Percolation¶
Core Idea¶
Percolation names the structural pattern in which local, often random or short-range connections, accumulated across a population of elements, suddenly produce system-spanning connectivity once the density of connections crosses a critical threshold. Below the threshold, the connected clusters remain small and local: most elements are not reachable from most others, and any signal, transport, or contagion dies out before crossing the system. Above the threshold, a giant component emerges that contains a finite fraction of all elements, and global reachability becomes a generic property of the system rather than a rare accident. The transition is sharp. Small further increases in local connection density above the threshold barely move the giant component, and small decreases below it barely matter either, but the qualitative connectivity status of the whole system flips at the threshold itself.
The commitment has three sharp parts. First, local rules, global outcome: nothing in the local connection rule — each pair joined with probability \(p\), each site occupied with probability \(p\) — knows anything about system-spanning connectivity, so the giant component's formation is purely emergent. Second, phase-transition shape: there exists a critical density \(p_c\) below which giant components are vanishingly rare and above which they are typical, and near \(p_c\) the system exhibits long-range correlations, critical fluctuations, and power-law distributions of cluster sizes. Third, connectivity is the order parameter: the relevant macrostate is not energy or density per se but connectedness — whether transport, signal, or contagion can cross the whole system or only local pockets. The canonical mathematical case is bond or site percolation on a lattice, each bond or site independently present with probability \(p\), but the same structural behavior extends to random graphs, configuration models, spatial random networks, and continuum percolation, with substrate-dependent values of \(p_c\) but the identical qualitative phase transition. This is what makes percolation a prime rather than a fact about porous rock: the connectivity-density relation is stated in pure relational terms and carries by literal mathematics wherever elements connect.
How would you explain it like I'm…
Stones Across The Pond
Suddenly It Connects
Emergent Spanning Connectivity
Structural Signature¶
a population of elements — a local connection rule applied independently across pairs or sites — a control parameter setting connection density — connectivity as the order parameter — a critical threshold separating local clusters from a system-spanning giant component — a sharp phase-transition shape with critical fluctuations near the threshold
The pattern is present when each of the following holds:
- An element population. A set of components that may become connected — lattice sites, network nodes, individuals, trees, banks.
- A local connection rule. A short-range or random rule joining elements, applied independently and blind to any global outcome. Each bond or site is present with some probability.
- A control parameter. A scalar — occupation probability, mean degree, reproduction number, fuel density — that sets the density of local connections and drives the system.
- Connectivity as order parameter. The relevant macrostate is not energy or density but connectedness: whether transport, signal, or contagion can cross the whole system or only local pockets, measured as the largest component's fraction of the whole.
- A critical threshold. A value \(p_c\) below which connected clusters stay small and local and above which a giant component spanning a finite fraction of elements becomes typical.
- A phase-transition invariant. The flip is sharp: near \(p_c\) the system shows long-range correlations, diverging correlation length, and power-law cluster-size distributions, so the global status can change abruptly relative to smooth change in the control parameter.
These compose into the signature move — local rules accumulate until connectivity crosses a threshold — and the mathematics is substrate-independent, carrying by literal equation across substrates within a universality class.
What It Is Not¶
- Not contagion. See
contagion. Contagion is the dynamic process of something spreading along connections over time. Percolation is the static connectivity condition that determines whether such spread can span the system at all. Percolation sets the stage (is a giant component present?); contagion is the play that runs on it. - Not propagation. See
propagation. Propagation is the traveling of a signal or influence outward through a medium. Percolation is silent about the traveling; it answers only whether a connected path spanning the system exists. A medium can percolate yet carry no propagating signal, and vice versa. - Not a generic tipping point. See
tipping_points_or_phase_transitions. That prime covers any abrupt regime change in a control parameter. Percolation is the specific phase transition whose order parameter is connectivity — the emergence of a system-spanning giant component — with its characteristic critical exponents and universality classes. - Not criticality as a tuned state. See
criticality. Criticality names a system poised at a critical point with scale-free fluctuations. Percolation is the connectivity transition that defines one such critical point; a percolating system is critical only at \(p_c\), and is generically sub- or super-critical away from it. - Not critical mass. See
critical_mass. Critical mass is the minimum quantity of participants or material needed to sustain a self-propelling process. Percolation's threshold is about the density of connections producing global reachability, not the count of active elements sustaining a reaction; the two thresholds answer different questions. - Common misclassification. Treating a smooth-looking aggregation as percolation, or reading local connection density as global connectivity. The catch: percolation requires a sharp connectivity threshold. If the system has no \(p_c\) at which a giant component abruptly appears — if connectivity grows smoothly with the parameter — it is aggregation or diffusion, not percolation.
