WHAT THE TRIAL ESTABLISHES
One recovery project.
Different reasoning contracts.
Toy scenario
Each method reveals, derives or rules out something specific. A common scenario makes their differences easier to inspect.
Read the results at the right scale
The recovery example is inspired by the collapsed footbridge at Gimårasten on the Gimån, near Stavre. Nothing here diagnoses the actual collapse, establishes its cause or proposes engineering work. Per Sonnvik’s walk account, dated 11 May 2024, provides background about the crossing; no construction assumptions below are inferred from it.
The new illustration is an imagined autumn scene generated with OpenAI, using the locality’s broad river, low wooded banks, long broken crossing and adjacent railway as cues. It is neither a photograph nor a reconstruction of the actual structure. Sonnvik’s photographs remain his copyright and are linked rather than reproduced. For another view of the river, Andreaze’s 2008 Gimån photograph is released into the public domain; its precise reach is not established here.
Initial boundary: the crossing is closed, damaged material has been made safe and removed, and the supplied condition evidence supports the declared abutment and installation assumptions. Placement permission is a separate prerequisite. Low-water access and lifting capacity are shared controls. The optional decision survey concerns extra logistical readiness; it cannot satisfy missing structural-condition evidence or permission.
The examples do not form one integrated engineering model. Logical moves, abstract schedule time, calendar windows, full-path budget and expected decision costs have different meanings. Curing is explicit in the state model, aggregated into placement duration in the scheduler, and omitted from the small semantic action catalogue. No figure is a real estimate or verified construction limit.
Model correctness, research contribution, reader understanding and usefulness in practice are separate questions. Automated checks do not establish human learning or bridge safety. “Best for” judgements on the comparison page refer to evidenced uses, not a universal numerical ranking.
110 · Process to plan
Seven explicit facts feed a projection. Five constraints are retained; two narrative facts are omitted with recorded reasons. Two process descriptions yield the same four-part summary, but only one contains bank-based placement when the low-water work window disappears. Permission and condition evidence remain separate supported or missing claims.
The counterexample demonstrates information loss under a projection. It is not proof of functoriality or an inverse reconstruction algorithm. Having a plan does not grant permission.
120 · Dynamic project states
The state records temporary supports, two deck-section stages (absent, curing, ready, accepted), clearance and reopening. Explicit update rules generate valid moves; breadth-first exploration finds a shortest completing continuation. At the default start, one lifting resource needs nine logical moves; two allow a joint placement and need eight. These are not durations. Reopening follows acceptance of both sections and removal of supports.
Two states can both show “two sections installed” while only one allows inspection. State sufficiency belongs to the declared update law. Missing access blocks the selected low-water method; changing to bank-based placement changes the mechanism. This is finite dynamics, not a sheaf-gluing result.
140 · Local to global
Three accounts constrain month and method. The access review also records independent permission and condition-evidence domains. The solver intersects every relevant domain and enumerates joint assignments. The pairwise trap uses {July, August}, {August, September} and {July, September}: every pair intersects, but all three do not.
These finite constraint sets are not declared sheaf sections with a gluing theorem. Missing evidence is not agreement; fixing a date does not repair missing permission. A real review must also distinguish units, versions, scope and authority.
150 · Mechanisms to possible plans
Nine once-only transitions model supports, preparation of A/B, placement of A/B, inspection of A/B, support removal and reopening. All 52 complete atomic execution orders are enumerated. Placement reads permission, condition evidence and method access; a lifting resource is reusable. Acceptance records are retained through support removal and reopening: the signatures carry them forward as evidence, and the scheduler reads completed prerequisites.
A separate scheduler reserves lifting capacity throughout placement and credits completion at its end. Its placement duration includes the settling/curing allowance. With one resource there are two earliest-start schedules finishing in nine abstract units; with two resources there is one finishing in seven. These priority schedules do not enumerate deliberate waits. Independent replay checks durations, precedence, coverage and aggregate capacity.
Typed signatures do not implement a general Petri-to-SMC functor. WBS grouping is separate from causal order and does not recover every dependency.
