The Organism Appears First in Possibility-Space

The instrument

Can the whole forecast itself better than its best split?

We take a short history of the current state and ask how much it tells us about the next state. Then we split that state into parts and ask again. The whole-over-parts difference is our Flow Lenia analogue of integrated predictive information.

Only the past chooses the measurement or action. Later morphology stays sealed until the prediction or intervention is fixed.

The headline number needs its scale next to it: the whole-state predictive signal runs about 8.76 nats and the prospective rise of +0.364 is roughly 4 percent of that, while the whole−sum level itself stays negative through all of development, so the best split out-predicts the whole at every phase and what we measure is the whole catching up with its best split, still behind it when development ends.

1 · Same evidence.
The whole and split see the identical past. 2 · The hardest split.
We freeze the best partition. 3 · The difference.
Positive values mean cross-part organization matters.

Past window onlyfreeze split or actionthen open the future

Two related readings. The diagram shows the frozen best-partition whole−sum primary. The Φ-R MMI companion adds its redundancy correction: +0.364 and +0.171 are not numerically interchangeable.

Before the evidence

The recurring terms, in plain words

q8, q12, q32, q48: The organism's age counted in growth pulses. q8 is very early, well before anything looks like a coherent body.
passage (P1–P12): A fixed development window used as the time unit for futures. P6–P12 means six to twelve windows after the intervention.
whole−sum: How much better the whole state predicts its own future than its best split into independent parts. Positive means the coordinated whole carries predictive structure the parts alone do not.
Φ (phi): Shorthand for that whole-over-parts score. Φ-forward pushes the state along the direction that raises it; Φ-backward pushes the opposite way.
the information turn: The developmental moment when the actions favored by Φ and by ordinary predictability stop agreeing. It happens before visible form.
action, fan of futures: An action is one of eight fixed local pokes applied to a state. A fan of futures launches matched copies under each action and watches how far the outcomes spread.
predeclared vs post-hoc: Predeclared (or frozen) means the analysis and its pass rule were written down before results were seen; post-hoc means the reading was made after looking at the data. Predeclared results carry weight; post-hoc results are leads.
primary vs secondary: The primary is the one predeclared outcome an experiment stakes itself on. Secondaries are also named in advance but carry less weight; anything else is exploratory.
family, group: Organisms grown from the same founder. Effects are averaged within families first, so the statistical unit is usually 16 or 19 families, not the larger organism count.
source-bound: Every number in this report is copied from a result file whose SHA-256 hash appears in the ledger, so each claim can be traced to the exact JSON that produced it.

Integrated predictive information gives us one partial coordinate of development and one useful local direction to push along, and what it turned out to be best at here is showing how the action-world is organized and when that organization changes.

How the question changed

From steering replication to watching an organism commit.

We began with a three-link steering programme on fed reservoir lineages, looking for a number that rose before replication and could be pushed, and what actually happened is that the flagship failed its frozen test while the experiments around it kept pointing at how the developing state constrains what a poke can do, which is what the rest of this report measures.

01

Silhouette splitting was the wrong event.

Reservoir recurrence and committed morphogenesis gave us cleaner Flow Lenia events than silhouette splitting.

02

The flagship steering test failed.

reservoir-causal-steering-v1 completed its 1,920 scheduled trajectories and did not meet its predeclared causal chain: the required fixed-horizon onset-reward link failed, and the Φ-specific steering test failed with it. That null is what moved the endpoint from restoring fed lineages to single-organism morphogenesis.

03

More Φ was not a universal push.

The information turn replicated prospectively, while Φ-high versus Φ-low changed sign as an actuator.

04

Matched future fans instead of single endpoints.

Launching matched futures exposed canalization that one final body or one scalar score could not show.

05

Spatial structure at q8.

At q8, nearly identical visible states already carried strongly different causal futures, and scrambling the cells took that structure away.

06

Several kinds of future, several schedules.

Timing, action meaning, dominant-body allocation, and fine shape each close on their own schedule.

The evidence chain

From prediction to causal closure.

The cards below run in the order the questions got harder, from whether the signal appears at all, to whether it forecasts, to whether we can move it and whether the organism remembers being moved, and finally to which kinds of future actually close.

Predictive whole

The whole-over-parts signal rises before the body does.

prospectivereplicated precursorspecificity unresolvedevent-aligned atlas

Across a fresh 64-specimen cohort, 54 had complete primary windows: frozen whole−sum predictive organization rose before an independently certified sustained break from the artificial square, while the specificity contrast remained unresolved. A second predeclared companion also came back null: adding whole−sum to ordinary visible measurements left held-out event prediction essentially unchanged (AUC 0.8089 to 0.8093), so the rise marks the transition without yet improving a practical forecast. Following that same developmental arc with exact local probes showed a further change: the actions favored by integration and by ordinary prediction stopped agreeing.

Fresh pre-form whole−sum+0.364

54 complete primary windows; 95% interval +0.231 to +0.502; 39 positive, with 15 of 16 family means positive. The rise is about 4 percent of the 8.76-nat whole-state signal, and the whole−sum level stays negative throughout.

Developmental atlas−1.75 → −0.32

Whole-over-parts increased by +1.43 from early development to transition; 16 of 16 families rose. The phase axis is each organism's own already-observed transition step, so this panel is a retrospective alignment, not a forecast.

Compass rotation44 → 4

Φ-high and TDMI-high selected the same action in 44 of 64 organisms early, then only 4 at transition. TDMI along Φ flipped +2.16 → −0.73.

