Stigmergy: the most important concept you’ve never heard of
coordination without planning or communication
Many of the concepts we rely on every day were once considered abstruse, technical, or even incomprehensible.
Energy was an abstract bookkeeping device invented by nineteenth-century physicists.
Feedback emerged from control theory and cybernetics.
Gravitation began as an invisible force acting at a distance, which many of Newton’s contemporaries found metaphysically troubling.
Natural Selection was Darwin’s explanation for how complex organization can evolve, initially in biology, but eventually extended to psychology, economics, computer science, and other domains.
Today, these ideas are not only familiar — they are indispensable. We use them to reason about physical systems, organisms, social dynamics, and even our own lives.
What changed was not the concepts themselves, but our recognition that they captured something fundamental.
The concept of “stigmergy” is at precisely this stage today: sufficiently well-defined to provide us with a fundamental new insight about nature, life and society, but not yet widely known.
The Concept
Stigmergy is a concept with a clumsy name and a modest origin. It was first proposed by the French entomologist Pierre-Paul Grassé to explain how termites build complex nests without blueprints, leaders, or explicit communication. Yet beneath that narrow biological context lies one of the most powerful and general solutions to a problem that pervades modern life:
How can agents coordinate their actions without central control, shared plans, or direct communication?
We usually treat coordination as a cognitive problem — one that requires intelligence, negotiation, foresight, planning or authority. When coordination fails, we blame ignorance, incentives, or human nature. But in many cases, the problem is far simpler and far deeper.
It is a problem of information flow between agents and their activities.
Stigmergy solves this problem by shifting where information is kept. Instead of residing inside individual minds or in messages transmitted between them, information is offloaded into the environment. Agents act on a shared medium, and those modifications leave traces that guide the actions of others. The trace becomes both a memory of what has already been done, and a signal to do what still needs to be done.
The Mechanism
The principle underlying stigmergy is so simple that once once you have understood this, it becomes impossible to unsee:
An action produces a trace or mark in an external medium. Perceiving the mark stimulates a next action.
The mark represents the work already done, while suggesting the work that still needs to be done. It does not matter who makes a mark or who perceives it. As long as the mark remains clearly visible in the medium, the information it provides can be picked up by anyone at any time, without needing to know who made it or for what purpose.
The action stimulated by a mark creates a next mark, adding to what is there. This new mark will in turn incite a next action, and so on … Thus, actions by different (or the same) agents build on each other, in a virtuous cycle of activity eliciting more activity. As a result, the accumulated marks are gradually elaborated to form an increasingly sophisticated structure.
Two prototypical illustrations of such stigmergic development are the collective intelligence of insects, such as termites and ants, and the collective development of Wikipedia.
Where the Idea Came From: Termites and Ants
The concept of stigmergy did not originate in computer science or social theory, but in the careful observation of insects — in particular, termites and ants.
At first glance, termite mounds look like the result of intelligent design. They can be several meters high, contain intricate networks of tunnels and chambers, regulate temperature and humidity, and even manage airflow and gas exchange. Yet no termite oversees the construction. No termite has a blueprint. No termite understands the structure as a whole.
Each termite follows extremely simple rules:
Pick up a grain of soil
Drop it where similar grains already accumulate
Take into account local cues such as texture, or chemical traces
That’s it.
Yet from these local actions, coordinated by changes in the shared environment, an astonishingly complex structure emerges. The mound itself becomes a guide for further construction. What has already been built constrains and directs what will be built next.
Ants provide an even clearer illustration.
When ants search for food, they initially wander more or less randomly. As they move, they deposit faint chemical traces — pheromones — on the ground. If an ant happens to find food and returns to the nest, it reinforces the trail it followed. Other ants are more likely to follow stronger trails, and in doing so they reinforce them further.
Very quickly, a branching network of paths appears, with the most efficient routes becoming the most strongly marked. This network now functions as a collective map of the environment, guiding ants to the most productive food sources. No ant calculates distances. No ant compares alternatives. No ant decides which path the colony should use.
The “decision” is embodied in the trail itself.
This is the key insight that led to the concept of stigmergy: coordination does not require communication between individuals, only interaction through a shared environment. The environment stores the results of past actions and channels future ones.
What made this idea revolutionary was not that insects behave this way, but that the same principle can operate wherever many agents interact with a modifiable medium — including human societies, technologies, and knowledge systems.
