Rotation is real.
The preferred Φ/TDMI action directions rearrange sharply with developmental age.
Before the replicator · fresh 64-organism causal test
The future narrows in time,
not in form.
Early Flow Lenia organisms already hold a broad space of possible causal futures. Across q32, q40, and q48, that space stays almost exactly the same size—even while the local information compass rotates.
The direct result
No singular opening at q40.
We generated 2,048 fresh systems, selected 64 bounded organic forms using a classifier trained on the earlier human-reviewed atlas, mapped eight interventions at five early ages, then opened 1,536 untouched long counterfactual futures.
64fresh organisms
1,536long causal futures
+0.093q40 rotation ↔ repertoire curvature
-0.0023mean q40 repertoire curvature
The causal volume is conserved.
Mean future-field repertoire: 0.2588 at q32, 0.2565 at q40, and 0.2588 at q48. The q40 curvature interval is [-0.0069, +0.0023]. There is no population-wide bulge or collapse.
That is not an empty result. It says the developing form is already causally open by q32.
The organism does not wait until q40 to acquire alternatives. It carries a wide future repertoire throughout this window.
What changed in the theory
The compass is not a volume knob.
Expansion is not.
A more exceptional q40 compass turn does not reliably produce a more exceptional q40 future volume: ρ=+0.093, bootstrap interval [-0.178, +0.352].
So ask a better question.
Information geometry may tell us which futures become reachable, which arrive sooner, or which directions become canalized—not simply how much future exists.
1Broad final-form alternatives already exist.
2The information compass rotates inside that space.
3Transition timing becomes committed before endpoint variety collapses.
The cross-experiment synthesis
Temporal commitment precedes morphological closure.
In the earlier fresh full-repertoire test, higher whole-over-parts organization predicted a narrower range of transition times under the same eight-action inventory: partial-rank estimate -0.248, 95% interval [-0.420, -0.045]. But its association with final-field repertoire breadth was weak.
The emerging developmental variable is not “how many adult forms remain.” It is “how tightly scheduled has the route into them become?”
This is closer to canalization than simple loss of possibility: many eventual forms remain, while the timing and pathways into them become increasingly organized.
Representative causal futures
Same assay. Very different individual geometry.
The cohort result is stable, but individual organisms can show a strong q40 widening, a strong narrowing, or essentially no curvature. Pink points in the scatter mark organisms whose sharpest local compass turn occurred at q40.
organism 20mystic-amoeba-2656
q40 curvature +0.0567 · rotation excess +0.071
organism 01flowing-cell-2320
q40 curvature -0.0621 · rotation excess -0.175
organism 10vibrant-orbiter-3596
q40 curvature -0.0003 · rotation excess -0.278
organism 07mystic-walker-2053
q40 curvature +0.0022 · rotation excess +0.476
The next direct swing
Track paths, not volume.
The next experiment should preserve action identity across ages and ask whether the information compass predicts the orientation and timing of downstream trajectories: which intervention reaches which morphological basin, how quickly, and whether that mapping stabilizes after the visible organism appears.
Transport the same perturbation.
Use state-corresponding spatial/material coordinates rather than reseeding the eligible-cell shuffle at every age.
Measure route geometry.
Compare transition time, basin identity, and path topology—not just pairwise endpoint breadth.
Look for stabilization.
If causal emergence is developmental commitment, the action-to-future map should become more persistent even while multiple endpoints remain available.