Three grayscale Flow Lenia field frames from the same prospective trajectory. The first is a dense artificial square; by step 42 it has reorganized into an elongated striped form; by step 74 that form persists with further internal rearrangement. — **Frames from one held-out trajectory**The predictive rise is measured before the independently certified sustained departure from the square. These frames show the morphology only; they do not enter the past-only choice.

The system is becoming a coordinated whole before the body detector fires, and the kind of prediction that integration favors is changing as it does.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`organism-birth-signal-v1/before-the-creature-prospective-result.html`	`organism-birth-signal-v1/prospective-results.json`	`complete_fresh_prospective_replication` prospective precursor replication without specificity	64 specimens; 16 families; 54 complete primary windows; 63 observed events; 1 right-censored	`prospectiveCompletionSHA256 2dbf4f19d65f53490906d111f8262e10502390e5213603199dcea371634e8559`
`developmental-basin-atlas-v1/before-the-organism-developmental-basin-atlas.html`	`developmental-basin-atlas-v1/results.json`	`complete_open_ended_developmental_basin_response_atlas` open-ended causal atlas, event-aligned to each organism's known transition	64 specimens; 16 families; 5 phases; 1,280 trajectories	`fullCompletionSHA256 4f81f0418fc9a29da4033838c339c320d80ec1ebad38bc72982017ba82dc31e8`

Geometry clock and actuator test

Past-only geometry tracks developmental progress, but pushing along Φ does not advance the transition.

future-blind clockfresh cohortopen causal atlas

Past-only information geometry tracked actual developmental progress, then reproduced its turn in a separate cohort of unseen organisms. But pushing the instantaneously Φ-high action did not reliably advance the visible transition. The measurement reads a real developmental coordinate; force along one local axis is not universally the mechanism.

Phase coordinate · predeclaredρ +0.523

Whole-over-parts versus actual progress, pooled across checkpoints; 39 of 62 eligible organisms were strictly monotone. Exact countdown did not improve: MAE 39.88 with ordinary controls, 40.46 after adding information geometry.

Fresh turn · post-hoc25% → 75%

Turned-state prevalence from step 24 to 48 in the open-discovery panel of the causal atlas; 33 organisms crossed into the state and 1 crossed back. This reading was made after the data were seen and sits outside the frozen clock claim.

Actuator testNo fixed sign

Φ-high minus Φ-low timing effects changed sign across four clocks. The order parameter survived; the universal one-shot control claim did not.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`prospective-geometry-clock-v1/the-state-turns-before-the-form.html`	`prospective-geometry-clock-v1/results.json`	`complete_fresh_prospective_geometry_clock` future-blind prospective coordinate	64 specimens; 16 groups; 64 observed events	`fullCompletionSHA256 664cdad167dceff3fc7a880509ea32399187210bd667be862cd44be343e2a88a`
`geometry-turn-causal-test-v1/geometry-turn-causal-atlas.html`	`geometry-turn-causal-test-v1/results.json`	`complete_open_causal_susceptibility_atlas` fresh prospective replication plus open causal atlas	64 specimens; 62 untouched observed events; 832 full causal futures	`fullCompletionSHA256 3e2106a47f2ae33ea466bde2f6e7dc8695d5207262f0dcbd40287946a4f2e21f`
`geometry-turn-causal-test-v1/geometry-turn-causal-atlas.html`	`geometry-turn-causal-test-v1/open-discovery.json`	post-hoc open discovery; source of the 25% → 75% turn-prevalence reading	same cohort, turned-state panel at steps 24 and 48	`SHA256 2074c1bd5abcfef399d5070f1d0b2f1c551ab8ec138d6d36e69bd58272a29753`

From forecast to intervention

More whole-over-parts predicts a narrower timing future, and an early push leaves a small trace in later microscopic outcomes.

prospective forecastdirect causal

Every eligible state received the same full eight-action repertoire. Before visible form, higher whole-over-parts organization forecast a smaller range of transition times beyond ordinary morphology. A separate within-ancestor experiment then changed an early commitment coordinate. The outcome it staked itself on, whether high-commitment siblings would spread their transition times over a narrower range, did not resolve: the median difference is exactly zero, with interval [−6.67, +1.0]. What did move is a secondary the plan also named in advance, and it is small: under the same eight actions, high-commitment descendants ended in slightly less varied final fields, a 2.3 percent shift against the low-commitment baseline of 0.341.

Timing repertoire−0.248

Partial rank association, 95% interval [−0.420, −0.045], controlling checkpoint, shape, size, topology, and raw field change; 63 of the 64 cohort specimens entered the analysis.

Same eight actions222 states

All eligible pre-transition states received untouched plus all eight fixed perturbations; 1,840 long futures.

Branch secondary · timing primary null−0.0078

Mean pairwise final-field repertoire, high minus low commitment; interval [−0.0152, −0.0008]; 30 pairs. The predeclared timing-range primary was null, median 0.0.

Drawn fans, measured contrastsThe drawn fans explain the measured direction; the coefficients and intervals above are the bound results.

