The organism's own temporal direction is causally special.
This is no longer “any hidden rewrite changes development.” Forward and backward moves used the same invisible dose. The meaningful difference was alignment with the organism's own recent history.
Direction controls form
At +256, forward remains more compact than reverse in 13 of 16 families.
Direction controls possibility
At +256, the same four perturbations spread farther apart after forward than reverse.
Not just a sideways shove
Forward exceeds the equal-energy orthogonal-forward control; 95% CI +0.0015 to +0.0156.
Forward and backward begin invisibly. The body then chooses different geometry.
The native forward-minus-reverse radius separation appears by +64, stays near −0.10 through +256, and remains visible at +384. The orthogonal sign contrast is smaller and less coherent across time.
Across all five shared perturbation branches, the native signed radius effect exceeds the orthogonal signed effect at +128 by -0.084 [-0.137, -0.028]. The private direction survives nearby disturbances.
The forward state becomes more open—at a particular developmental time.
The repertoire effect is not a permanent scalar property. It is small early, peaks at +256, and relaxes by +384. That looks like a developmental window: the forward move changes when the organism becomes most susceptible to branching futures.
Forward vs reverse
Relative increase from 0.2127 to 0.2209 in mean pairwise future-field distance at +256.
Forward vs orthogonal
The native direction opens more future spread than an equal-dose scrambled direction.
Dose has structure
Double-minus-half native dose on q256 branch-mean radius; 95% CI -0.117 to -0.004. The response is directional and nonlinear.
Same young body. Same four perturbations. Different spread.
Representative source 23, chosen mechanically as the organism nearest the cohort-median q256 forward-minus-reverse repertoire effect. The starting visible field is identical across all conditions.
native forward · +256
native reverse · +256
Each row contains the unbranched future plus the same four local coupled-cell-swap branches. Aggregate repertoire is the mean of all ten pairwise visible-field Hellinger distances among those five futures. The claims use all 64 organisms and 16 families; these frames are a median illustration.
A developmental state has a direction—and that direction controls its causal horizon.
The organism's hidden trajectory does not merely describe where it has been. Extrapolating that trajectory changes both what body appears and how many nearby futures remain available later.
The causal-emergence connection
We have found an early, state-derived coordinate with interventional consequences for later organization and later susceptibility. This is the kind of precursor Levin's result points toward: information geometry before visible transition.
The sharper next question
Does this causal aperture peak before fission, identity turnover, ecological succession, or another spontaneous morphological event—and can its timing forecast or trigger that event?
Next swing: repeat the native-tangent intervention at several ages—q24, q32, q40, q48—and map the moving window where forward motion maximally opens the future. That gives us a developmental clock made from causal possibility, not just morphology.