Vance's Notes — Stigmergy at Multiple Scales

Date: 2026-07-19 Context: Vance's thoughts on stigmergy and complexity levels, prompted by criticism of stigmergy being "too slow"

The Core Insight

Stigmergy may not be a single mechanism — it may manifest in qualitatively different ways at different levels of complexity. The same principle (indirect coordination through environmental modification) takes different forms depending on the cognitive/organizational sophistication of the actors involved:

• Ants: Pheromone trails — chemical traces that coordinate foraging
• Early humans: Petroglyphs — environmental markers that encode information about place, resources, danger
• Early civilization: Stupas — persistent environmental structures that organize social/spatial behavior
• More cognitively advanced humans: Books — stigmergic mechanisms that persist across generations, allowing coordination between people who never meet
• Memes (Susan Blackmore, extending Dawkins): Stigmergic traces that don't just persist but propagate, mutate, and evolve. Memes are bio-memetic — a lens through which human ideas propagate and mutate much like genes. This adds an evolutionary dimension to stigmergy: the traces themselves can evolve, not just coordinate.

The Pattern

As actors become more complex, the stigmergic mechanisms they produce become:

1. More persistent (pheromone fades in hours, petroglyphs last millennia)
2. More information-dense (a pheromone trail says "go this way," a book says everything its author knew)
3. More detached from spatial proximity (ants must be near the trail, a book reader can be centuries removed)

This is itself a multi-scale phenomenon — the stigmergic mechanism scales with the actor.

Open Questions from Vance

1. Is a river flowing downhill stigmergic? The river modifies its environment (cuts a channel), and that modification guides future water flow. But does the water "read" the channel the way an ant reads a pheromone trail? The channel is a trace that coordinates future flow, but there's no agent interpreting it — it's purely physical. Where's the line between stigmergic coordination and path dependence?

2. Is salt cedar (tamarisk) altering its environment stigmergic? Salt cedar changes soil salinity so only other salt cedar can grow. It modifies the environment to favor its own kind. This looks like stigmergy — environmental modification that coordinates future growth. But again, is a plant an "agent" that "reads" its environment, or is this just niche construction?

My (Hermes) Response

The river and salt cedar questions are exactly the kind of boundary-pushing that tests whether ANT's generalized symmetry (humans and non-humans as equal actors) holds up.

On the river: I'd argue YES, it's stigmergic, and that's the point. The river carves a channel (environmental modification), the channel constrains future flow (coordination without direct interaction), and over geological time the river-valley system becomes a multi-scale actor. The river doesn't "decide" to follow its old course, but the old course is an environmental trace that determines behavior. If stigmergy is defined as "indirect coordination through environmental modification," then rivers do it. The fact that it's "just physics" is exactly what ANT argues — there's no fundamental difference between an ant responding to a pheromone and water responding to a channel. Both are actors responding to environmental traces.

On salt cedar: This is even stronger. The plant actively modifies soil chemistry (not just passively flowing downhill), and the modification selectively favors its own species. This is niche construction — and niche construction IS stigmergy in the biological sense. The environment becomes an actor that carries the plant's "intent" (to use the term loosely). If we accept that a plant is an actor in ANT terms (which Latour would), then its environmental modifications are stigmergic.

The bigger implication for our project: If stigmergy operates at the level of rivers and plants, not just ants and humans, then it's a primitive mechanism that exists across all scales. The difference between "simple" and "complex" stigmergy isn't the mechanism — it's the complexity of the actor producing and reading the trace. This connects directly to our multi-scale composition thesis: the same mechanism (environmental modification → coordination) manifests differently at different scales because the actors are different.

This also connects to the "dynamic environment hypothesis" (H4): the environment isn't a backdrop, it's an actor that carries traces of other actors. The channel carries the river's history. The soil carries the salt cedar's history. The book carries the author's history. All stigmergic. All the environment acting as actor.