Broad Use¶
- Materials science. Gel formation and the sol-gel transition, conductivity of disordered conductor-insulator mixtures, porosity and fluid flow through porous media, polymer gelation.
- Epidemiology. An epidemic reaches a finite fraction of a population only when the effective reproduction number crosses a threshold; herd-immunity thresholds and the phase transitions of compartmental models are percolation phenomena on contact networks.
- Wildfire. Forest-fire models on a lattice show percolation transitions in fire spread as fuel density crosses threshold; the spanning combustible cluster is a giant component.
- Network science. The connectivity of large networks under random or targeted node removal; the internet's robustness to random failure but fragility to targeted attack is a percolation problem on a heterogeneous-degree network.
- Collective behavior. A social movement reaches systemic scale only when local-recruitment density crosses a threshold; threshold models of collective action are percolation models.
- Neuroscience. Neuronal avalanches, the critical-brain hypothesis, and the seizure threshold as a percolation transition in recurrent excitatory networks.
- Computer science and infrastructure. Random-satisfiability transitions, routing connectivity in ad-hoc wireless networks, and cascade-failure thresholds in power grids and inter-bank networks.
- Ecology. Habitat connectivity under fragmentation, and the critical fraction of preserved habitat below which a species loses continental movement corridors.
Clarity¶
Naming the percolation transition makes legible a class of phenomena that linear or smooth-threshold reasoning systematically misses. It explains why "almost there" is not there: a system just below \(p_c\) looks locally identical to one just above it — the same density of short-range connections — yet their global connectivity is qualitatively different, and no purely local measurement can distinguish them. It explains why interventions can be wildly leveraged near the threshold: small changes in local connection density near \(p_c\) produce large global changes, so the same effort buys almost nothing far from the threshold and almost everything near it. It explains why network heterogeneity changes the calculus: on highly heterogeneous networks the threshold depends on the degree distribution and can be vanishingly small, meaning even sparse contact suffices for system-spanning connectivity, which is why targeting hubs is efficient and random intervention is wasteful on such networks. And it explains why the transition seems to come out of nowhere: because percolation transitions are genuine critical phenomena, with a correlation length that diverges at \(p_c\), the global flip can appear abrupt relative to the slow, smooth change in the upstream control parameter that drives it. The clarifying move is to recognize that the relevant variable is connectivity, not the surface quantity, and that connectivity has a threshold.
Manages Complexity¶
The pattern collapses a large family of disparate-looking phenomena — gel formation, epidemic spread, forest fires, network robustness, social cascades, conductivity, brain dynamics — into one structural question: what is the local connectivity, where is the threshold, and which side of the threshold is the system on? This compression does real analytic work, because it licenses method transport: the renormalization-group, mean-field, scaling-theory, and simulation methods developed for one percolation problem transfer to the others without re-derivation, since they all describe the same underlying transition. The compression also yields a sharp diagnostic test. If connectivity exhibits a sharp threshold in the relevant control parameter, the phenomenon is percolation-like, and many systems first modeled with smooth-threshold or linear-aggregation reasoning turn out, on inspection, to be percolation transitions in disguise. Recognizing this changes the entire intervention vocabulary: instead of acting on the system-wide order parameter directly, the analyst acts on local connection density relative to the threshold. The complexity that overwhelms a phenomenon-by-phenomenon treatment — each gel, fire, and epidemic studied on its own terms — is managed by recognizing that all of them are the same transition over connectivity, governed by the same critical structure.
