An Ecological Commons

Ecological knowledge lives all around us: In published research, in careful spreadsheets, in seed records, survey notes, species lists, and restoration logs. Some of it is rigorously produced. Some of it is quietly maintained by one person who has been watching the same place for years. Much of it is hard to find and harder to connect.

A commons is a shared space or a way of holding a resource collectively as a community. Land, water, libraries, and seed banks, for example, are shared resources that stay healthy when the community around them tends them with care. An ecological commons extends the same idea to ecological knowledge. Not a warehouse, but a shared practice of describing, registering, and reaching what we steward, so that anyone who needs it can find it and work with it honestly.

What this is

This site is the beginnings of an idea, a home for an ecological commons being built in the open with as much careful intent as possible. It holds the ideas behind the work, the infrastructure that supports it, and a basis for a community of people and organizations to shape what it becomes.

At its center is a simple commitment: knowledge enters the commons on its own terms. Attribution stays with the source. Each source is described honestly. A researcher, a homeowner, a student, or a tool working on someone’s behalf can all reach for it and know what they’re holding.

The knowledge pattern

The pattern is small and repeatable. A source honestly describes what it is — what it holds, who made it, and how it came to be. That description is what makes it findable. A source that no one could find yesterday is integrated and made reachable.

Further, sources that are integrated become navigable and reachable, no matter the form they arrived (a spreadsheet, a file, a partner’s system, a public API). Each integrated source gets its own explorer, a public way to glimpse what’s inside and understand what it means.

The pattern works at every scale. Local, regional, national. Each level is real in its own right, not a zoom on the others.

Why honesty about sources matters

Not all knowledge is the same kind of knowledge. Some was observed in the field. Some was compiled from many observations. Some was synthesized across sources using methods that themselves deserve careful description. A commons that flattens those differences does its users a disservice. A commons that makes them visible lets everyone who reaches for a fact know whose hands it passed through.

That honesty is also what lets a new tool trust what it finds here. As technological tools emerge that help people build, what we can imagine might soon be within reach. Those tools will need ground to stand on. A commons that is honest about its sources is exactly the kind of ground that helps a powerful tool be safe.

What gets built on top

What a commons enables is not bounded by predictions or institutions. Field guides that know your specific site. Restoration planning tools that show the gap between what grows there and what could. Classroom exercises where students explore ecological networks by adding species and watching the interactions emerge. Regional coordination between organizations that no longer have to rebuild the same data work separately.

And beyond any of that — whatever the people who use this commons find themselves able to imagine once the ground beneath them is trustworthy.

Get involved

An ecological commons is a community activity. It depends on the care of the people and organizations who choose to be part of it.

If you hold ecological data — a species list, a survey, a record of observations — and want it to be findable without taking on a technical project, there is a path for that.

If you lead an organization whose work depends on ecological knowledge — a land trust, a native plant society, a restoration program, a research program, a nursery, a university, an agency — there are ways to connect your work to the commons and benefit from what others have contributed. These organizations have different things to offer: data, expertise, staff time, space, reach, resources. A community activity asks each participant to bring what they can. If your organization wants to explore what that might look like, get in touch.

If you build tools, design systems, or want to help shape this infrastructure, we are open to any level of collaboration.

We wonder about the feasibility of a gathering of data holders, regional organizations, and indigenous community members to help shape how this grows. If that sounds like something you’d want to be part of, get in touch.

EcologicalCommons@gmail.com

The Idea

A commons is shared — a resource that sustains many, held and tended by the very community that depends on it. No single participant owns it. It is neither free of care, nor free of rules. It persists because those who draw from it also give back to it.

The original commons is the ecological world itself. A prairie is a commons shared by the grasses, the pollinators, the soil microbes, and the birds that pass through it. A forest is a commons shared by the trees, the fungi connecting their roots, the animals that live within it, and the streams that run through it. Each participant takes what it needs and returns what it can. The system holds because the participants hold it together, not because anyone manages it.