The information coordinate forecasts how redirectable development still is, and one early sibling push changes what later pushes can reach.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`causal-repertoire-contraction-v1/the-future-closes-before-the-organism-appears.html`	`causal-repertoire-contraction-v1/results.json`	`complete_fresh_full_causal_repertoire_test` fresh future-blind prospective repertoire forecast	64 specimens; 222 eligible states; 9 futures per state; 1,840 long runs	`fullCompletionSHA256 c04c782477371fb6d504f919467f38ce280680d98157c48dfd52034d391de432`
`commitment-branching-intervention-v1/one-early-push-narrows-the-later-possibility-space.html`	`commitment-branching-intervention-v1/results.json`	`complete_fresh_within_ancestor_commitment_intervention_test` direct nested causal intervention	31 fresh ancestors; 30 selected pairs; 18 futures per ancestor; 540 long runs	`fullCompletionSHA256 4813d53663f893d47f7ba9a992057ab1b372412c6cb063a880a49583dbd00b70`

The action-world at q8

The action-world is already broad at q8, and the spatial arrangement is what gives it structure.

fixed-age atlasexact spatial ablation

At the earliest measurable two-pulse state, visually close histories already led to strongly separated later futures. Scrambling the exact same cells, while preserving genotype, masses, composition, and world, made formation slower and its action geometry less coherent. Leaving the cells where they were did not make more futures reachable, it made the reachable futures faster and better sorted by where the poke landed.

q8 causal world104.4%

q8 already carried 104.4% of the q32 history-conditioned future separation.

Matched futures0.378

q8 history-conditioned future displacement, versus only 0.088 between the corresponding release states.

Scramble delays+12.85 / +22.51 / +30.59

Passages for 8×8 tiles, 4×4 tiles, and cells relative to intact. All 39 untouched states still formed.

Sensitivity to a local poke at q8 depends on where in the arrangement the poke lands, since scrambled copies of the same field lost that dependence, and the intact arrangement both makes nearby interventions more legible and constrains their distant consequences.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`initial-action-world-v1/plasticity-is-there-from-the-start.html`	`initial-action-world-v1/results.json`	`complete_fixed_age_q8_to_q32_action_world_curve` fixed-age causal atlas	39 organisms; 156 states; 5,616 causal futures; 312 matched-history pairs	`completionSHA256 7de5b00378d093bfa0dfa85fa4f4a2ce4164664436fc88765e0b777c2aa55359`
`spatial-scale-action-world-v1/organization-disciplines-the-action-world.html`	`spatial-scale-action-world-v1/results.json`	`complete_multiscale_spatial_ablation_of_q8_action_world` direct same-state spatial ablation	39 organisms; 4 exact spatial conditions; 5,616 causal futures	`completionSHA256 c0b1e358be4043343104691412d9709e5b86a2cffbffd9f45a190a7e743c81c4`

State-setting and memory

A brief feedback policy holds the causal gate open, then the state carries itself forward.

closed-loop causalselected-future follow-up

The successful controller usually delivered one large pulse and then stopped. Open and closed branches stayed separated after release. In the long follow-up, their field divergence grew even as the special advantage of the originally selected action vanished. After that the two branches kept diverging on their own, so whatever the pulse did is now carried by the state. One detail changes how to read the +0.075: against untouched controls the opened branch itself barely moved (window means +0.002 to +0.007), while the closed branch fell below untouched by −0.014 to −0.029 everywhere, so most of the +0.075 separation comes from closing the gate costing something rather than from opening it adding something.

Post-release memory+0.075

Cumulative open-minus-closed causal-history separation; interval [+0.059, +0.090]. All 16 families positive.

Field divergence0.389 → 0.448

Open versus closed from selection frame to step 256; all 39 organisms remained separated.

Action advantage0.040 → 0.002

Max-selected minus min-selected advantage faded while the state divergence continued to grow.

A two-by-two grid of real Flow Lenia fields for specimen s63. Columns compare release at step 32 with the later field at step 256. Rows compare the max-selected and min-selected feedback histories. Both histories form structured bodies, but their later morphologies remain visibly distinct. — **Separation after the pulse ends**One representative paired future. Across all 39 organisms, branch separation persisted after release. The separation shows the state remembers the pulse, while which direction the pulse ought to have had is the separate question the actuator test left open.

A perturbation can set a developmental condition that outlives it, and from then on the basin carries the difference on its own.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`commitment-gate-controller-v1/the-gate-stays-open.html`	`commitment-gate-controller-v1/results.json`	`complete_closed_loop_commitment_gate_control_and_release_test` direct closed-loop control and release	39 organisms; 6,240 candidate probes; 2,496 release readouts	`controllerCompletionSHA256 12009169931a2b3ba330336293f006eb19bf9e9f904935419fc8704583e764fa`
`commitment-gate-formation-v1/the-hidden-state-becomes-a-body.html`	`commitment-gate-formation-v1/results.json`	`complete_selected_future_visible_formation_analysis` selected-future long follow-up	39 organisms; 312 trajectories; 79,872 frames	`completionSHA256 c0aa1de549c693b060ac885d63bcc63f9e9f7ba1cf2e9d43b5146bdc09500ad9`

Hidden-channel transplant

The same visible body can contain a different future.

exact hidden transplantfresh reciprocal cohortfrozen primaries unresolved

We preserved the visible mass field at every pixel and swapped only the hidden channel mix underneath it. The outcome this experiment staked itself on, that hybrids would slide toward their donor along a single expansive-versus-compact axis, did not come through (+0.050, interval −0.258 to +0.370), and one pre-named secondary even moved the wrong way: hybrids ended farther from their donor's field than their host's in all 16 groups. What survived is a different pre-named secondary: later bodies moved 27.9 percent along the direction their donor naturally grows, though mostly for one direction of the swap, 39.8 percent for compact bodies given expansive composition against 16.1 percent the other way. A fresh experiment swapping hidden state between young and slightly older organisms then asked its own frozen question, whether the older mix advances the developmental clock; it does not, with the shift leaning the other way, while a second look at the same data found a hidden direction that held back later body growth.