Abstract Reasoning¶
Percolation licenses reasoning about critical thresholds in connectivity as distinct from thresholds in energy, density, or other order parameters, which keeps the analyst's attention on the right macrostate. It licenses reasoning about hub versus random targeting as intervention strategies with sharply different efficacy on homogeneous versus heterogeneous networks. It licenses reasoning about the robustness-fragility trade-off: a network can be robust to random failure yet fragile to targeted attack, and the difference is a percolation property of the degree distribution rather than a contingent fact about any particular network. It licenses reasoning in terms of universality classes: percolation transitions share critical exponents across substrates within a class defined by dimensionality and symmetry, which allows quantitative predictions to transfer between substrates that share no surface vocabulary whatsoever. It licenses partitioning system behavior into sub-critical, critical, and super-critical regimes, three qualitatively different states with different intervention affordances — far from threshold, near it, and across it. It licenses distinguishing spatial from network percolation, which give different thresholds and therefore different intervention calculus depending on whether the relevant adjacency is geometric or topological. And it licenses reading long-range correlations near the threshold — the power-law distributions of cluster sizes and avalanche durations — as a recognizable signature that a system is sitting close to criticality.
Knowledge Transfer¶
Percolation transfers across domains with unusually strong precision, and the reason is structural rather than analogical: the underlying mathematics is substrate-independent, so the same equations literally describe the new domain rather than merely resembling its behavior. The scaling theory that predicts conductivity onset in a conductor-insulator mixture predicts the epidemic threshold on a contact network, because both are percolation on a disordered medium and share a universality class. The Erdős-Rényi giant-component transition, carried over to internet topology, produces the robust-to-random-fragile-to-targeted result directly. Forest-fire percolation models carry into wildfire risk management as the prescription to reduce fuel density near the threshold, and the field's intervention vocabulary — fuel breaks, prescribed burns, defensible space — is already percolation-aware, targeting local connection density rather than the spanning cluster itself. Contagion-on-networks percolation carries into vaccination strategy as the result that immunizing high-degree nodes raises the effective threshold faster than random immunization on a heterogeneous network. The same framework, applied to inter-bank exposure networks, identifies systemically important institutions whose failure would cross a cascade-failure percolation threshold. Applied to landscape ecology, it predicts the critical habitat fraction below which a species loses its movement corridors. Applied to neural dynamics, it carries the conductivity-threshold mathematics to neural-avalanche predictions under the critical-brain conjecture.
The transferable structural vocabulary is itself the carrier: threshold, giant component, universality class, order parameter, critical exponent, and sub-/super-critical regime travel across substrates and provide a shared analytic language in which a result proven in one field is immediately stateable in another. The role mappings carry cleanly: the element population is the lattice site, the network node, the individual, the tree, the bank; the local connection rule is the bond probability, the transmission probability, the inter-bank exposure, the spatial adjacency; the control parameter is the occupation probability, the reproduction number, the fuel density, the mean degree; the order parameter is in every case the size of the largest connected component as a fraction of the whole. Because the structure and its mathematics are shared, the diagnostic and intervention questions transfer as one recipe: identify the connectivity, locate the threshold, determine which side the system is on, and act on local connection density rather than on the global order parameter. A materials scientist's intuition about pushing a mixture below the conductivity threshold and an epidemiologist's intuition about pushing transmission below the herd-immunity threshold are, structurally, the same move on the same object. The static case covers cleanly; where real systems have dynamic connectivity that forms and dissolves, the prime supplies the conceptual scaffold onto which the temporal-network and adaptive-percolation extensions attach.
Examples¶
Formal/abstract¶
Bond percolation on a two-dimensional square lattice is the canonical worked instance, and it makes every element of the signature concrete and computable. The element population is the set of lattice sites — picture a large grid. The local connection rule opens each edge (bond) between neighboring sites independently with probability \(p\) and leaves it closed with probability \(1-p\); the rule is purely local and blind to any global outcome. The control parameter is \(p\) itself. The order parameter is connectivity: the fraction of sites belonging to the largest connected cluster. As \(p\) is raised from zero, clusters of open bonds grow, but below the critical threshold \(p_c\) (for this lattice, one-half) every cluster remains finite and local — no path of open bonds spans the grid, so a fluid poured in at the top cannot reach the bottom. At \(p_c\) the phase-transition invariant manifests: the correlation length diverges, the cluster-size distribution becomes a power law, and a spanning cluster — the giant component — appears for the first time, abruptly conferring system-wide connectivity. The intervention the formalism enables is sharp leverage near criticality: a small increase in \(p\) far below \(p_c\) does almost nothing to global connectivity, while the same increase right at \(p_c\) flips the system from disconnected to spanning. Critically, the renormalization-group and scaling methods that compute \(p_c\) and the critical exponents here are not lattice-specific tricks; they describe a universality class and therefore predict behavior in any substrate sharing the same dimensionality and symmetry.