Human communities have long organized around commons of their own — grazing land, fisheries, irrigation systems, village wells, public squares, libraries, seed banks. The work of Elinor Ostrom showed that these arrangements persist, successfully, when the community around them holds practices of reciprocity, restraint, shared decision-making, and mutual accountability. Human commons do not survive on their cleverness. They survive on the care of the people who tend them, and they work best when they sit honestly within the ecological commons that sustains them.

An ecological commons, in the sense this project means, is a human arrangement in service of something larger than data: the ecological knowledge and wisdom that have accumulated in particular hands — researchers, practitioners, field observers, stewards, indigenous communities, long-standing programs — and that most people have no ready way to reach. A commons is one way that knowledge becomes available to everyone, so that more of us can live, decide, and act in closer relationship with the living world.

The material of this commons is broad. Species lists, field observations, survey notes, restoration logs, seed records, community classifications, taxonomic references, soils, weather, images, measurements, histories. Some of it is the product of rigorous science. Some of it is carefully held by people and programs working close to the ground. All of it is knowledge that someone, somewhere, already knows — and that someone else, somewhere else, could act on if only they could find it.

The commons is a shared practice and shared infrastructure for describing that knowledge honestly, making it findable, and making it reachable by the people and tools that need it. Attribution is real and visible. What grows from the commons is a community of participants and layers of trust, not a database.

What makes this particular type of commons possible now is that the infrastructure for such a thing is finally unremarkable. Databases, APIs, provenance tracking, consistent interfaces — the pieces required to make a commons of knowledge reachable are standard. What is hard in this time is not the technology. It is the combination of knowing what the knowledge means, caring about how it is represented, and holding the human practice that makes the whole thing more than a technical exercise.

An ecological commons is an attempt to do that work in the open, at whatever scale people and organizations are ready to hold. It begins where it begins — in this case, in Southeast Michigan with a small number of integrated sources and a regional program already practicing what a commons asks of its participants. It might be a timely idea and have a useful shape. It grows as others choose to be part of it.

Why infrastructure isn’t enough

Building the infrastructure is the easier half. The harder half is everything that determines whether a shared system actually gets used — how organizations decide to contribute data, how different communities find ways to work without forcing consensus, how decisions get made, how systems persist, and how trust is held as new kinds of participants enter the picture.

Two traditions of scholarship and practice have already done serious work on what this commons is trying to become. The information sciences — commons governance, data stewardship, boundary objects, knowledge infrastructures — have studied for decades what makes shared systems work or fail. The computing sciences — federated architectures, grounded synthesis, autonomous development — are active frontiers where the tools being built today will shape what an ecological commons can become. The lessons from both are available to be applied and this application, we think, holds some community-enhancing potential.

• An information frontier →
• A computing frontier →

What becomes possible

The cost of building working software is collapsing. So is the cost of having a question and getting a grounded answer. Against a commons of honest data about the living world, these two shifts mean something specific: the questions people actually ask, and the answers too, can grow into tools that people actually use.

People ask “these” questions constantly. At native plant sales. At workshops. At preserves. In classrooms. On porches. The same questions, over and over, because the answers are scattered across databases, scientific papers, nursery staff memory, and the accumulated experience of volunteers — and there has never been an approach or a system that could simply answer them. The bones of such a solution now exist.

What should I plant here, given my soil and my light, to support the most life?

What was growing on this ground before it was a lawn?

Which plants in my garden are safe around my dog? Around children?

Where are the gaps in my yard’s bloom calendar, and what could fill them?

Is this plant invasive, and what should replace it?

What would this half-acre look like in five years if I kept planting? In ten?

How does my yard connect to my neighbors’ yards as habitat? Where would a single well-placed plant close a gap?

What seeds can I collect nearby, at what time of year, and how do I grow them?