Visible residual · by constructionexactly 0.0

All 156 transplant starts preserved channel-summed visible fields bit-for-bit. The transplant constructor enforces this and the capture audit confirms it; it is a validity check, not a finding.

Donor-native pull · predeclared secondary27.9%

Mean movement along the paired donor’s native morphological direction at +128; interval [12.1%, 47.2%]; 14 of 16 groups positive. An unbounded projection, not a fraction of distance covered, and direction-asymmetric: 39.8% one way, 16.1% the other.

Hidden counterweight · second pass−0.096

q8→q12 hidden displacement lowered later radius at +96; interval [−0.156, −0.043]; 14 of 16 groups negative. For scale, the natural q8→q12 radius gap at capture is +0.198, so the displacement is about half that gap, read over a different time base.

Channel allocation schematicThe blobs are a schematic of channel allocation, not specimen screenshots. The residual is enforced by construction, and the donor-ward displacement is the predeclared secondary projection; the frozen primary axis shift did not resolve.

Two organisms with pixel-identical visible fields can carry different hidden mixes, and the hidden mix biases where the body goes next, though so far that shows up in a predeclared secondary projection and a second-pass growth restraint rather than in a frozen primary.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Completion binding
`latent-composition-transplant-v1/the-body-stays-the-future-leans.html`	`latent-composition-transplant-v1/results.json`	`complete_hidden_composition_causal_transplant_analysis` direct visible-field-preserving transplant	39 organisms; 16 groups; 156 exact transplants; 1,248 frames	`completionSHA256 b3c1354772459d0f7824dcd567c551ca9162789c3c4034dd0c45fc7ea43fe294`
`developmental-ignition-transplant-v1/the-hidden-future-pushes-back.html`	`developmental-ignition-transplant-v1/results.json`	`complete_fresh_reciprocal_hidden_age_transplant_analysis` fresh reciprocal direct intervention; frozen clock question resolved no	64 organisms; 16 groups; 128 captures; 640 futures; 122,880 frames	`completionSHA256 a20346a43025df0b7ef652d73acb1a8f3b934e7e60089959039a2d24bca7a728`
`developmental-ignition-transplant-v1/the-hidden-future-pushes-back.html`	`developmental-ignition-transplant-v1/counterweight-results.json`	post-hoc second-pass counterweight analysis; source of the −0.096 growth-restraint reading	same cohort; q8→q12 hidden-vector future effects at +96	`SHA256 086729c395b6d6c06b79969a0eacb7fb615f8ec0a3ac7a333efb28bcbbfc2f33`

Causal grammar

After the information turn, morphology becomes harder to steer and actions acquire more stable meanings.

fixed intervention panelturn fixed upstream

The same four founder interventions lost spatial leverage just after the structural turn while compositional leverage stayed open. In a two-action experiment, the later action’s spatial meaning also became less rewritable by which action came first. One predeclared check on the map itself did not pass, the rule asking the closure to still be there four passages past the peak, so the claim rests on the turn-versus-after contrast and the grammar experiment rather than on a persistent plateau. Past the turn, what a founder input does depends more on the organism it lands in and less on the input itself.

Morphological steerability−13.9%

Turn versus two passages after: +0.00290 leverage before closure, interval [+0.00049, +0.00529].

Action-geometry rewrite−11.1%

Before-minus-after rewrite +0.002935, interval [+0.000493, +0.005409].

Selective closureSpatial first

Founder-composition leverage remained open while later spatial futures became less redirectable and less context-rewritable.

The organism-level state increasingly constrains how local founder inputs propagate, without freezing every internal degree of freedom, which is the intervention-level form of causal emergence this programme set out to find.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Manifest / completion binding
`developmental-causal-susceptibility-map-v1/the-future-narrows-here.html`	`developmental-causal-susceptibility-map-v1/results.json`	`complete_dense_organism_relative_causal_susceptibility_map` fixed-panel causal commitment map	19 families; 38 schedules; 2,736 futures; 32,832 decoded fields	`manifestSHA256 45aace45f3fc07b211aa71a1930182205a869e276b38ae31784a8d215770cde8` `completionSHA256 e9634826f44cf4e1e74c13bb8af6e31aca7bde21872b662dda7aa3131985ad70`
`developmental-causal-grammar-v1/the-organism-learns-what-actions-mean.html`	`developmental-causal-grammar-v1/results.json`	`complete_two_action_developmental_causal_grammar` future-blind two-action causal grammar	19 families; 38 schedules; 152 complete 4×4 panels; 2,432 trajectories	`manifestSHA256 dbb2a36d8d77d80d895f07f6f8795f42188aab27a4aaeab8e32dce75590f4dd9` `completionSHA256 3753423409b2f86c8df44da7bf4b7942c6049fb514cc2dd63962421d8409db54`

Selective future closure

A Φ-directed push narrows different organism-level futures at different developmental moments.

direct causalpost-hoc coarse leadpredeclared fine-shape secondary

A post-hoc reading of one fixed early-state panel found that Φ-forward made eight unseen futures agree more about which connected body would dominate; the frozen primary in that study was unresolved. In the three-age follow-up, the coarse primary again stayed unresolved, but at each organism’s structural peak the predeclared fine-morphology panel converged. Two caveats: the peak arm is selected after the fact, since both predeclared timing contrasts straddle zero, and the fine-shape panel is one of 78 predeclared contrasts in this experiment, of which 5 resolved. The eight futures still differ in coarse layout and founder mix, and only their fine shape drew closer.