160 · Process wiring
The diagram checks four nominal interfaces. A mounted section cannot satisfy an accepted-section input. A separate simultaneous-use contract adds lifting demand and checks capacity, both evidence prerequisites and low-water access. Repairing a type mismatch need not repair a resource failure.
This is a small composition check, not a full wiring-diagram operad. Matching a port does not certify structural or hydraulic performance. The handover output means the two declared acceptance interfaces match; the full clearance and reopening process appears in the other examples.
Composing work packages (the broader native investigation) · Comparison implementation
170 · Shared constraints and review
The model enumerates two placements in August/September under accepted access, evidence and capacity. Each named month represents one indivisible placement slot; sequential placements within a month are not modelled. Two placements in the same slot therefore demand two lifting resources. With one resource and both months, two opposite orders remain. Restricting access to one month removes them unless capacity changes.
A reviewer proposes an access change; accepting it updates only this local demonstration and its reproducible URL. Proposed and accepted values remain distinct. This is neither live collaboration nor approval in a project system. Calendar assignments are not physical differential dynamics.
Composing open systems · Finite shared-constraint adaptation. The native investigation has a broader open-system scope than this two-window example.
180 · Feasible decks and their construction paths
Three deck configurations combine with two placement methods. Feasibility checks a supplied load-rating floor of 10, lifting-load compatibility, access, both evidence prerequisites, capacity, temporary supports and full-path budget. Rating and mass units are model attributes; no structure is calculated.
The one-piece deck with low-water placement costs 68 units before temporary works and 78 with them, exceeding the default budget of 70. Two standard sections remain feasible at 64 units with low-water placement or 60 with bank-based placement. The first has shorter duration and site occupation. Light sections fail the rating floor even when cheap. The nondominated set therefore expresses trade-offs without inventing a universal weighting.
These are six finite witnesses. They do not search the full design space or embed the separate co-design feedback solver. One-piece versus two-section configurations are local design alternatives; the other examples hold two standard sections fixed.
210 · Meaning, goals and required work
The desired end is accepted sections, clear temporary works and an open crossing. A local typed vocabulary compiles explicit method rules into planning actions. Setup destroys “clear”; removal restores it after acceptance; reopening requires clearance and acceptance. Empty recovery scope generates no work. A two-section method requires seven actions; one section requires five.
Necessity is tested by removing each function and searching again. It is necessity within this catalogue, not all possible repair techniques. Changing product/process packaging preserves action identities. The semantic route searches fact sets; a separately declared parameter-free STRIPS domain is parsed into an independent bit-mask search. Both receive the same initial facts and goal, and parity is computed.
This is not full RDF/OWL reasoning or evidence of superior planning. The supplied-WBS validator checks a declared single-parent acyclic hierarchy and an output for every leaf, with summaries exempt.
220 · Evidence before commitment
Three decision opportunities, a fixed hidden favourable/adverse site-readiness regime and one optional survey define a finite belief model. A controlled low-water placement attempt succeeds with conditional probability 0.9/0.2 and costs 10 units. Non-completion safely stops the attempt, consumes an opportunity and updates belief. Bank-based placement completes delivery for 27 units. Continuing closure through the horizon costs 50. These are logistics outcomes, not structural-failure probabilities or the whole recovery budget.
Required structural-condition evidence, permission and lifting capacity gate either method. The optional survey concerns additional logistical readiness; it cannot supply missing prerequisites. Dynamic programming compares low-water attempts, bank placement, waiting, continued closure and a survey. Waiting provides no new signal. The hidden regime does not evolve; river levels are not simulated.
Survey signals and observed non-completion update the posterior. Informative early evidence can change the chosen action; noise can cost money without changing belief; late evidence cannot help before the horizon. A 0.6 success-probability threshold policy uses the same costs and observation rules. Independent tests enumerate deterministic policy costs and compare their minimum with the calculated optimum. Optimality holds only inside this finite contract.
How to interpret the comparison
The shared disturbance removes a low-water work window. Some methods include a bank-based fallback; other rows explicitly select low-water placement. Their different outcomes expose declared scope and do not rank techniques by a common score. A stronger comparison asks the same question with controlled assumptions and a suitable simpler baseline.