What does a healthy version of this kind of place look like — a mesic southern forest, a wet prairie, a fen — and how far is this site from that?

Where did this recommendation come from, and should I trust it?

I’m a teacher. Can my class build an ecological network for our school grounds?

I’m a volunteer crew leader. What should we prioritize at this preserve today?

I’m twelve years old, and I want to do a real science project on the pollinators in my neighborhood. What should I look for?

Each one of these is something someone can now build an answer to. Not a funded team — anyone with the question and the care to see it through.

The questions in the list are the ones people already ask. The more interesting possibilities are the ones no one has thought to ask yet — the ones that require both a commons of trustworthy knowledge and the ability to build against it cheaply, conditions that have not existed together before this moment.

For the first time, what gets built can be shaped by what would actually help.

What this could become

More data can be added. The integration pattern is small and repeatable. As more sources are described and brought in, the commons gets richer. Some of that will be data from regional organizations and programs. Some of it will be data from community science — iNaturalist, eBird, Nature’s Notebook, and the many smaller observation projects that exist. Hosting community science datasets directly, with their full context and provenance preserved, is one of the simpler things a commons can do.

People can build against it. The commons is most useful when the people closest to the questions are the ones building the answers. Bootcamps for community scientists, crew leaders, teachers, and young naturalists — showing them how to use the tools now emerging to build against the commons — are a direct implication of what is becoming possible.

People can gather around it. A commons requires the people who hold and use the knowledge to talk to each other. A community symposium — data holders, regional organizations, indigenous communities, practitioners, researchers, agencies, nurseries — is where the shape of the commons gets discussed by the people it serves.

What’s Built

An ecological commons begins with the infrastructure that makes knowledge findable. Below are descriptions and links to what currently exists.

The data layer

The data layer holds descriptions of registered ecological data sources, provides a few ways to find them through a consistent set of software interfaces and offers a basic explorer into what they offer. New sources are revealed as they are described and integrated.

The data layer browser lives at: data.ecologicalcommons.org.

Above the data layer, a trust layer exists, with named and author linked domains capture human or expert judgement about which data sources are applicable for different objectives (regions, knowledge areas, or activity).

• OpenAPI JSON
• OpenAPI YAML
• Source registry
• MCP endpoint

Demo Apps

Applications that demonstrate that kinds of things that can be built.

Ecological Augur

An ecological question-answering tool that consults the commons for real occurrence records, taxonomy, and conservation data before answering. Every response shows four sections: the answer, what came from the commons, what came from the AI’s training knowledge, and where gaps remain. You can inspect the sourcing — you don’t have to take the AI’s word for it.

Connecting Concepts

A demonstration and explanation of how large language models work. Submit three concepts; the model finds connections across what it learned during training. The prompt — the instructions shaping which patterns it emphasizes — is shown, not hidden. Same concepts, different context, different connections.

What’s Coming

• Integrating additional sources that have been identified.
• Developing more demonstration applications.
• Inviting community members to understand and collaborate.
• Requesting data owners to discuss and share their valuable products.
• Discussing our shared approach to technical and governance topics.

Old, but still interesting

An early proof-of-concept application demonstrated what can be built on a static version of the commons and one way of dealing with the challenge of trusting agentic systems. It has its own documentation and is available to explore:

• s2c-species.replit.app

A field report describing the proof of concept journey and the early work, through January 2026:

• Field Notes (PPT)

The Trust Layer

A piece of infrastructure that lets people work from the sources their context calls for.

The data layer of an ecological commons can hold hundreds or thousands of sources. Citizen science projects, research datasets, nursery records, park inventories, regional floras, herbarium databases, observation programs, seed collection logs, government surveys, community projects. Each one is real ecological knowledge produced by people who care about something specific — a region, a taxon, a method, a question. The substrate has room for all of them.

Like the current data landscape in the real world, this is also overwhelming.

