 1Hive and are being assembled into reusable patterns by projects like the Commons Stack. It is not surprising that these elements of social logic have been the primary focus of computational constitutional development thus far.

These early examples have created a lot of enthusiasm around the idea of DAOs, but there is a lot of exploring ahead to discover workable institutional patterns (and anti-patterns). It’s also important to remember that the computational constitution is not the whole constitution, and that automation does not magically result in a healthy institution. We believe that taking an institutional approach, and adopting some concepts from social science, will help us to establish and steward healthy DAOs. The institutional approach can also help us organize the range of different projects in the DAO space (see diagram above), and identify common research questions. We’re researching design principles and institutional patterns in hopes of establishing a mature science and engineering practice for DAOs."