Post-hoc dominant-body lead−0.01358

Φ-forward minus Φ-backward; interval [−0.02509, −0.00181]; 14 of 19 families narrowed.

Fine shape at peak−0.0040

Pairwise shape distance −0.00399, interval [−0.00713, −0.00117]; 13 of 19 families narrowed. On the frozen effective-shape estimator the same contrast is −0.00426, about 0.4% of a saturated baseline near 1.08 effective shapes of 8; −4.55% is the pairwise-distance framing of the same effect.

Delayed effectP6–P12

Every fixed passage interval from 6 through 12 pointed toward canalization. Founder mixture and coarse body organization stayed open.

Point estimates with family-clustered 95% intervals at each passage and no smoothing; the direction holds from P6 onward while the size varies from passage to passage.

Our current reading is that commitment is multidimensional, with development closing specific kinds of future on their own schedules rather than shutting all possibility at once.

Source receipts for this card

Report	Result	Status / claim class	Cohort	Manifest / completion binding
`tangent-future-aperture-v1/the-future-chooses-a-main-body.html`	`tangent-future-aperture-v1/exploration.json`	`complete_post_hoc_source_bound_exploration` post-hoc direct-intervention lead; frozen primary unresolved	19 families; 38 schedules; 4 treatments; 1,216 trajectories; 8 futures per treatment	`manifestSHA256 f50ead16920b9749a029a728c946459d863cbf172ac1b9ab341a7acfbd60c8b2` `completionSHA256 29dad2bbd870ff2eff907f8b31a93e0904b1f31c95d1af602d5008db3588e69f`
`developmental-commitment-aperture-v1/the-shape-space-closes-late.html`	`developmental-commitment-aperture-v1/results.json`	`complete_source_bound_future_blind_analysis` predeclared secondary direct causal result; coarse primary unresolved	19 families; 38 schedules; 3 timings; 4 treatments; 3,648 trajectories; 8 futures per treatment	`manifestSHA256 9d123a6fff24d9675dd0c692c98a214299194b4edef0c15d64339403077eb952` `completionSHA256 cc947fc2bf181f4f263312b0c7a7d0710afce04c2a61fd1e66ce84d92d356de1`
`developmental-commitment-aperture-v1/the-shape-space-closes-late.html`	`developmental-commitment-aperture-v1/morphology-timing-exploration.json`	`complete_source_bound_secondary_and_descriptive_analysis` source-bound predeclared secondary and descriptive timing analysis	19 families; fixed P1–P12 morphology panel	`manifestSHA256 9d123a6fff24d9675dd0c692c98a214299194b4edef0c15d64339403077eb952` `completionSHA256 cc947fc2bf181f4f263312b0c7a7d0710afce04c2a61fd1e66ce84d92d356de1`

The synthesis

Commitment is selective closure.

The organism arrives on several schedules at once, since the timing repertoire narrows before visible form while action meanings settle only after the information turn and the fine shape fan closes later still, and that is why one scalar sometimes predicts well, sometimes acts weakly, and sometimes points at one specific aperture.

Five apertures, five schedulesHorizontal position shows developmental order. Fan width is explanatory; labels carry the measured effects and evidence class.

The campaign record

What did not survive, in the open.

The synthesis above is built from sixteen experiments. The campaign that produced them ran ninety-one, and the follow-up programme below brings the map to ninety-five. Reading the cited few against the ninety-five is what keeps the claim honest.

The programme opened as a three-link steering campaign on fed reservoir lineages: find an integrated-information precursor of replication, steer it, and show that the steering moved replication. Five early designs were stopped before analysis and sealed with forensics, and three more runs stopped at their own mechanical gates without reaching a scientific verdict. The flagship completed its 1,920 scheduled trajectories and then failed its frozen causal chain, so the campaign pivoted to single-organism morphogenesis, and everything above comes from that second phase. Inside the second phase, two predeclared prediction tests also came back null and are disclosed on their cards, the held-out event forecast and the exact countdown clock, and the nulls below actively bound how the positive results may be interpreted.

Each square below is one experiment folder, in alphabetical order; hover for its name and, where a frozen verdict exists, a one-line outcome. The bar for the green class is strict: a prediction declared before the outcomes were opened that then resolved in its stated direction. Of the ninety-five, ten clear that bar, twenty-four predeclared questions came back null or unresolved, three resolved a predeclared quantity in the wrong direction (two as pre-named secondaries on top of a null primary, one as the primary itself), eight were sealed as aborts or stopped at their own mechanical gates, and fifty are exploratory or descriptive work that never declared a prediction at all. The blue outline marks the twenty experiments cited by this synthesis or by the follow-up programme; several of them sit in the exploratory class, including the scale atlas and the basin atlas, and their cards above carry that framing. Two frozen results files, causal-compass-peak-v1 and causal-history-order-v1, still carry headline strings asserting the positive branch their own primary verdict rejected; this map reads the verdict fields, not the headlines.

predeclared prediction met · 10 predeclared null or unresolved · 24 resolved in the wrong direction · 3 aborted or mechanically stopped · 8 no declared prediction · 50 outlined: cited by this synthesis or the follow-up · 20

reservoir-causal-steering-v1 · the flagship

The frozen chain required a fixed-horizon onset-reward success and it did not arrive; the report's own label is primary_chain_not_met, and the Φ-specific steering test failed with it. Pushing Φ directly does not steer replication in fed lineages.

prospective-results-v1b/prospective-analysis-report-v1b.json · eab15f3c894edf4c8865f8ebc22f7ee84a9052b630d1376c593ae29967ca1815

tangent-induced-causal-grammar-v1

Pushing along the information tangent at fixed age did not reliably move the causal grammar. This bounds the mechanism behind the grammar card: the turn marks when action meanings stabilize, but the tangent itself is not the lever.