Some sources are more rigorous than others. Some have decades of curation behind them; some are weekend projects by enthusiasts; many sit somewhere in between. Most focus on a particular domain — a geographic region, a kind of organism, a category of observation, a stage of research. None of them is the right answer for every question. Most support answering interesting questions.

For a subject matter expert, this proliferation usually navigable, given time. They know which regional flora to consult for their region. They know which observation network is rigorous enough for their work. They know which nursery’s data they trust because they have visited the operation. The judgments are real and they are useful, but they live in the expert’s head, learned through years of practice. This is true of all of those who have dove into the work in a serious way.

For a homeowner, a student, a person curious about the plants in their yard, the same proliferation can be crushing. There is no way to know what to trust. The wall of sources looks the same from the outside, and choosing among them seems to require already knowing where to look, how to get access, and how to assemble necessary pieces.

The trust layer addresses this asymmetry. It lets the expert’s judgment become a named, shareable object — a trust group. A trust group is an ordered set of tiers, each tier holding the sources that the group’s author considers most appropriate for a particular context of work. Someone in Southeast Michigan working on native plants might define a trust group whose top tier is Michigan Flora, MNFI, a favorite nursery, and a regional seed program — sources that are local, careful, and trusted by the community of practice in that region. A second tier might hold broader regional sources whose information overlaps usefully but is less locally specific. A third tier might hold global aggregators that fill gaps when the first two are silent.

The point is not that this is the correct ranking. The point is that this is somebody’s ranking, made deliberately, in the open, for a particular kind of work. Other people doing different work in different places will have different rankings, and those will also be correct for their contexts. The Ecological Commons treats every group as equally legitimate at the data layer. It does not endorse. It records and exposes.

Applications use trust groups. An application given a trust group can answer a question without dragging the software or the user through the wall of sources. It reaches for the top tier first, falls through to lower tiers if needed, and stops when it has an answer. The user sees an answer grounded in sources their community trusts. The application did not have to invent its own judgment; it inherited the practitioner’s. The judgment is legible — apps can present which sources were consulted, in which order, and why.

This is a small social action with infrastructural consequences. A practitioner who publishes a trust group is sharing what they trust, for what kinds of work, in their context. The act of defining the group is the act of making implicit knowledge explicit. Other practitioners can adopt the group as-is, fork it, argue with it, learn from it. Communities of practice acquire a shared artifact that previously lived only in conversations and apprenticeships.

The architecture is straightforward. A trust group is a record in a database. Its tiers are ordered. Its members are FERNS sources. An API exposes the group’s contents as a tier-organized list. The data layer is not modified. The source response payloads are not modified. The trust group is a view onto the existing registry, nothing more.

Each group is a small contribution. Each one helps someone who would otherwise have been crushed by the wall. The layer makes room for communities to be themselves.

An initial implementation is built. The first trust groups will be local to Southeast Michigan. Other practitioners, in other places, with other priorities, are exactly the people the layer is built for. Operational integration with applications, partnership with communities whose practitioners want to publish trust groups.

Justifying the Structure

The Ecological Commons has a specific architecture. The data sources are described and kept as themselves, not absorbed into a single rewritten store. Trust is a named, declared layer that filters the data below it rather than a logic baked silently into applications. Names from all traditions and places are treated as first class knowledge, not labels attached to one true form. This page describes the architectural choices made and the reasoning behind them.

Aspect 1 - Ecology is about relationships and place.

Plants, animals, fungi, soil, water, fire, weather, people, and history — all of them interacting with each other in each specific place over time. The work of ecological stewardship is the work of understanding those relationships: what depends on what, what was present and is now absent, what does balance look like, what does diversity look like.

Key Phrase: Relationships in places, not facts about species.

Aspect 2 - The decisions that shape most places are made by non-experts.