Mapped back: The lattice sites are the element population, the bond probability \(p\) is the control parameter, the largest-cluster fraction is the order parameter, and the appearance of a spanning cluster at \(p_c\) is the critical-threshold flip from local clusters to a system-spanning giant component.
Applied/industry¶
Epidemic spread on a contact network is percolation operating in public health, and the mapping is mathematical rather than analogical — the same scaling theory applies. The element population is the set of individuals; the local connection rule is transmission, where each contact between an infectious and a susceptible person transmits the pathogen with some probability, applied locally and independently along each edge. The control parameter is the effective reproduction number, set by transmissibility, contact rate, and the fraction still susceptible. The order parameter is the fraction of the population reachable by the outbreak — the giant component of the transmission network. Below the critical threshold (reproduction number one, the herd-immunity boundary) outbreaks fizzle into small local clusters that die before spanning the population; above it, a giant outbreak reaching a finite fraction of everyone becomes typical. The frame's distinctive payoff is that network heterogeneity changes the calculus: on a contact network with a few high-degree hubs (super-spreaders), the threshold can be vanishingly small, so even sparse contact suffices for system-spanning spread. This directly prescribes the intervention — immunize or isolate the high-degree nodes, which raises the effective threshold far faster than the same number of random immunizations. The identical percolation structure, applied to inter-bank exposure networks, identifies systemically important institutions whose failure would push the network across a cascade-failure threshold, and applied to wildfire it prescribes reducing fuel density (the local connection parameter) below the spanning threshold via fuel breaks and prescribed burns. A public-health official targeting super-spreaders and a forester cutting fuel breaks are, structurally, making the same move: push local connection density below \(p_c\) so the giant component cannot form.
Mapped back: Individuals are the element population, transmission probability is the local connection rule, the reproduction number is the control parameter, and the herd-immunity boundary is the percolation threshold separating fizzling local clusters from a population-spanning epidemic — with hub-targeting as the heterogeneous-network intervention the frame uniquely prescribes.
Structural Tensions¶
T1 — Local Rule versus Global Outcome (Scale Gap). Percolation's defining move is that nothing in the local connection rule knows about system-spanning connectivity — the giant component is purely emergent. This is its power and its trap: no local measurement can tell a system just below threshold from one just above, because they are locally identical. The failure mode is inferring global connectivity from local density, declaring a system safe or spanning based on the connection rule alone. The diagnostic is to measure the order parameter directly — the largest component's fraction — rather than the control parameter: connectivity is a global property that local sampling cannot resolve, so "the connections look sparse" is not evidence the system is sub-critical.
T2 — Smooth Control Parameter versus Sharp Transition (Temporal Surprise). The control parameter changes smoothly while the global connectivity status flips abruptly at \(p_c\), so the transition appears to come from nowhere relative to its slow upstream driver. The failure mode is linear extrapolation: projecting the gradual trend in fuel density, transmissibility, or mean degree forward and being blindsided when the system spans all at once. The diagnostic is to locate \(p_c\) in advance and track distance-to-threshold rather than the raw trend — and to watch for the critical signatures (diverging correlation length, power-law cluster sizes) that announce proximity to the flip before it happens, since the smoothness of the driver gives no warning on its own.
T3 — Far-From-Threshold versus Near-Threshold (Leverage Asymmetry). Intervention efficacy is wildly non-uniform: the same change in local connection density buys almost nothing far from \(p_c\) and almost everything near it. The failure mode is uniform-effort reasoning — budgeting intervention as if each unit of connectivity reduction yields constant return, over-investing in a far-from-threshold system where it is wasted or under-investing near the threshold where it is decisive. The diagnostic is to estimate where the system sits relative to \(p_c\) before sizing the intervention: in the sub-critical and super-critical interiors, marginal effort is nearly inert, so leverage is concentrated in a narrow band around the threshold and effort should be concentrated there too.
T4 — Homogeneous versus Heterogeneous Connectivity (Threshold Dependence). The clean lattice picture assumes roughly uniform connectivity, but on heterogeneous networks the threshold depends on the degree distribution and can fall to nearly zero, so even sparse contact spans. The failure mode is importing a homogeneous threshold intuition into a hub-dominated network — assuming sparse connection is safe, or that random intervention is efficient, when a few high-degree nodes carry the giant component. The diagnostic is to characterize the degree distribution before trusting any threshold: where it is heavy-tailed, target hubs (immunize super-spreaders, identify systemically important banks), because random removal barely moves \(p_c\) while hub removal moves it sharply.