results.json · 3ff7652b91f2670201364275e46dc11bb806c0316f574cff79847f889971436e

developmental-causal-plasticity-window-v1

Structural information peaks did not amplify founder-action fanout, so the peaks used as intervention moments elsewhere in this report are not general sensitivity windows.

results.json · 18eed6194ce1125f77e12ca0966b3104b8388d89b09fdce13678d5faf43d3ac6

fresh-organic-q40-replication-v1

At fixed q40 the link between compass rotation and causal future breadth did not replicate: Spearman +0.093 with interval [−0.178, +0.352], prediction not passed.

results.json · a321b7922195f0a8cf3198d6961f2159ed4c6065b8cef69a3b9e616d21f7bc2e

causal-compass-peak-v1

The sharpest compass turn did not mark a broader future repertoire: +0.0023 with interval [−0.0028, +0.0073], prediction not passed.

results.json · 779277c93261976e9d95eacb2812cbc4f69e03005e9af3063d7329f4a30fb8f3

Eight stopped runs

Five were sealed aborts with forensic notes: goal-state-bite-calibration-v1, goal-state-matched-wound-recovery-v1, perturbation-recovery-engineering-v1 and -v2, and reservoir-cycle-emergence-census-v1. Three more stopped at their own mechanical gates without a scientific verdict: goal-state-tomography-v1 ended with 0 of 8 valid states under a protocol that forbids reading that as a negative result, preinjury-predictor-v1 stopped when its primary feature was unavailable and never ran the test, and orthogonal-pulse-release-v1 never bound its declaration after its qualification runs aborted.

One decisive next replication · since executed

Freeze the shape claim as the only primary.

This was written as the experiment that would turn the newest signal from a strong predeclared secondary result into a clean headline result. The protocol below is preserved exactly as it was published here; the verdict follows it.

1

Fresh, unseen cohort

Sample new organic families and new matched future worlds. No reuse of the 19-family outcome panel.

2

Endogenous timing

For each organism, find its structural peak from past-only information geometry and intervene there. The peak-selection rule is itself predeclared this time; in the current data the peak arm was chosen after the results were seen.

3

Five matched arms

Φ-forward, Φ-backward, orthogonal, score-blind, and untouched, with the same local dose and the same future seeds.

4

One sole primary

Translation-normalized pairwise fine-morphology distance over future passages P6–P12, analyzed by family, with the effective-shape estimator predeclared beside it so the framing cannot be chosen after the results.

5

One hard question

Does Φ-forward again compress the future shape fan relative to backward, orthogonal, blind, and untouched controls?

The protocol has since been run literally as fresh-cohort-fine-shape-confirmation-v1: a fresh 512-seed cohort with byte-identical search parameters, endogenous peaks under a predeclared selection rule, five matched arms, and the P6–P12 shape fan as the sole primary. The primary came back −0.0019 with interval [−0.0076, +0.0038], six families negative and six positive, so the fine-shape compression is not confirmed at this sample size; twelve families exclude only effects stronger than about −0.008, and the original −0.0040 sits inside what this cohort could easily miss. Three companion experiments ran beside it under their own frozen plans, and the four verdicts together are reported in The Organism Hardens Without Narrowing.

fresh-cohort-fine-shape-confirmation-v1 · the confirmation

Sole primary phi-forward minus phi-backward over passages P6–P12: −0.0019 with interval [−0.0076, +0.0038], prediction not passed, and no arm contrast separates from zero. Twelve families were too few to settle it either way.

results.json · 5318aa930cd789d446372828dd0fc1bcf55edbeafc7b07d7868feb76d5bc77d4

steerability-age-profile-v1 · prediction met

The same standard push moves an organism less the older it is, with the age slope of divergence +0.0038 and interval excluding zero across sixteen families; divergence falls about 69 percent from the youngest to the oldest push age, and the fall tracks each organism's past-only turn.

results.json · 7f88560f446752adb31483f8f9e5f851a67069e97050d40393ab26784f98f949

basin-width-init-variation-v1 · wrong direction

The predicted contraction of the sibling fan with age resolved backwards: fresh-seed siblings spread apart monotonically, width 0.032 at step 60 to 0.079 at step 900, early minus late −0.038 with the interval entirely on the wrong side. Older siblings sit farther apart, so whatever commits them is not pulling them toward one shape.

results.json · f71fd77f180cd081ea9011970ac2e109ec7b25179b5b7982f467a8f40deed91e

history-disambiguation-curve-v1 · null with a resolved secondary

The matched-ratio primary stayed flat at 0.76 to 0.83 across ages, null. The pre-named look-alike count collapsed from 6144 pairs at step 60 to 1665 at step 780, Spearman −0.988, so organisms become individuals even as their sibling fans widen.

results.json · 6cb3e4a06fa6244bc2347d11543d0a54484105edb7fbd346c0e9202380bf68f1

Open the source-bound evidence ledger · 19 result artifacts

Result SHA-256 values bind the exact JSON read for this synthesis. Completion or manifest SHA-256 values are copied from each bound result artifact.