Once this biological insight is generalized, termite mounds and ant trails stop being curiosities. They become the simplest, clearest demonstrations of a universal mechanism of self-organization.
The Coordination Problem We All Live With
Coordination problems are everywhere.
How do thousands of volunteers write and maintain an encyclopedia?
How do teams collaborate without endless meetings?
How does science progress without anyone overseeing the whole enterprise?
How do traffic patterns, languages, or social norms emerge without designers?
The traditional answers invoke hierarchies, planning, agreements, or communication. But these solutions scale poorly. As the number of participants grows, centralized control becomes brittle, and explicit communication becomes overwhelming.
What stigmergy reveals is that coordination does not require that agents understand the global picture or agree among each other about who would do what. It only requires that they can:
Act locally
Leave visible traces of those actions
Respond adequately to traces left by others
The intelligence is not in the agent, nor in a controller above them, but in the interaction between agents and a shared environment.
What Is Stigmergy, in Plain Language?
In the simplest terms:
Stigmergy is coordination of actions through the traces of past activity.
An action changes the environment.
That change stimulates the next action.
Over time, coherent patterns emerge.
No one needs to plan the whole. No one needs to even know that a “system” exists.
Stigmergy works because the environment does part of the cognitive work. It stores information, highlights what is relevant, and channels future actions. Thus, it functions like a “collective mental map”, a growing system of guidelines that channels activity to where it is most effective.
This makes stigmergy not just a coordination mechanism, but a form of distributed cognition.
Systems We Already Rely On (Without Noticing)
Once you look for stigmergy, you find it everywhere, not only in biology, but in society, psychology, computer science and even chemistry. It is common especially online.
Consider Wikipedia: the largest encyclopedia ever built, providing millions of in-depth articles in dozens of different languages covering about any subject you could imagine. Starting such a gigantic enterprise in the traditional way, by having a board of editors assigning different articles to be written to different authors, and then checking those articles for accuracy, would never have succeeded. That is why Wikipedia from the start chose a method of distributed self-organization, opening up pages to anyone willing to contribute.
Contributors do not coordinate by assigning tasks or negotiating who writes what. They simply add to and improve what they see: correcting errors, adding references, reorganizing sections. Each edit leaves a trace. Those traces guide the next contributor.
For example, when I read a Wikipedia page on a topic I have an interest in, I will notice when there is a mistake, or something relevant missing. That will stimulate me to correct the mistake or add the missing information by editing the page. That newly edited version will now be read by others, who will similarly be stimulated to correct my mistakes or omissions.
The same method can be applied to shared, editable documents, such as Google Docs. Here, collaboration does not happen primarily through meetings or the exchange of emails, but through direct modification of a document visible to everyone in the working group. The document itself becomes the focal point that aligns everyone’s actions.
Open-source software, scientific citation networks, shared databases, even collaborative playlists work in the same way. Contributors coordinate with the medium, not with each other individually.
These systems succeed precisely because they minimize the need for explicit agreement.
Why Stigmergy Works So Well
Stigmergy works because it reduces coordination to simple, local decisions, while still allowing global structure to emerge.
Each participant only needs to answer questions like:
What needs improvement here?
What is missing?
What looks wrong or outdated?
The shared environment ensures that:
Useful contributions accumulate
Redundant work is discouraged
Errors tend to be corrected
Structure gradually stabilizes
In this sense, stigmergic systems are self-correcting. They adapt not because anyone oversees them, but because feedback is built directly into the environment.
This makes stigmergy remarkably robust, scalable, and tolerant of individual mistakes.
What Happens without Stigmergy
Traditional methods of collaboration that rely on planning and communication are vulnerable to errors: messages being misunderstood, agents not being available, tasks failing to be performed … The more precise, rigid and centralized the plan, the higher the chance that a single error will throw everything off course, resulting in failure. With stigmergy, on the other hand, no plan needs to be communicated, and activity adapts directly to the state of the work. If one agent fails to act on the medium, sooner or later another one will do what is needed.
The following joke illustrates what can go wrong if workers tightly stick to the plan instead of paying attention to the trace:
A pensioner watches two city workers busy in the municipal park. The one digs a series of deep holes at regular intervals. The other one then shovels the mounds of earth carefully back into each hole, and flattens the soil. The pensioner asks him: “Isn’t that a waste of effort what you are doing?”, to which the worker replies: “No, we always work this way, and it is very efficient. It is just that the third guy who plants the trees did not show up today.”