Most of the land in any region is held by people who are not ecologists. Homeowners, landowners, municipal maintenance staff, volunteer boards, weekend stewards, neighbors with a backyard and some curiosity. The trained experts — the restoration ecologists, the botanists, the land managers with years of field experience — are vastly outnumbered. That gap matters because what aggregates into the ecological condition of a region is not what the experts decide to do with the land they manage. It is the daily, unremarkable decisions made by the many: what gets planted, what gets removed, what gets mowed, what gets left, what is valued by society.

Key Phrase: Aggregate decision-makers are not the experts.

Aspect 3 - People make most decisions alone, regardless of how many experts exist.

Face-to-face, inspired connections are irreplaceable for society. To meet the people where they are, when experts are not on hand, we must provide the best data and translation possible.

Most decisions, in most domains, happen alone — at a kitchen table, in front of a screen, walking a piece of land before the nursery opens, at a hardware store with a phone and a question. Even in domains crowded with experts and workshops and online forums, the actual moment of decision is often private. Ecology itself is crowded with geological, botanical, and chemical facts that no expert holds, no discussions can cover, or human minds can transmit. Any system that pretends the alone-channel does not exist has missed the place where most of the work happens. We would argue for both more in-person sharing and also systems that share hard-earned wisdom at scale with everyone, everywhere.

Key Phrase: Face-to-face is necessary and insufficient; most decisions happen alone.

Aspect 4 - The data landscape is broken in three specific ways.

First, we are awash in information, challenged to judge, often without training or clear insight. Overwhelming volume, scattered across websites, books, seed catalogs, agency PDFs, extension bulletins, forum threads, and every manner of species descriptions written more for marketing than accuracy. A person trying to answer a real question in a real place has to read across all of it with no systematic way to evaluate what is reliable and what is not.

Second, significant sources are walled off. Some because of institutional incentives, some because of the shape of delivery, and some because of missing social cohesion. The sources that would matter most — rigorous regional flora, conservation status data, peer-reviewed phenology, field observation records — are often the hardest to reach personally and programmatically.

Third, the technology of access keeps moving. Books, then websites, then spreadsheets, then REST APIs, and now MCP protocols. The method will keep changing. Each generation of access opens what the prior generation could not reach, and makes implications inherent in the previous one obsolete. Knowledge that wants to be genuinely reachable has to be structured so that it can move with the access methods.

Key Phrase: Too much, walled off, and the methods of access keep moving.

Aspect 5 - New tools change what is possible.

Against this backdrop, a new class of tools has arrived. Called AI loosely, the term is used with varying interpretations and perspectives about what it actually is. It is a new tool class. Like chainsaws, fire, vehicles, propane, or chemicals … AI has real failure modes, many possible uses, and also genuinely changes what is possible. The failure modes are themselves strange in comparison to other tools: confabulation, training bias, confident wrongness.

Specifically, a model that has absorbed the internet, Reddit, marketing copy and social media produces answers that mirror the quality level of what it was fed. Analysis that feeds on Michigan Flora, MNFI, BONAP, FQA analysis, peer-reviewed work, and field observations can produce amazing insight.

The current capabilities and the pace of advancement are specific and real. For our domain, it’s not a replacement for experts. It could be a way to extend expert knowledge into all the alone-moments where most decisions actually happen.

Key phrase: New tool class, real failure modes, genuinely changes what is possible.

Aspect 6 - Make knowledge findable, well-described, reachable, and connected.

If the problem is that good data is hard to find, hard to use at scale, and the tools that could help are only as good as the data they are fed, then one concrete answer is to make existing data reachable and honored for what it is.

Every source is kept as itself. The Ecological Commons describes each source in depth — what it covers, how it was made, what its limits are, who maintains it — and provides uniform ways to reach it. Keeping sources as themselves means keeping their authority, their governance, and their perspectives intact. Aggregation through absorption and assimilation does not make a vibrant Commons.

Key phrase: The tools are downstream of diverse data. Describe and reach; never absorb.