T5 — Random Robustness versus Targeted Fragility (Sign of the Perturbation). A percolating network's response to node removal is double-signed: robust to random failure yet fragile to targeted attack, a property of the degree distribution rather than the network's apparent solidity. The failure mode is summarizing resilience with a single number — "the system tolerated 30% loss" — without specifying whether the loss was random or adversarial, since the same network can shrug off random failure and shatter under hub-targeted attack. The diagnostic is to ask who or what is removing nodes: stress-test against the worst-case removal order, not the average one, because robustness-to-random offers no guarantee against an adversary who removes hubs first.
T6 — Static Connectivity versus Dynamic Rewiring (the Frozen-Graph Assumption). Classical percolation assumes a fixed graph whose bonds are set once; many real systems rewire continuously — contacts form and dissolve, fuel regrows, exposures shift — so connectivity is a moving target rather than a frozen draw. The failure mode is reading a single-snapshot percolation analysis as a verdict on a temporally dynamic system, concluding it is sub-critical because it is disconnected now while accumulating connections push it across threshold over time. The diagnostic is to ask whether the graph is static on the timescale of the spreading process: where connectivity evolves faster than or comparably to the contagion, the static \(p_c\) is the wrong object and temporal-network or adaptive-percolation extensions are required.
Structural–Framed Character¶
Percolation sits at the structural pole of the structural–framed spectrum — a structural prime with a 0.0 aggregate, and one of the cleanest in the catalog. It is a pure mathematical pattern: local, often random connections accumulated across a population suddenly yield system-spanning connectivity once their density crosses a critical threshold. The same equations describe every substrate, and every diagnostic reads zero.
The pattern carries no home vocabulary that must travel with it; it carries the mathematics itself. Porous rock, epidemic contact networks, wildfire fuel beds, random graphs, neural connectivity, and the satisfiability transition are not analogies tied by a shared word but instances of one theory, each told in its own substrate's terms while obeying the same scaling laws near \(p_c\) — so vocab_travels is zero. It carries no inherent approval or disapproval: a giant component is neither good (a connected power grid) nor bad (a system-spanning epidemic) until you specify what is percolating, so evaluative_weight is zero. Its origin is formal — statistical physics on a lattice, with \(p_c\) as a mathematical object — owing nothing to any human institution, so institutional_origin is zero. It runs indifferently in physical, biological, and abstract substrates with no human practice required to instantiate it: bond percolation on a crystal lattice and connectivity in a configuration-model graph are the same phenomenon, so human_practice_bound is zero. And invoking the prime recognizes a connectivity phase transition already present in the system rather than importing any interpretive frame — the threshold exists in the mathematics whether or not anyone names it "percolation," so import_vs_recognize is zero. Every diagnostic points the same way; the structural label with its 0.0 aggregate is exactly right.
Substrate Independence¶
Percolation is about as substrate-independent as a prime can be — composite 5 / 5 on the substrate-independence scale, and one of the cleanest in the catalog. Its breadth is maximal: the connectivity-phase-transition structure recurs in materials science (gelation, conductivity, porous flow), epidemiology, wildfire, network robustness, collective behavior, neuroscience, random satisfiability, and landscape ecology. Its abstraction is total: the signature — an element population, a local connection rule, a control parameter, connectivity as order parameter, a critical threshold, a sharp phase-transition shape — is stated in pure mathematics with no domain content, so the connectivity-density relation carries by literal equation, not analogy. Transfer is exceptionally concrete and documented: substrates within a universality class share critical exponents and scaling laws, so the renormalization-group and mean-field methods derived for one problem predict another without re-derivation, and the transferable vocabulary itself (threshold, giant component, universality class, order parameter, sub-/super-critical regime) is the carrier — a materials scientist pushing a mixture below the conductivity threshold and an epidemiologist pushing transmission below the herd-immunity threshold make the identical move on the same object. Maximal breadth, maximal abstraction, and mathematically guaranteed transfer all line up, which makes it a canonical 5.