Why We’ve Missed the Concept
If stigmergy is so powerful,
why isn’t it widely known or taught?
One reason is that we are biased toward intentional explanations. We like to attribute order to plans, leaders, or minds. Stigmergy, by contrast, explains order without invoking anyone who “knows what they are doing” at the system level.
Another reason is that stigmergy sits between disciplines. It is not just biology, not just sociology, not just cognitive science, and not just technology. As a result, it often falls through the cracks. Yet this in-between position is precisely what makes it so important.
A third reason is that the mechanism is indirect. Stigmergy is a form of self-organization, but one we tend to ignore because it is less visible. In better-known forms of self-organization, such as swarms of birds flying in sync, the coordination is direct: each bird aligns its flight with the one of its immediate neighbors as it senses them. Thus, alignment happens in real time.
But in the case of stigmergy, there is an intermediate stage—the trace left in the medium—that allows agents to coordinate without knowing about what the others do or even being present at the same time. Because we focus on direct interactions, we tend to overlook this passive trace which is lying in wait to be sensed.
Why Stigmergy Matters Now
We live in a world of increasing scale and complexity.
Scientific knowledge is exploding.
Digital platforms connect billions.
AI systems interact with human contributors.
Global problems demand coordination without global control.
In this context, stigmergy is not a curiosity — it is a design principle.
Understanding stigmergy helps us:
Build better collaborative platforms
Design healthier online communities
Rethink governance and participation
Understand collective intelligence
See cognition as something that extends beyond individual minds
Most importantly, it offers a way to think about coordination that does not rely on authority, coercion, or constant communication — but on shared structures that quietly guide action.
A Concept Whose Time Has Come
Like energy, feedback, or gravitation, stigmergy names something that was always there, but only recently became understood scientifically.
We are already living in stigmergic systems. We depend on them daily. We design them — often blindly — and suffer when we design them poorly.
Giving this phenomenon a name is not just an academic exercise. It is a step toward understanding how intelligence, order, and meaning can emerge from collective activity without central control.
Stigmergy deserves to become common knowledge — because it explains how much of our world already works, and how it might work better.
Further reading
My review paper, in which I clarify how precisely stigmergy works, is becoming a citation classic on the subject:
Heylighen, F. (2016). Stigmergy as a universal coordination mechanism I: Definition and components. Cognitive Systems Research, 38, 4–13. https://doi.org/10.1016/j.cogsys.2015.12.002
Its follow-up paper classifies different forms of stigmergy and applies it to the origin of cognition and cooperation.
Heylighen, F. (2016). Stigmergy as a universal coordination mechanism II: Varieties and evolution. Cognitive Systems Research, 38, 50–59. https://doi.org/10.1016/j.cogsys.2015.12.007
Here are two older classic citations that inspired me:
Theraulaz, G., & Bonabeau, E. (1999). A Brief History of Stigmergy. Artificial Life, 5(2), 97–116. https://doi.org/10.1162/106454699568700
Parunak, H. V. D. (2005). A survey of environments and mechanisms for human-human stigmergy. International Workshop on Environments for Multi-Agent Systems, 163–186.
Here is my paper applying stigmergy to the web and open-source development:
Heylighen, F. (2007). Why is Open Access Development so Successful? Stigmergic organization and the economics of information. In Bernd Lutterbeck, Matthias Bärwolff, & Robert A. Gehring (Eds.), Open Source Jahrbuch 2007 (pp. 165–180). Lehmanns Media.
What a great concept.
Didn't know it.
First thought is that probably the concept explains why "Self-Discipline" is mostly a scam. We don't act based on what we "want." We act based on the signals in our immediate environment.
If your phone is on the desk, the "trace" says: "Scroll."
If your running shoes are by the door, the "trace" says: "Run."
We are engineers of our own stupidity because we design environments that trigger the wrong algorithms.
Great article.
Bravo!
Could stigmergy be a solution to the Prisoner’s Dilemma? I don’t see how. It works with ants and termites because all the “players” have the same motivation (reproductive success), but does it work in a human society when each individual have their own goals, some that conflict?