Integration Capacity Analysis · The Great Homecoming research programme · explainer · June 2026
Why do some systems hold together under pressure while others — often richer, larger and better equipped — come apart? A country, an institution, a company: each faces shocks continuously, and the same shock that one absorbs will shatter its neighbour. The difference is not luck, and it is rarely resources. It is an internal state, already present before the shock arrives.
We frame that state as a gap between two capacities. Interaction capacity is everything that lets a system do more, faster: connectivity, technology, finance, information, programmes, interfaces. Integration capacity is what lets the system cohere around all that doing — whether its parts still pull in a shared direction, whether bad news still travels upward intact, whether stated purpose still describes actual behaviour, whether the bonds that hold people to the work are being renewed or merely consumed. Societies and institutions today are running 2026-level interaction on integration capacity that has stopped keeping pace, and the gap shows up everywhere the same way: institutions that keep functioning but stop meaning, rising polarisation, money spent without traction.
The Great Homecoming is an independent research programme on that gap: why systems cohere or fragment, how the difference can be measured, and what durably coherent systems are built on. The aim is constructive — not cataloguing decline, but understanding coherence well enough to read it in advance, where the reading can be checked.
It helps to separate two layers that are easy to confuse.
The model is the set of premises about what a system is, and it does not change. The interaction-versus-integration gap above sits inside a fuller picture: a system is held by three capacities at once — integration (pulling its own diversity into coherence: holding a direction, making the parts act as one), interaction (meeting, taking in and responding, inside and out), and bonding (holding: forming durable structure and producing what endures). From these follow a few fixed rules: a system is only as sound as its weakest load-bearing capacity, never an average; coherence depends on a working correction loop, so a system still reachable by its own feedback can recover while one whose loop is cut cannot; and a damaged loop is repaired from a healthy level above, not from inside.
The metrics are how we currently read those capacities — the named diagnostics that appear in a report. Unlike the model, they are meant to improve: as the instrument is tested and extended, the way each capacity is measured will sharpen. When a reading changes, that is the measurement getting better, not the model moving. And today's report deliberately surfaces only part of what the model can read — it covers integration and correction most fully, with a deeper diagnostic layer and three further measures still in development: a system's external interaction (how it actually couples with its environment), its output (the real effect it imposes on the world), and its legacy (what it transmits to what comes next). The development roadmap is open about which of these is built, which is being built, and which is still on the list.
The programme’s output is not commentary. It is an instrument — Integration Capacity Analysis — in two parts.
The first is a measurement instrument that reads a system’s internal state from observable material — what it declares as its purpose, what it actually does, how it responds when correction is attempted — and converts that material into a structured reading under written coding rules.
The second is a simulation engine that takes such a state and runs its dynamics forward. It does not produce a date or a single number. It produces a pattern: which commitments drift, which bonds hollow, where polarisation concentrates, and in what sequence.
Two design properties distinguish the build. First, scoring is derived, not assigned: readings come from the system’s own declared and enacted structure rather than from the analyst’s judgment, and at every step of a simulation the engine cross-checks its structural state against that layer — a per-step self-audit that catches the model drifting away from the system it claims to describe. Second, the same dynamics have been run, without per-case tuning, on historically opposite cases: one system that consolidated under stress, one that hollowed out and collapsed. Reproducing both outcomes from one set of rules is the transferability test most frameworks in this field never take.
Four things — none of them a prophecy.
Shock-readiness. Whether the system, as it stands today, would absorb its next shock or shatter under it. The shock itself cannot be predicted; readiness can be read, because it is decided entirely before the event. The event only reveals it.
The two clocks. Every organisation runs on a solvency clock, which counts the stores — cash, credit, brand, legitimacy — and a coherence clock, which counts whether the thing still works as one purposeful system. The clocks can diverge, and when they do, the solvency clock always reads later: stored surplus keeps a hollow organisation looking alive. The gap between the clocks is the earliest warning available, often visible years before the balance sheet notices anything.
Where effort evaporates. Units of money, work and attention that move dashboards without moving outcomes. A hollowed bond consumes input without converting it; a reading shows where in the system that conversion is failing — before the next budget is committed to it.
The arc it is on — rising as well as falling. A reading is not only a decline-detector. It places a system on its full cycle — growing, consolidating, declining, or reorganising after a fall — and reads the quality of a rise as sharply as a fall: whether growth is building genuine new coherence or merely expanding, and whether a system is renewing itself rather than only surviving. Which of these a system is in is often the most consequential call, because expansion and integration, like crisis and collapse, can look alike from outside and be opposite underneath.
This field has produced many frameworks that explain everything after the fact. Ours runs under a discipline designed to prevent exactly that.
Every forward analysis is dated when it is made and tracked on a fixed register, which has run since June 2026. We hold ourselves to what we wrote, not to what we remember having meant. Results are reported in both directions: where early out-of-sample tests on third-party historical and ecological datasets did not beat simple baselines, those results stand and are stated plainly. They were useful — they marked the boundary of the instrument’s lane (systems read in depth and in motion, not sparse aggregates read from a distance) and fed directly into the current design. A programme that reports only its hits cannot be trusted.
The instrument is under live forward test, and we describe it exactly that way until the register adjudicates; the first windows close in 2027–2028. Until then, nothing here is described as validated, because nothing yet is.
The framework is stated entirely in structural terms that stand on their own. It was built from first principles about how systems cohere and fragment, and tested against historical and contemporary cases — not derived from any single tradition. Every claim is required to survive translation out of any particular vocabulary, and a claim that cannot is dropped.
The entry engagement is a bounded structural assessment: a written report on one system, with the reasoning shown and every reading paired with what would prove it wrong.
The Great Homecoming is an independent research programme on why systems cohere or fragment. Contact: Wim Van Laere.