The research programmes remain independent. Update a trial when its source method, assumptions or evidence changes; preserve useful counterexamples and interactions. Model tests, source-backed explanations and human usefulness remain separate evidence claims.
Return to the working examples → · Compare the wider collection →
One contained construction from each remaining enquiry
The following twelve adaptations were executed on 21 October 2026. They preserve the native question at a deliberately small scale. The rules, records and quantities are authored. A model calculation is separate from a reader study, independent reuse, real-world validation or commercial acceptance. Where an ordinary method receives the same information, its equal answer is retained.
230 · Risks and tasks
Execute deliveries, placements and an outside maintenance passage under one lane and crane account; compare placement target outcomes.
Two deliveries, two placements and at most one maintenance passage. Shared access, permit, condition and capacity apply. Budget is ignored. Acceptance and reopening are outside this timing model.
Native source: the farm-track model reserves and releases one lane for concrete delivery and cattle passage, with separate task, intention and threat interpretations.
Adaptation: two bridge deliveries share one bank-haul lane with a maintenance crew; placement separately borrows lifting resources. Starts and finishes are enumerated, and timed priority witnesses use durations 1, 2 and 3.
What this establishes: The shared-lane event model retains both actors’ processes; outside-first passage delays placement while the task-only account misses the interaction.
Limit: Finite event traces and duration witnesses do not estimate risk likelihood, validate engineering or prove that priorities have been agreed.
Integrated risks and tasks — One lane, two plans · Supporting source 2
240 · Package boundaries
Compare full state and labelled joint-step graphs for direct assembly and separately retained fragments, including early output and shared resource supply.
Six once-only events, count tokens and selected-work completion. All shared inputs gate mounting; budget is ignored. No schedule or reopening is inferred.
Native construction retains partial occurrence sets, boundary input/output vectors and joint steps; it compares composition with directly glued count nets.
Adaptation uses two prepared deck panels, unrelated supplier packing and a receiver that mounts then accepts both panels. All joint inputs are checked before returned lifting tokens are credited.
What this establishes: Direct assembly and open fragments agree; closing a receiver deletes mounting, whole-package barriers hide early release, and copied crane supply adds unsupported joint steps.
Limit: Finite equality here is a package-preservation check, not autonomous engineering, a general composition theorem or measured planning benefit.
Higher-autonomy enquiry — Boundary Contract Lab · Supporting source 2
250 · Spines and slices
Compute transitive precedence, a maximum-cardinality chain and all finite antichains; validate a covering partition with one spine task per part.
Seven activities; cycle detection precedes poset claims. Shared capacity checks lifting demand separately; access, permission and condition are readiness gates only. Budget is ignored.
Native finite shed selects a maximum-cardinality chain and checks coverage, transitive incomparability, minimum partition size and one chain member per slice.
Bridge adaptation checks the same certificate and makes the review-loop obstruction explicit. A cycle is rejected before height or width is reported.
What this establishes: The mixed placement-and-acceptance bucket fails transitive incomparability; height-based slices pass, while shared lifting capacity remains a separate condition.
Limit: A finite spine certificate is existential for the displayed partition. It is not a universal infinite-poset result, a resource-feasible schedule or human-use evidence.
Project Spines — finite shed and antichain slice audit · Supporting source 2
260 · Interaction and hiding
Compose two finite signed partial functions by routing their internal interface until an exterior exit, missing rule or exact repeated state is found.
Access-report and proposed-method messages for both sections. Shared evidence and capacity can supply a blocked report; access chooses low/high report. Budget is ignored. No work execution occurs.
Native Island Reading Room supplies finite typed partial-function routing with signed A/B/C interfaces and two independently initiated exterior episodes.
The bridge replaces landing offers with low-water/bank method offers. A request episode never silently becomes a supplied report; missing and looping interactions remain distinct diagnostics although both leave the exterior function undefined.
What this establishes: Routing computes exterior proposals and distinguishes a missing reply from a repeated internal exchange; an external request ends at a request for evidence.