Experiment / report	Result path · SHA-256	Result status	Claim class	Completion / manifest binding
`organism-birth-signal-v1/before-the-creature-prospective-result.html`	`organism-birth-signal-v1/prospective-results.json` `c6e3c6634d04de86ee89f5d11dec423e936befe81e022240a82dbdbc85daf963`	`complete_fresh_prospective_replication`	prospective precursor replication without specificity · 64 specimens / 16 families; 54 complete primary windows	`prospectiveCompletion 2dbf4f19d65f53490906d111f8262e10502390e5213603199dcea371634e8559`
`developmental-basin-atlas-v1/before-the-organism-developmental-basin-atlas.html`	`developmental-basin-atlas-v1/results.json` `bf1e0f3a082f256747f2d80f054a9d590f34ae9db28a83a6f6b9aa81d2fd2b1a`	`complete_open_ended_developmental_basin_response_atlas`	open-ended causal atlas, event-aligned · 64 specimens / 1,280 trajectories	`fullCompletion 4f81f0418fc9a29da4033838c339c320d80ec1ebad38bc72982017ba82dc31e8`
`prospective-geometry-clock-v1/the-state-turns-before-the-form.html`	`prospective-geometry-clock-v1/results.json` `269d280de5e93b8b93882ff900c41fb1f2750a256dfb145e284cbeca2cf4d742`	`complete_fresh_prospective_geometry_clock`	future-blind prospective coordinate · 64 specimens	`fullCompletion 664cdad167dceff3fc7a880509ea32399187210bd667be862cd44be343e2a88a`
`geometry-turn-causal-test-v1/geometry-turn-causal-atlas.html`	`geometry-turn-causal-test-v1/results.json` `98a5aa04dba2c35aacd0035fbbb0747c8d792cffc276e7ff480d3f80172a693e`	`complete_open_causal_susceptibility_atlas`	fresh replication plus open causal atlas · 64 specimens	`fullCompletion 3e2106a47f2ae33ea466bde2f6e7dc8695d5207262f0dcbd40287946a4f2e21f`
`causal-repertoire-contraction-v1/the-future-closes-before-the-organism-appears.html`	`causal-repertoire-contraction-v1/results.json` `9fe83df68e0e2cef8ba487673e4bf1005cacc5bb1dbbf84c963b90d8d2602fd4`	`complete_fresh_full_causal_repertoire_test`	future-blind prospective repertoire forecast · 222 states / 1,840 futures	`fullCompletion c04c782477371fb6d504f919467f38ce280680d98157c48dfd52034d391de432`
`commitment-branching-intervention-v1/one-early-push-narrows-the-later-possibility-space.html`	`commitment-branching-intervention-v1/results.json` `2ccd2af0ed14f2f329ca31fc982f0e1e0d314681db43ba9fd2994c2e586dcf64`	`complete_fresh_within_ancestor_commitment_intervention_test`	direct nested causal intervention; timing primary null, field-repertoire secondary resolved · 30 sibling pairs / 540 long futures	`fullCompletion 4813d53663f893d47f7ba9a992057ab1b372412c6cb063a880a49583dbd00b70`
`initial-action-world-v1/plasticity-is-there-from-the-start.html`	`initial-action-world-v1/results.json` `99727f878bab51518dddfb618cec0cd98c15bb6fe8b3667f515adb6d933e4817`	`complete_fixed_age_q8_to_q32_action_world_curve`	fixed-age causal atlas · 39 organisms / 5,616 futures	`completion 7de5b00378d093bfa0dfa85fa4f4a2ce4164664436fc88765e0b777c2aa55359`
`spatial-scale-action-world-v1/organization-disciplines-the-action-world.html`	`spatial-scale-action-world-v1/results.json` `6b3a76b4eb347010d9cb14a9bf57d1c70c69f4612a1bf7fee1b41fb57537f186`	`complete_multiscale_spatial_ablation_of_q8_action_world`	exact direct spatial ablation · 39 organisms / 4 states	`completion c0b1e358be4043343104691412d9709e5b86a2cffbffd9f45a190a7e743c81c4`
`commitment-gate-controller-v1/the-gate-stays-open.html`	`commitment-gate-controller-v1/results.json` `7969f195c566d69717b1430095266bab6132472e42d380015e273a46e5b71fc6`	`complete_closed_loop_commitment_gate_control_and_release_test`	direct closed-loop control and release · 39 organisms	`controllerCompletion 12009169931a2b3ba330336293f006eb19bf9e9f904935419fc8704583e764fa`
`commitment-gate-formation-v1/the-hidden-state-becomes-a-body.html`	`commitment-gate-formation-v1/results.json` `b7fc337afa1f605e7de16a59c9e9797804b5039b29fa45f600529dd930d6aac3`	`complete_selected_future_visible_formation_analysis`	selected-future long follow-up · 39 organisms / 312 trajectories	`completion c0aa1de549c693b060ac885d63bcc63f9e9f7ba1cf2e9d43b5146bdc09500ad9`
`latent-composition-transplant-v1/the-body-stays-the-future-leans.html`	`latent-composition-transplant-v1/results.json` `785e801c2f7c8ad7feba79c25463f195efe55467677f8f04eb78bd28a1b68888`	`complete_hidden_composition_causal_transplant_analysis`	direct exact-field transplant · 39 organisms / 156 runs	`completion b3c1354772459d0f7824dcd567c551ca9162789c3c4034dd0c45fc7ea43fe294`
`developmental-ignition-transplant-v1/the-hidden-future-pushes-back.html`	`developmental-ignition-transplant-v1/results.json` `f6d6078b264fb0285a1dd8982fe9bd3a749f01a0ace89f766a8735297f822e40`	`complete_fresh_reciprocal_hidden_age_transplant_analysis`	fresh reciprocal direct intervention; frozen clock question resolved no · 64 organisms / 640 futures	`completion a20346a43025df0b7ef652d73acb1a8f3b934e7e60089959039a2d24bca7a728`
`geometry-turn-causal-test-v1/geometry-turn-causal-atlas.html`	`geometry-turn-causal-test-v1/open-discovery.json` `2074c1bd5abcfef399d5070f1d0b2f1c551ab8ec138d6d36e69bd58272a29753`	post-hoc open discovery	source of the 25% → 75% turn-prevalence reading · same 64-specimen cohort	bound to the atlas run above
`developmental-ignition-transplant-v1/the-hidden-future-pushes-back.html`	`developmental-ignition-transplant-v1/counterweight-results.json` `086729c395b6d6c06b79969a0eacb7fb615f8ec0a3ac7a333efb28bcbbfc2f33`	post-hoc second-pass analysis	source of the −0.096 growth-restraint reading · same 64-organism cohort	bound to the transplant run above
`developmental-causal-susceptibility-map-v1/the-future-narrows-here.html`	`developmental-causal-susceptibility-map-v1/results.json` `ef00428ea459e805f786f32fc7f399528c3cb589f1991ed6f39b0a9f20d37a00`	`complete_dense_organism_relative_causal_susceptibility_map`	fixed causal commitment map · 19 families / 2,736 futures	`manifest 45aace45f3fc07b211aa71a1930182205a869e276b38ae31784a8d215770cde8` `completion e9634826f44cf4e1e74c13bb8af6e31aca7bde21872b662dda7aa3131985ad70`
`developmental-causal-grammar-v1/the-organism-learns-what-actions-mean.html`	`developmental-causal-grammar-v1/results.json` `200e1c14daef2d965fff12ccb167ca70b7ecaad410393ab2a7d445401d920e17`	`complete_two_action_developmental_causal_grammar`	future-blind causal grammar · 19 families / 152 panels	`manifest dbb2a36d8d77d80d895f07f6f8795f42188aab27a4aaeab8e32dce75590f4dd9` `completion 3753423409b2f86c8df44da7bf4b7942c6049fb514cc2dd63962421d8409db54`
`tangent-future-aperture-v1/the-future-chooses-a-main-body.html`	`tangent-future-aperture-v1/exploration.json` `aa28dfa388708b7679acce9b7845f48922d9f1886e2bc2a733bef9198d943b23`	`complete_post_hoc_source_bound_exploration`	post-hoc direct-intervention lead; frozen primary unresolved · 19 families / 1,216 trajectories	`manifest f50ead16920b9749a029a728c946459d863cbf172ac1b9ab341a7acfbd60c8b2` `completion 29dad2bbd870ff2eff907f8b31a93e0904b1f31c95d1af602d5008db3588e69f`
`developmental-commitment-aperture-v1/the-shape-space-closes-late.html`	`developmental-commitment-aperture-v1/results.json` `abd5bbe063064cc4b1d1c55d331a51b4970d764412ab776bcd79010c236c948d`	`complete_source_bound_future_blind_analysis`	predeclared secondary direct result · 19 families / 3,648 trajectories	`manifest 9d123a6fff24d9675dd0c692c98a214299194b4edef0c15d64339403077eb952` `completion cc947fc2bf181f4f263312b0c7a7d0710afce04c2a61fd1e66ce84d92d356de1`
`developmental-commitment-aperture-v1/the-shape-space-closes-late.html`	`developmental-commitment-aperture-v1/morphology-timing-exploration.json` `e31b5f34e1dfc8f5573c8433c0ff6d05777f7fc52fe116a71cbbc140f47fec21`	`complete_source_bound_secondary_and_descriptive_analysis`	source-bound predeclared secondary and descriptive timing analysis · 19 families / P1–P12 panel	`manifest 9d123a6fff24d9675dd0c692c98a214299194b4edef0c15d64339403077eb952` `completion cc947fc2bf181f4f263312b0c7a7d0710afce04c2a61fd1e66ce84d92d356de1`