Aspect 7 - Trust is something to share

The data layer is rich with information, but also so operationally vast as to not be directly usable for most purposes. Most uses need a narrower field before they can begin. The Trust Layer is one possible implementation of how to execute such a filter. Any kind or style of group can be shaped: peer-reviewed botanical sources only, Michigan-focused regional authorities only, named community contributors plus major synonymy publishers, species focused, whatever is useful. The sources of these groups will emerge with the data sources. We are just at the beginning of that necessary Commons governance work. We would expect many passionate users to share their best judgement of the layers as we all experiment with new tools.

When using an application or creating your own, select (or create) any group that is appropriate to the purpose. The data layer itself remains untouched and complete. The application layer above does not have to invent its own trust logic. It can rely upon a declared list from a trusted creator.

Key phrase: Trust is a layer of named, declared scopes — applied above the data, exposed to the user, and explicit.

Aspect 8 — No canonical names; every equivalence is an attributed claim

Living things are known by many names: Scientific Latin, common local names, regional vernaculars, the languages of indigenous communities who lived alongside these species for thousands of years. Each tradition is anchored in a particular relationship between a community and the world.

Indigenous botanical knowledge was, in many places and times, not just unrecorded but actively suppressed. Plants were renamed. Knowledge holders were displaced. The naming conventions that emerged through that period reflect that history — and the silence around what was lost was itself manufactured, not natural. The medicinal, food, ceremonial, and ecological knowledge carried by those traditions was detailed practice, often more developed than what replaced it.

We made two architectural decisions associated with this topic, which sit within the Name Translation Layer.

The first is that no naming tradition is structurally privileged. Every name is a node in a graph. No node type sits above another. They only appear in the Name Translation Layer if their community gives permission.

The second is that every claim of equivalence between names is recorded structurally — who said two names refer to the same living being, when, and on what basis. Both ends of such an assertion of equivalence must be represented in the Name Translation Layer before such equivalences are allowed.

When interacting with a data source, the user can choose to use the Name Translation Layer to bridge the gap between the name they know and the names accepted by the data source. Whatever name a user knows is the center of a small search. The system walks outward from that input, finds what works against the data source, and shows the trail it took.

Key phrase: No structurally privileged name; every equivalence is an attributed claim; the user’s input is the center of the search.

Aspect 9 - Three types of reciprocity as architecture

The Ecological Commons holds three things in relationship, and the architecture is only sound if all three support each other.

The ecological work and posture is the content. A million years of evolution, specialization, and mutual adaptation — organisms in relationship with other organisms in specific places over deep time. The Commons treats this as substance, not subject. It is the reason the architecture exists at all, not to force source conformance, not to impose the “right” perspective. Every decision described above — keeping sources as themselves, refusing a canonical naming tradition, making trust visible and attributable, keeping the Commons neutral — is accountable, finally, to the ecological reality it exists to serve.

The technical work and posture is the shape and structure of the system. The structural design and architectural decisions determine whether the system can be trusted. Sources kept as themselves. Names without a canonical tradition. Trust as a named, declared layer. Applications as where opinions live. These are the architectural disciplines that let the Commons stay honest about its own job. The technical here is what keeps each layer accountable to the layers around it, rather than accumulating quiet assumptions that compound into hidden bias.

The social work and posture is the human relationships: the cross-cultural trust, the decision-making, the regional vision. The machinery of how humans across cultures, regions, institutions, and generations decide what counts, who is heard, what is honored, and what direction the whole thing moves. The Commons treats these relationships as load-bearing, not decorative. They are built into the substrate as named, attributed, accountable contributions rather than assumed at the edges where no one examines them.

None of the three is sufficient alone. The ecological without the technical is unsharable at scale. The technical without the ecological is empty scaffolding. Either without the social is a Commons structure no one is responsible to and no one is responsible for.

Key phrase: The Ecological Commons architects the ecological, the technical, and the social in reciprocity.