Limit: This adapts the native finite Int(Pfn) construction; it does not establish general functor laws, legal game strategies, learned policy or completed delivery.
270 · Enough local knowledge
Intersect admissible actions over whole observation classes; enumerate every route-case partition and separately check the reopening owner’s completion information.
Three logistical route contexts, optionally doubled by uncertain reserved-route setup. All shared gates apply; access removes low-water actions. Budget is ignored. Checkpoints do not imply eventual completion.
Native whale construction uses whole-observation-class action intersections and separate route and reopening owners; its second essai adds a known robust service route.
The bridge has low-water, bank and reserved-bank actions. Exhaustive partitions measure theoretical information outcomes, not necessarily implementable sensors. Known inspection and reliable report delivery are strong supplied assumptions.
What this establishes: Blind route choice fails despite physical options; a cue or known robust route repairs it. A possibly missed completion report still prevents a uniform reopening decision.
Limit: Finite authored worlds, observations and checkpoint rules are not a temporal epistemic planner, evidence of real communication or measured coordination benefit.
Knowing enough — Moving the whale and One workable choice · Supporting source 2
280 · Capability compilation
Compile the current method, inputs, evidence conditions, procedure and reversal rules; test an explicit receiving-context fixture and retain failed conditions.
A deterministic packet for two-section recovery; all shared conditions are retained. A capture allowance does not establish completed capture or review. Budget is ignored. No physical work is executed.
Native Frontier compiler assembles a packet from the demand/evidence context, retains stop/reversal conditions, checks adaptation to a different access setting and distinguishes packaging from independent reuse.
The bridge adaptation selects this one construction. The demand, truth-sensor, allocation and tournament lenses remain separate; no universal integration or higher-order innovation result is claimed.
What this establishes: A parameterised packet passes the compatible bank fixture; a literal low-water method, short receiving window, stale evidence or insufficient capture allowance exposes an explicit boundary.
Limit: Packet and transfer checks do not establish learned capability, independent reuse, recipient acceptance, commercial value or completion of the wider innovation enquiry.
Meta Project Innovation — Project Capability Compiler · Supporting source 2
290 · Causal assumptions
Enumerate eight equally weighted structural records, compare raw association, readiness adjustment and equation replacement, then remove observability or overlap.
A preparation score, not duration, money or physical safety. Shared work gates are reported separately and never alter the assumed causal score; budget is ignored.
Score = 4 + 4 × readiness + selected coordinator effect; the readiness population is held fixed across interventions.
The default treatment occurs in one of four low-readiness draws and three of four high-readiness draws.
Removing overlap restores readiness-dependent assignment even when the random graph was selected. A hidden readiness value is unavailable to both observational comparators.
The intervention result remains an equation-based assumption when the observed record cannot support adjustment.
What this establishes: The eight-record witness has a raw difference of +1 and an assumed intervention difference of −1. Matched ordinary stratification agrees; hidden readiness and absent overlap block adjustment.
Limit: The arrows and equation are authored. Adjustment and simulation do not establish an actual causal effect or validate a repair policy.
Counterfactuals: Causal Lab and arrow elicitation · Supporting source 2
300 · Current evidence
Keep exact cited passages, change a source packet, invalidate its prior inspection, and explicitly refresh without resolving absent or contradictory evidence.
One evidence-librarian intervention from a 32-app collection. Shared access, permission, condition and capacity seed the packet; budget is ignored.
A local inspection binds to the complete source packet, including new contested records.
Changing or adding any record preserves old passages and invalidates that packet-wide inspection. Refreshing passages alone does not record a new inspection.
The ordinary comparator sees identical current and retained passages, versions and review bindings.
Contradictory sources remain visible together. No algorithm chooses the authoritative account.
What this establishes: A changed source preserves the retained passage but reopens inspection. A current ordinary versioned register reaches the same gate result; explicit refresh cannot cure missing or contested evidence.
Limit: Deterministic version handling and a preidentified contradiction do not demonstrate semantic retrieval, user benefit or organisational adoption.
310 · Scoped feedback
Execute four explicit policies on eleven authored use cases, preserve original-use checks, and compare scoped corrections with ordinary access to identical reasons.
