Abstract
Decentralised platforms such as Bitcoin and Ethereum were created to enable people and machines to transfer value without the need for platform companies (Brunton, 2019; Maurer et al., 2013). While parts of these systems replicate financial products occurring on centralised platforms (albeit with novel governance structures – see Schär, 2021), others are developing alternative approaches to value creation. One mechanism that has so far received little attention is contribution systems. The cases considered here operate with a different platform logic to that typically associated with finance – namely, big tech's processing of proprietary data into payments and credit lines (Frost et al., 2019). Contribution systems instead observe things that are being done in the world and trace the dependencies that led to each specific action, typically using collective processes for deciding what is important. They invite us to rethink platforms not simply as systems for extracting value from the behaviours of users, but as a means to compute the relationships that underpin value production, broadly defined.
A contribution system is a mechanism that recognises, measures, and may reward contributions made by individuals or machines to a collective endeavour even when the participants are unknown to each other. While they have existed in limited form for some time (Kealey and Ricketts, 2014), contribution systems have become a fundamental mechanism in public blockchain networks and are sometimes used within permissionless organisations, such as decentralised autonomous organisations (DAOs) (Bauwens and Niaros, 2016; Davidson et al., 2018). Contribution systems are increasingly being conceived and built where knowledge and the resources to achieve an outcome are distributed beyond a typical organisation or firm, including land regeneration efforts that utilise blockchains to produce, record, establish, verify and transfer value.
Bibliographic citation systems are a simple version of a
Contribution systems that are created within or on platforms are a potentially significant alternative to price and exchange-based value as they are capable of creating value and incentivising activities that would otherwise go untraced and unrewarded. Before exploring this, I will first specify how contribution systems create what is sometimes called
How contribution systems produce value
A simple representation of a contribution system for citations would connect references to each other; nodes in a graph with lines (edges) between them, showing that one has built on the other. For example, although it is not described as such in the documentation relating to the decentralised blockchain, Ethereum, its proof-of-stake consensus mechanism is a contribution system. Ethereum is a network of computers (nodes) rather than a central entity or infrastructure. The nodes must coordinate among themselves and reach agreement on the state of the blockchain, collectively determining what information is valid and creating a base of shared knowledge for activities that occur on the platform. Validating nodes (those that have contributed to secure the blockchain by depositing or ‘staking’ resources) are randomly selected to propose a block of transactions and to attest to the validity of other blocks. As anyone with the required stake (Ether deposit) and skills can become a validator, these nodes are effectively assessing the contributions of other nodes and, through Ethereum's software clients, will penalise malicious or unreliable nodes who threaten the network's shared reality (i.e. consensus on the state of the blockchain). Although the benefits of Ethereum – its ability to hold common knowledge – extends to all who transact there, validators gain disproportionately for their contributions in the form of Ether-denominated rewards and can be punished for failing to provide useful contributions, including having their deposit destroyed by the protocol.
The Ethereum example demonstrates how contribution systems can operate at the protocol level to derive value through the actions of machines and the people who run them. The same approach to gathering data about both actions and the reception of actions can be applied between layers of the network's wider ecosystem, producing value when a particular component, design or technology is picked up by another human or software agent. In such cases, a contribution system determines a value based on the relationship between the original contribution and the subsequent action that recognised or drew on it, which may then be used to calculate a benefit (pay) from the latter to the former. As van Epps (2024) writes, what is commonly discussed in blockchain design discussions as maintaining public goods is not about public goods at all; rather, it's about recognising ‘dependencies’ between networks, software and libraries. A contribution system is a more accurate descriptor as it monitors dependencies and calculates value from them, then potentially administers funding for recognised non-rivalrous but excludable ‘contribution goods’ (Kealey and Ricketts, 2014).