- Composite substrate independence — 5 / 5
- Domain breadth — 5 / 5
- Structural abstraction — 5 / 5
- Transfer evidence — 5 / 5
Relationships to Other Primes¶
Parents (1) — more general patterns this builds on
-
Percolation is a kind of Tipping Points (or Phase Transitions)
The file: percolation is 'the SPECIFIC phase transition whose order parameter is connectivity' — the emergence of a system-spanning giant component, with its own critical exponents and universality classes. A species of the generic tipping-point/phase-transition.
Path to root: Percolation → Tipping Points (or Phase Transitions) → State and State Transition
Neighborhood in Abstraction Space¶
Percolation sits among the more crowded primes in the catalog (30th percentile for distinctiveness): several abstractions describe nearly the same structure, so a description that fits it will tend to fit its neighbors too — transporting it usually means disambiguating within this family rather than landing on it exactly.
Family — Graphs, Networks & Connectivity (12 primes)
Nearest neighbors
- Percolation Threshold — 0.87
- Critical Mass — 0.72
- Contagion — 0.71
- Phase Separation — 0.71
- Permeability — 0.70
Computed from structural-signature embeddings · 2026-06-14
Not to Be Confused With¶
The deepest and most consequential confusion is with contagion, the embedding-nearest neighbor at similarity 0.90 — and the two are so entangled that epidemiology runs both at once. The distinction is between a condition and a process. Percolation is a statement about the structure of connectivity: given a population and a local connection rule, does a system-spanning giant component exist? It is, at root, a static question answered by the order parameter — the largest component's fractional size — and it is silent about time, rate, and dynamics. Contagion is the process that spreads something — a pathogen, a belief, a default — from element to connected element over time. The relationship is that percolation determines the reachable set within which contagion can operate: an epidemic confined to a sub-critical contact network fizzles not because transmission is slow but because no spanning cluster exists to carry it, whereas above \(p_c\) the same transmission process reaches a finite fraction of everyone. What percolation captures that contagion does not is the threshold geometry of reachability — the sharp flip in whether global spread is even possible. What contagion captures that percolation does not is the temporal unfolding — the rate, the curve, the timing of arrival at each node. A practitioner needs both: percolation says whether a giant outbreak is on the table; contagion says how fast it arrives and crests if it is.
A second confusion is with tipping_points_or_phase_transitions, which percolation is a special case of, and which a reader may reach for as the general label. Tipping points name any abrupt qualitative regime change driven by smooth movement of a control parameter — a lake flipping from clear to turbid, a market from boom to crash, ice to water. Percolation is the specific such transition whose order parameter is connectivity and whose machinery is the appearance of a giant component on a graph. The added specificity is the whole value: because percolation is a particular universality class, it brings exact critical exponents, the robustness-to-random-fragility-to-targeted result, the hub-targeting intervention calculus, and the substrate-independent scaling mathematics — none of which follow from the bare fact of an abrupt transition. Calling a percolation transition merely a "tipping point" is correct but loses the predictive apparatus; conversely, importing percolation's hub-targeting intuitions into a tipping point that is not a connectivity transition (a bistable chemical reaction, say) is an error, because there is no graph and no degree distribution to target.
A third, subtler confusion is with criticality as a system property. Criticality names a state of being poised at a critical point, exhibiting scale-free fluctuations, long-range correlations, and power-law event-size distributions — the signature of self-organized critical systems like sandpiles or the critical-brain hypothesis. Percolation generates one canonical critical point — exactly at \(p_c\), the percolating system shows all the critical signatures — but a percolation system is critical only at that one value of the control parameter and is generically off-critical (sub- or super-critical) elsewhere. Criticality-as-a-state often refers to systems that self-tune to their critical point and sit there persistently; ordinary percolation has no such self-tuning and must be driven to \(p_c\) externally. The shared vocabulary — correlation length, power laws, exponents — masks the difference between a transition you pass through (percolation) and a state a system maintains (self-organized criticality). Reading the power-law cluster sizes near \(p_c\) as evidence of self-organized criticality, when the system was simply tuned to threshold, is the characteristic error.
For a practitioner, sorting these apart routes the analysis correctly. Ask first the percolation question — is there a spanning cluster, and where is \(p_c\)? — because it bounds what is possible. Then ask the contagion question — how fast and along what curve does spread actually unfold on the reachable set? Use the generic tipping-point frame only when no connectivity graph underlies the transition, and reserve the language of sustained criticality for systems that genuinely self-tune to their critical point rather than ones merely parked at threshold.
Solution Archetypes¶
No catalogued solution archetypes reference this prime yet.