Deterministic policy execution, not a new language-model learning trial. Shared inputs generate two cases; budget is ignored.
Explore means a reversible desk comparison that does not assert readiness or perform physical work.
Commit means the fixed low-water placement proposal: a known closed window or zero capacity warrants rejection; otherwise missing permission or condition evidence warrants inquiry.
History, scoped and global policies see the same correction reasons; the frozen reference intentionally lacks the new commitment rule.
Keep, ask and reject are defined by the above contracts. Outputs are programmatic decisions, not captured participants.
What this establishes: An unscoped commitment rule blocks valid exploratory cases. Purpose-scoped rules and ordinary reason-bearing history agree on the entire authored bank, so no scoped-method advantage is awarded.
Limit: These cases and answer contracts are designed together. Scripted conformity establishes neither transfer of judgement nor a gain over ordinary practice.
Learning from what we keep: Lessons with a scope, Essai 2 · Supporting source 2
320 · Equal facts, two forms
Present identical complete bookings and rules in two forms, calculate conflicts, and separately test loss of permission evidence.
Six bookings with half-open integer intervals. All shared work gates apply; budget is ignored. Acceptance work here is a booking label, not proof of cured or accepted structures.
Each crew serves one booking at a time; global simultaneous lift demand is separately limited by shared capacity.
Both forms carry full IDs, crew identities, start and end values and resource demand. Aligned bars are redundant.
Known resource or access contradictions take priority over unresolved evidence. Missing permission never becomes approval.
The native screenshot pilot found 16/16 decisions in each form; these bridge calculations add no new human or model-session observation.
What this establishes: The table and aligned view preserve the same twelve statements across touching, nested and different-crew controls. An independent interval oracle checks the answer; no representation advantage is claimed.
Limit: No bridge reader experiment has run. Equal statements and computed answers do not measure which form improves judgement, speed, confidence or effort.
Picture and judgement: aligned-time pilot and next-question interpretation · Supporting source 2 · Supporting source 3
330 · A bounded commission
Audit six declared dependency predicates, deliver findings or a hold, and independently accept only a complete accurate receipt tied to the current source.
A dependency extract commission only. Shared physical settings and budget are excluded from its verdict; lifting capacity and work gates are displayed as separate counterexamples.
The unit uses half-open integer days and finish-to-start zero-lag links under a declared governing source.
Convention, unique identity, positive intervals, predecessor existence, acyclicity and date agreement are the six checks.
Unresolved conventions, conflicting identities or an expanded readiness request cause an explicit hold; downstream checks are withheld.
Acceptance compares the exact source contents, report coverage and findings with a separate ordinary-table implementation. This is machine acceptance only.
What this establishes: The graph audit and ordinary table agree; early links, cycles, absent predecessors and ambiguous conventions are exposed. Stale and erased adverse receipts are rejected, while external resource infeasibility stays outside the dependency claim.
Limit: No buyer, payment, practitioner acceptance, observed effort saving or commercial viability is established. The input may omit necessary work.
Professional work by the piece: dependency-check commission · Supporting source 2
340 · Where a layer fails
Enumerate four placement subsets, compare section-local permission with joint lifting demand and an ordinary time-slot table, then state the valid range.
An authored application of common-coupler and failure-range heuristics. Shared capacity and evidence gates apply; timing and method are local controls; budget is ignored.
The local model checks each selected section in isolation; the refined model retains simultaneous demand on the shared lifting pool.
The ordinary table retains the same complete record and global constraint as the refined model.
The six capacity/timing regimes describe the resource-only two-placement boundary; missing evidence is a separate common gate.
Keep the valid local use and add an update trigger; never treat the authored interaction weights as observed empirical coupling.
What this establishes: Two individually feasible placements conflict when their intervals overlap at capacity one. Touching intervals and capacity two restore agreement; a complete ordinary time-slot table always matches the refined model.
Limit: This is a constructed stress test of a proposed heuristic kit, not empirical abstraction discovery, a general theorem, or evidence that a richer diagram improves work.
Abstraction layers: common-coupler check and layer-shift protocol · Supporting source 2