Can the whole forecast itself better than its best split?

From steering replication to watching an organism commit.

Silhouette splitting was the wrong event.

The flagship steering test failed.

More Φ was not a universal push.

Matched future fans instead of single endpoints.

Spatial structure at q8.

Several kinds of future, several schedules.

From prediction to causal closure.

The whole-over-parts signal rises before the body does.

Past-only geometry tracks developmental progress, but pushing along Φ does not advance the transition.

More whole-over-parts predicts a narrower timing future, and an early push leaves a small trace in later microscopic outcomes.

The action-world is already broad at q8, and the spatial arrangement is what gives it structure.

A brief feedback policy holds the causal gate open, then the state carries itself forward.

The same visible body can contain a different future.

After the information turn, morphology becomes harder to steer and actions acquire more stable meanings.

A Φ-directed push narrows different organism-level futures at different developmental moments.

Commitment is selective closure.

There is already a result worth sharing.

What we would claim

The causal-emergence connection

What we have not shown

The record

What did not survive, in the open.

reservoir-causal-steering-v1 · the flagship

tangent-induced-causal-grammar-v1

developmental-causal-plasticity-window-v1

fresh-organic-q40-replication-v1

causal-compass-peak-v1

Eight stopped runs

Freeze the shape claim as the only primary.

Fresh, unseen cohort

Endogenous timing

Five matched arms

One sole primary

One hard question

fresh-cohort-fine-shape-confirmation-v1 · the confirmation

steerability-age-profile-v1 · prediction met

basin-width-init-variation-v1 · wrong direction

history-disambiguation-curve-v1 · null with a resolved secondary