Platform · when computation meets Ontology · links 3–4 of 5
ConstitutionsPlus is the platform of the Computational Ontology project: where the matrix's grammar and the project's validated systems become an executable instrument — agent-based models, system dynamics and network analysis over constitutional orders, with a governed human verdict at the end of every run.
Where it sits
The same chain of five links runs through all three sites. The matrix states it; the project derives and validates systems from it; the platform executes those systems as a coupled model and hands the result back to a human.
The grammar: being / knowing, the register as operator of identity and belonging, the four social domains, the constitution as meta-register. Also home of the book manuscript and its open companion data.
The systems derived from the grammar and validated on corpora: SOCE (the exclusion cycle) and EMERGENZA (constitutional endurance), with their codebooks, preregistration and deposits.
The instrument: the validated systems lifted to the whole constitutional order, read as a complex adaptive system, simulated under counterfactual pressure, and played — never as a source of canonical verdicts.
being · the in-emendable
inscription · identity · belonging
who controls the inscription; closure loops
CCP-compatible units; coupled functions
plural democracy — a human judgement
The platform occupies links 3 and 4 of the chain. Link 5 is behind a firewall: no simulation output, no user reading and no model ever produces a normative claim; that link is written by a person, and the platform only shows what it rests on.
For the jurist: a simulation here never tells you what the constitution ought to say; it shows which coupling of functions is about to shut into a loop, so that the judgement you make is made earlier and with the record in view.
The research problem
Constitutional theory reads a constitution as a static pyramid of sources to be interpreted, or as a legal-historical object to be described case by case. Both capture the consolidated structure and miss what the order demonstrably is in operation.
A macro-order not present in any single actor or article, which then conditions which micro-behaviours may update next.
Episodic strain that suddenly becomes self-sustaining — a closure loop that shuts, after which only multi-nodal intervention works.
Ratchets that do not relax on their own, and the same macro-outcome reached from opposite micro-configurations — structural invariance to content.
The generalisation gap. The project has validated single-domain cycles of exactly this kind. There is no integrated instrument yet that reads a constitutional system as a whole as a complex adaptive system, on a common ontological footing, with a reproducible structural notation and a governed simulation layer that moves from diagnosis (what the system is) through prospection (where it is tending) to controlled simulation (what it would do under pressure). ConstitutionsPlus is specified to be that instrument.
Four curated lines
Each line is a curated scientific script — a system with its own corpus, codebook and validation history — that the platform executes. Two exist; two are named and awaiting specification.
Five nodes (inscription · ideology · imagined community · obedience · alterity), an eleven-link catalogue, a structural signature Φ = ⟨σ, V, E⟩, eleven historical cases in five topological families, descriptive and prospective modules (SOCE-D / SOCE-P).
Constitutions as transgenerational objects; survival analysis, contagion and agent models over the Comparative Constitutions Project record (1789–), with a transgenerational index calibrated on adaptability, flexibility and durability.
Institutions carry obligations between generations; the constitution as promise and as world-artifice that outlives its founders. To be specified as its own curated line.
What in the order is unamendable-as-inscribed and what is amendable-as-interpreted; documentality as the mode in which constitutional facts acquire durability. To be specified as its own curated line.
On vocabulary: across the three sites, domain is reserved for the four social domains of the matrix (biological, economic, political, cultural). The four research lines above are therefore called lines here, so that a domain in the matrix and a line in the platform never collide.
Six constitutional functions
Within the constitutional region, the platform reads six coupled functions. They are not silos: legitimacy feeds membership, emergency rewrites sources and stresses endurance, and an exclusion closure can consolidate the "We" that re-inscribes the founding act. The object is the coupled network, not six separate analyses.
Method
The simulation core turns each line's structural signature into an executable model. The methods are complementary layers, not alternatives; a transversal layer of nonlinear mathematics reads closure as a phase transition.
Citizens, officials and factions act under adaptive rules; compliance modes, membership boundaries and institutional behaviour emerge from the bottom up. Natality enters as rule-innovation.
Stocks and flows of the order: reinforcing loops (closures), balancing loops (checks, endurance mechanisms), ratchets and hysteresis.
The coupled topology of functions and links; centrality, community structure, topological families — compared by link configuration, never by ideological content.
Bifurcations and tipping points, path dependence, sensitivity to initial conditions — the early-warning reading of a loop that is about to shut.
Master constraint. If it cannot be read from a diagnosis, it does not enter the model. Resolution is declared and governed — ordinal and qualitative where the phenomenon is qualitative; no magnitude enters that a diagnosis cannot support. Reality, through retrodiction against known outcomes, is the judge.
Two results carry over from the validated lines as platform-wide expectations: multiple realisation (the same outcome on opposite bases) and intervention asymmetry (before closure one link may suffice; after closure only simultaneous multi-nodal intervention works) — the platform's most consequential practical output.
Visual evidence layer
The research question has three clauses: can exclusion be formalised as a cycle; does the cycle have reproducible topological properties; and can its detection be made legible for democratic practice? The visual layer is the instrument of the third clause. It is what turns a structural signature into something a jurist, a monitor or a citizen can read, check and disagree with.
The layer follows evidence-presentation and visual-forensic practice — the tradition of Forensic Architecture — rather than data-storytelling. A picture here is a reconstruction: it shows what the coded record contains, in the order it was inscribed, and nothing the record does not support.
The notation has no majority baseline: the same rubrics apply whichever group is the "We" and whichever is the Other. The visual grammar is symmetric accordingly — the picture cannot be tilted by the ideological content of the case, only by its link configuration.
Every node, link and intensity drawn on screen carries its provenance: the source span, the coder or model that proposed it, the human who validated it, the codebook version, and its state — datum, interpretation, or a normative claim that the platform never draws. Legibility means being able to follow any element back to its inscription.
A reading counts as legible when practitioners in different institutions, working from the deposited pack, its visual layer and an open model stack, independently reproduce it. The picture is validated the same way the diagnosis is: by reproduction, not by approval.
The layer is not decoration on the platform; it is the platform's public face. The exploratory interface, the visualisation of coupled functions and the playable instrument are all built on it, and it serves all four lines with one grammar.
The two-tier rule
The rule is not a compliance slogan: it follows from the standpoint. The observer is an existing, plural being, and the capacity to begin something new is precisely what no simulation forecloses. That is the ontological reason the canonical verdict is always human-validated.
Anyone may run the published scripts, read cases, teach, probe the model, or play the interface. Every output is tagged with its state and its uncertainty; nothing here is a validated diagnosis.
A diagnosis counts as canonical only when produced by trained human coders under the deposited protocol, with adjudication and a versioned codebook. Model outputs may propose values; only human coders validate them.
White-box, model-agnostic. Input → process → output: information and method go in, agents curate the scientific literature by matching language hermeneutically without violating copyright, and usable results come out — reproducible by any group at near-zero compute cost. No fine-tuning, no vendor lock-in, no trackers on this site.
The long-term playable interface renders the coupled constitutional system as a serious game: perturb a function and watch emergence, closure and intervention asymmetry unfold. It is an exploratory-tier instrument by definition — a public-engagement channel and a field test of the governed templates, never a source of verdicts.
Status
Stage IV of the research trajectory (SOCE → EMERGENZA → ConstitutionsPlus → advanced simulation). The specification is written; the software is not. This page will change as the build advances, and every change is versioned in the repository.