This capacity to recognise dependencies and assign value can extend beyond software environments. For instance, contribution systems are employed in some land regeneration projects to formalise and recognise contributions to ecological restoration. Contribution systems in these contexts monitor the contributions of humans and non-humans (animals, plants and soil) in relation to each other. They aim to express a relationship of reciprocity with nature rather than appropriation and compensation. Contributions here can have ritualistic properties – pure action that has an ordering factor to it (Hermann-Pillath, 2024) – where the value is created through the combination of everyday, routine actions performed by people and non-humans. These projects don’t reward based on impact but instead recognise that land management occurs as a cumulation of small, interrelated undertakings, the value of which is not captured in land ownership. One example is rotational sheep grazing in vineyards, which enhances soil and reduces fire risk, in a project associated with the Regen Network. The project deploys monitoring technologies to verify when short-duration sheep grazing has been provided (typically by someone other than the landowner) and allocates credits to shepherds (see Horning, n.d.). Those involved in the creation of such biocultural credits are experimenting with the value that is assigned to animals, trees and other non-human entities, which may become the major shareholder with the power to ensure that governance processes cannot be undermined. In this way, regenerative land projects are beginning to illustrate how contribution systems can be used to reposition non-humans as actors in ecological institutions, overcoming extractive, production-based approaches to creating value from land (Smith, n.d.).
From valuemeters to interobjective value
The value produced through contribution systems can be used to make a claim on resources, demonstrate reputation, or determine access rights. Importantly, this value is not calculated by measuring the production of goods or services (labour theory of value) or by price (exchange theory of value). Instead, value is established by computing relationships – considering not just actions themselves, but the context of those actions, including how and when they are received, and their connection to other actions. Contribution systems collate agreement on the existence of a contribution and locate it within the broader context of actions and the identities of contributors. This requires data collection but also an ability to represent dependencies between data points. These are not mechanisms for commodifying users or predicting behaviour (as in many critiques of datafication), but tools for tracing how contributions circulate and are recognised. Value emerges from contextual interrelation. Data is not the primary source of value; rather, it is a component in the computational process of recording value that is generated when a contribution is traceably appreciated or built on by a person or a machine.
In this respect, contribution systems function as what Latour and Lépinay (2009) – in their pamphlet on Tarde's economic anthropology – call a
Contribution system valuemeters create what is sometimes referred to by their creators as intersubjective value because they process relationships and aggregate judgments using both qualitative and quantitative metrics. Value is computed based on a potentially wide array of connections. In doing so, these systems make aspects of participation that might seem fleeting or ephemeral to humans (Kelty, 2019) legible to machines, which can use this information to coordinate other processes and capital (not only financial resources but also to the broader capacities of a system to mobilise and allocate attention, effort, and infrastructural resources toward collective goals). The value generated through a contribution system is not tied to a preexisting economic structure but emerges from interactions. It is also intertemporal – realised not at the moment of the action itself but at various points in time, sometimes long after the original contribution was made.
Rather than being intersubjective, we could say that contribution systems work with
Timothy Morton's concept of interobjectivity (based on object-oriented ontology) is useful here. Where intersubjectivity is a framework for understanding social relations among human actors through shared perceptions and intentions, interobjectivity repositions objects and humans as participants in a vast, interconnected network where agency and influence are not confined to human subjective perception. As Morton writes, ‘The phenomenon we call intersubjectivity is just a local, anthropocentric instance of a much more widespread phenomenon, namely interobjectivity’ (2013: 107). Relationships and exchanges are constantly occurring between all kinds of entities, whether human, non-human, living, or non-living. The interactions we traditionally label as ‘social’ also happen between objects independently of human perception. In this sense, we, too, are objects with material properties, influenced by and interacting with other objects, like the air, weather systems, or the devices we use. Interobjectivity is based on a philosophy that has jettisoned anthropocentric assumptions of modernity, accepting not only that things have a presence beyond perception by the human mind, but that objects leave footprints that are signs of causality, including being visible (albeit partially) through processes of computation and data. Contribution systems monitor and produce value from these interactions.
How can they fail?