Aspect 10 - The data sources are neutral; applications are flexible

End users don’t generally interact with the data layer, the Trust Layer or the Name Translation Layer directly. People use applications (explorers, guides, tools, interfaces) built on top of those layers. Those applications will use the source access methods best suited to their design. Application objectives can adapt as new sources are linked, application goals expand, or human needs shift.

As functionality advances, agents and agentic systems (that can take natural language direction as inputs, explore among selected sources, and return answers) will emerge. These new tools are themselves applications, though with strange and sometimes disconcerting failure modes. They need clear context and thoughtful guidance to avoid failure cases, which take time to learn and flexibility to apply. Technology advancement holds the potential to minimize these failure modes and to also allow anyone to request a new application tool of their own imagining, without the need for more technical human support. As these tools advance and trustworthy data remains essential, the Ecological Commons can offer a grounding in trusted truth.

Through all these uses, the Commons remains neutral and source-faithful, with its provenance intact and its sources stable and reachable. The Commons remains neutral and source-faithful. The applications carry the opinions and the audience. We bring the imagination.

Key phrase: The Commons is neutral; applications are flexible.

The Naming Layer

A piece of infrastructure that lets people reach ecological knowledge from the names they actually use.

Living things are known by many names. A plant in a Michigan field has a name in scientific Latin. It has names in English — sometimes several. It has names in other European languages, in indigenous languages of the people who lived alongside it before it was first written down by botanists, in regional vernaculars used by gardeners and naturalists, in the conventions of programs that grow it and share it. None of these names is wrong. None is more correct than another. Each is anchored in a particular relationship between a community and the world.

Existing ecological data systems do not treat the names this way. They pick one form — almost always scientific Latin — as the canonical identifier, and treat all other names as labels attached to the canonical form. A community whose primary naming tradition is not Latin is required, structurally, to translate itself into Latin to participate in shared infrastructure. The translation is not neutral. It encodes a hierarchy that the underlying realities do not support.

The Naming Layer takes a different approach. It treats every name as a node in a graph. It treats every assertion of equivalence between names as an edge with provenance — who said two names refer to the same thing, on what basis, when. It treats a species concept as whatever connected component emerges from the graph, not as a pre-declared identifier owned by a particular naming tradition. No naming tradition is structurally privileged. A user from any tradition can ask the layer for an answer from any source, and the layer translates between traditions at query time.

The layer learns from use. Each time a name is sent to a source, the layer records what happened. Names that are accepted by a source accumulate as operational knowledge of how the source behaves. Subsequent queries are fast because the layer remembers. Queries from naming traditions that are new to the layer pay a small first-time cost, and then are fast for everyone who comes after.

Communities can contribute names and equivalences as first-class graph contributions. A community member adding a name from their tradition is making the same kind of contribution that a major synonymy publisher makes — recorded, attributed, available alongside every other assertion. The data layer does not pick a winner among contributors. It records what was said, by whom, on what basis, and lets the application or the user choose what to honor.

This is, structurally, a small piece of infrastructure. The architecture is straightforward to implement and uses standard technology. What is unusual about it is what it refuses to do: it refuses to designate any naming tradition as canonical, refuses to invent equivalences the system was not told about, refuses to merge belief and operational behavior into a single fact. The discipline is in keeping the system small and honest, so that every layer built on top inherits the symmetry rather than inheriting a hidden privilege.

The Naming Layer is part of the Ecological Commons. It sits above the data layer of integrated sources and below the applications that people build to ask questions of the living world. The architecture has been designed in detail and is documented in a technical specification. Implementation, integration with a wide range of sources, partnership with communities whose naming traditions have been historically under-documented, and release of the architectural pattern for adoption by other projects are the work that follows.

The Naming Layer is documented in detail in a philosophy document and a technical specification. People interested in the architectural reasoning, the implementation, or potential collaboration can reach the Ecological Commons at the contact address on the main page.