Illustrative pair · largest field contraction
ethereal-pattern-1326
The coordinate gap was 0.432. Mean pairwise final-field breadth moved from 0.234 in the lower branch to 0.178 in the higher branch; the timing range moved from 5 to 4 steps.
01 · See the branches
Nearby siblings, then different response fields.
These are actual descendants from the assay. We show the pair with the largest measured field-repertoire contraction so the visual example is legible, although the cohort estimate below—not this post-analysis example—carries the claim.
02 · The frozen assay
The causal fork and the measured contraction.
At q33, eight first actions branched the same q32 ancestor. At q36, a frozen portable coordinate selected the two extremes within that ancestor; at q37, both selected descendants received one continuation and the same eight second actions, and their nine final fields were compared pairwise at q900.
Negative values mean that the higher-coordinate descendant exposed a narrower set of final fields under the common nine-arm repertoire. The 5,000-draw bootstrap resampled ancestors, which was the frozen unit of analysis.
03 · All 30 pairs
The average contracted, although not every organism did.
Eighteen paired branches narrowed and twelve widened, while the mean effect remained below zero after leaving out each ancestor in turn. Seven of the eight small morphology-group means pointed in the same direction, with three or four ancestors in each group.
paired ancestors with a narrower higher-coordinate final-field repertoire; 12/30 widened.
group means below zero; the leave-one-ancestor-out means ranged from -0.0089 to -0.0061.
04 · The primary
Timing moved the same way, but the interval still crosses zero.
The predeclared primary was the paired change in transition-timing range, not the field metric. Its mean was -2.57 steps, with a 95% interval from -6.67 to +1.00; that leaves both a modest contraction and no timing effect compatible with this cohort.
timing ranges narrowed, while 6 were equal and 10 widened.
both branches eventually showed the transition under every arm, so event incidence itself did not change.
05 · Across groups
The contraction was distributed, not uniform.
The group view is descriptive because each group is small, but it makes the heterogeneity visible rather than smoothing it away.
Counterexample · largest timing reversal
radiant-blob-1636
The coordinate gap was 0.545. Mean pairwise final-field breadth moved from 0.472 in the lower branch to 0.449 in the higher branch; the timing range moved from 19 to 35 steps.
06 · What follows
The first action changed later microscopic reach; it did not settle timing.
Because the higher and lower descendants were produced by different selected first actions, this is a causal effect of the coordinate-guided policy on descendant repertoire, not an isolated causal effect of the coordinate itself. The assay also tracks developmental transitions rather than autonomous replication.
A first local action can steer an early organism into a sibling whose later responses to the same action family occupy a smaller mean field repertoire.
Whether repeated pushes turn that microscopic contraction into a robust timing effect, and whether the change persists after the driving policy is released.
The next experiment can compare one, two, and four coordinate-directed pulses, then replay the same actions open-loop after release; that would test whether the contraction scales into a stable developmental change instead of remaining a short-lived steering effect.