When contributions are evaluated based on narrow information, such as reactions – for instance, likes in online channels – contribution systems can favour that which is most visible or easily understood. If contributors feel some contributions are not appropriately recognised then it can disrupt community cohesion. Contribution systems that are designed to produce value that is determined by social responses, therefore, require mechanisms through which those involved can ensure that evolving priorities are made known and adjusted. For instance, those involved in SourceCred's contribution system at one point chose to favour interactions on Discord as these could show informal and affective labour that was not visible through the GitHub plugin (although this was later found to be problematic – see Rennie, 2023). As SourceCred discovered, a system also needs to be able to adequately produce a membership based on contributions and to manage the boundaries of who is considered a contributor without becoming susceptible to extractive behaviour (also observed by Eghbal in her 2020 study of GitHub). If only a few individuals possess the necessary expertise to modify the system's rules, the system may become stagnant or power may become concentrated, undermining the participatory nature of the decentralised network. Systematic under-recognition for that which cannot easily be measured may mean that essential tasks are overlooked. Commons-style governance (Ostrom, 1990) can therefore be key to their design and success, which can include processes for deciding what the system should track and weight as it calculates value. Unlike platform governance, which tends to centralise control over infrastructure and decision-making, commons governance requires broad accessibility to the tools that govern value assignment, ensuring that adjustments can be made collectively. Systems that fail to acknowledge these complex dynamics risk misaligning incentives, stagnating, and ultimately losing any meaningful engagement. Bollier (2024) observes that those who participate in the
An unwritten rule of any contribution system is that those who contribute should be rewarded over those who do not. In the case of Ethereum, smart contracts have been built on the platform that provide opportunities for individuals who do not wish to directly participate in consensus to receive validator rewards via a third party who have staked on their behalf (known as liquid staking). As the benefits of liquid staking have come to outweigh the benefits of direct staking (running a validator node oneself), one consequence is that power has accrued to these large third parties, undermining the blockchain's ability to withstand censorship and collusion. The Ethereum example suggests that contribution systems may weaken when they compete with other forms of value creation. In this case, platform dynamics designed to maximise liquidity, scalability, and passive participation run counter to the logic of contribution systems, which depend on traceable primary inputs being prioritised over rent-generating abstractions.
Why research contribution systems?
The extent to which contribution systems may alter what is valued, and therefore what occurs in the world, is still unknown. We do know that these systems may help resolve practical problems related to sustaining infrastructure development and maintenance, particularly in cases where a protocol or system is not managed by a single entity. Protocol Guild, a funding mechanism for Ethereum's development and maintenance, is an example of this approach: developers who work on the Ethereum protocol collectively decide on the membership of Protocol Guild, and those who are accepted receive rewards administered by a smart contract that controls a basket of assets, donated by applications built on Ethereum. Such automated systems help to overcome some of the drawbacks of grant-based funding, including the costs and inefficiencies associated with evaluation. They may be transformative in similar contexts where change requires inputs across a distributed group of actors and systems, such as in land regeneration projects where the consequences of actions can extend beyond the boundaries of private property and demand forms of coordination that top-down regulation struggles to effectively manage. Understanding how contribution systems can both aggregate and value contributions made in pursuit of long-term goals, whilst enabling value to be created and utilised in the present, could provide a powerful model for tackling complex, global challenges.
If platforms can be designed to compute interobjective value, they cease to be merely sites of extraction and become engines of coordination across distributed agency. Where contribution systems succeed is when value accrues to activities that, although infinitesimal in isolation, mesh together to form an interconnected response to a challenge or mission that has otherwise proven difficult to coordinate. They may help address ‘hyperobjects’ (Morton, 2013); phenomena that cannot be experienced in their entirety but that make their presence known indirectly – like unseasonable raindrops on your skin, hinting at the hyperobject of climate change. How do we alter something that is so high-dimensional in time and space, like climate change, which is non-local and reveals itself interobjectively, as footprints of relationships between objects and never as a whole? Capitalism, as Morton points out, cannot effectively comprehend hyperobjects because it focuses on discrete, exchangeable objects – ‘whatever comes in at the factory door’ (p. 36). A different approach, as described here, is to meet the hyperobject with a machine that can take in a wider range of information – gathering the many small actions that follow a desired path, toward an outcome that is beyond the thought or influence of an individual human and which may occur far into the future. This machine takes inputs about actions and states, calculates a value from the connections between those inputs, and works in the expectation that coordination will follow.
Footnotes
Declaration of conflicting interests
Funding
ORCID iD
