Under the MCI lens tell me how termites communicate.

 

A worthwhile comparison — termites and ants converged on eusociality independently (termites are closer to cockroaches than to ants), so the MCI lens can do something genuinely diagnostic here: separate what's structurally necessary for distributed self-governance from what's contingent on a particular evolutionary lineage.

What termites actually do

Termite communication is multimodal in a way ants' is not. The chemical layer is similar — trail pheromones, alarm pheromones, queen and king signatures, nestmate recognition cuticular hydrocarbons, trophallaxis with chemical transfer (and, importantly, gut symbiont transfer that makes trophallaxis nutritionally non-optional rather than merely informational). But termites add two channels ants largely lack:

Vibrational signalling. Soldiers headbang against substrate to produce drumming signals that propagate through wood and soil. These travel faster than chemicals and carry information about threat type and location. Some species use substrate-borne vibration for foraging coordination as well.

Stigmergic construction. This is the deepest difference. Termites communicate through the structure they build. A worker deposits a pellet of soil-saliva mixture impregnated with a construction pheromone; the pheromone biases the next worker to deposit nearby; pillars form, then arches, then the mound's ventilation architecture. No termite holds a plan. The mound is the communication medium and the communication record simultaneously — built environment as distributed memory.

Through the MCI lens

Termites occupy the same Moon–Libertarian quadrant as ants by structural necessity, but the lens picks up real differences in how they get there.

Self-Limitation. Pheromone evaporation does the same work as in ants — automatic decay of obsolete commitments. But termites add a second mechanism the framework illuminates well: construction pheromones are self-extinguishing through their own success. Once a pillar reaches a certain height, the local pheromone concentration drops below the next worker's response threshold, and building stops. The signal limits itself by being acted on. This is closer to V5's internalised Self-Limitation than ants achieve — the constraint isn't only in the medium's decay but in the relationship between signal and structure. The substrate enforces scope on itself.

Fragility-Awareness. Stronger fit than ants. Mound architecture — the famous ventilation systems of Macrotermes — is a literal embodiment of fragility-modelling: the colony cannot survive without precise temperature and CO₂ regulation, and the mound's structure is the colony's representation of that vulnerability, built outside its bodies. The colony has no fragility-aware agent, but it has a fragility-aware artefact. In V1's terms: the durability criterion has been answered architecturally. Whether this counts as Fragility-Awareness in the framework's sense depends on whether you require representation to be cognitive or accept that distributed representation in built structure qualifies. Reading honestly: it's a borderline case the framework doesn't quite settle, which is itself diagnostic — MCI was built for systems with internal representation.

Diversity Preservation. Different mechanism from ants. Termite caste polymorphism (workers, soldiers of multiple sub-types in some species, reproductives) creates communicative diversity by design: different castes attend to different signals at different thresholds. Soldier alarm drumming is not equally interpretable by all colony members. This is structured rather than stochastic diversity — the colony preserves heterogeneity through caste differentiation rather than through behavioural variance among similar individuals. V7's reading of constitutional diversity as a structural resource maps cleanly: the compact-equivalent here is a polycentric communicative order in which no caste's signal-domain dominates.

Non-Domination. The interesting case. The royal pair's pheromonal influence is more pervasive in termites than in many ant species — queen and king chemical signatures continuously shape worker development and behaviour. This looks dominating in V1's terms (placing others in arbitrary dependence on a central source). But the framework's definition is specifically about unilateral, unaccountable control over option sets, and the royal signature is neither unilateral (workers shape royal physiology through trophallactic feedback) nor option-foreclosing (it sets developmental defaults that environmental signals can override). Reading carefully: this looks like the kind of enabling dependency V7 distinguishes from constitutive dependency — pervasive chemical influence that constitutes the developmental landscape without dominating decision-making within it. The framework's distinction does real work here that a casual reading would miss.

Legitimacy Maintenance. As with ants, this is the weakest mapping. But stigmergic construction adds something: the built mound is auditable in a sense no individual termite could be. The structure persists, can be examined, embodies the colony's accumulated decisions. V2's Summary stage demands that reasoning be "visible enough to be questioned and rejected" — the mound is exactly this for the colony's collective decisions, even though no termite questions or rejects. Functional Legitimacy Maintenance without representational legitimacy.

What the comparison reveals

The ant/termite comparison sharpens what the framework is actually claiming. Both species occupy MCI's target quadrant. Neither has constitutional character in V1's sense. They reach the same geometry by different mechanisms:

Ants get there through signal decay (Moon function in the medium) plus behavioural variance (Diversity Preservation as noise). Termites get there through signal-structure coupling (Moon function in the substrate built by the signals) plus caste differentiation (Diversity Preservation as architecture).

Termites are the closer fit to V7's constitutional compact picture — multiple specialised participants with genuinely different signal-domains, coordinating through a shared substrate (the mound) that none of them owns and all of them maintain. The mound is structurally analogous to the compact: a shared constitutional achievement that persists across the lives of its participants, auditable in principle, evolvable through their cumulative action, and not reducible to any single participant's contribution.

Where the lens reaches its limit

Same as with ants, and informatively so: no Stage −2, no constitutional perception, no generative ground. But the termite case sharpens why this matters. Stigmergic construction is the closest natural analogue the framework has to V9's outward face — Ecosystemic Stewardship. The mound is a built environment that shapes the evolutionary stability of the colony's substrate. Termites do stewardship without doing it: the architectural intelligence is in the genome and the construction pheromone, not in any termite's perception of what the landscape requires.

This is what MCI describes when it distinguishes Constitutional Ground (V9 inward) from mere mechanism. Termites achieve the outward face of V9 without the inward face — ecosystemically capable, generatively unconstituted, in the framework's sense not because they failed to constitute generative ground but because they never needed to. The framework's claim isn't that systems lacking constitutional ground are inferior; it's that systems capable of generative agency only count as constitutionally mature when that agency is grounded.

Termites are an existence proof that the target geometry is reachable. They're also a clean illustration that the geometry alone isn't what MCI is after — the geometry plus the grounded agency is. Ants show this too, but termites show it more sharply because their stigmergic intelligence is so much closer to what V9's outward face describes.