Represented futures unquestionably exist as present physical models. The open question is whether that is the whole story—or whether a precisely defined later boundary also enters the probabilities of actual histories.
Possibility is content. Causation requires an actual carrier or law.
The blue option cone is a present model’s representation. It is not automatically a set of physically coexisting worlds and does not push the present by itself.
solid · actualbranch · representeddash · conjecturereturn · present model
02 · The three branches
No agreement bonus. Each branch can lose.
The branches share one observation—the present contains models of alternatives—but explain its consequence differently.
F5-W · [I/C]
Present-model floor
Claim. Represented futures influence present action through actual models, signals, memories, institutions, and computations.
Mechanism
A D4 model token represents and reweights D5 alternatives; ordinary dynamics carries the change into action.
Strongest rival
Salience, reward, demand, memory, expectancy, reactive control, or ordinary planning explains the change without a distinct D5-capacity variable.
Discriminator
Intervene on represented content while holding relevant present means and confounds fixed; require incremental prediction beyond rich baselines.
Kill
A fair-baseline null, or no independently measurable future-content variable.
Survivor
Ordinary causal anticipation through present carriers.
F5-N · [D/C]
No-additional-interaction null
Claim. No additional interaction couples to representation as such; selection, verification, correction, and revision service D5.
Mechanism
Known physical interactions plus biological, cognitive, social, and verification mechanisms exhaust the effect.
Strongest rival
F5-R predicts a preregistered residual dependence on later boundary data after ordinary channels are excluded.
Discriminator
Compare a frozen accepted-physics and control baseline with the candidate history law on held-out intervention data.
Kill
A replicated force-specific residual with defined coupling and observable, in the tested range.
Survivor
Verification and revision remain necessary practices.
F5-R · [C] UNVALIDATED
Future-boundary history-selection wager
Claim. An actual D4 history-selection law indexed by declared later physical boundary data and a D5 option functional may alter present transition probabilities.
Mechanism
A two-boundary or weighted-history law changes history weights. No possible future is treated as an agent.
Strongest rivals
Accepted quantum or classical dynamics, time-symmetric redescription, present control, common cause, selection bias, superdeterminism, or postselection.
Discriminator
After an early outcome is irreversibly recorded, randomize a later physical boundary intervention without ordinary signalling or postselection; test the frozen residual.
Kill
No exact law or distinct observable fails admission. A preregistered null kills its promised parameter or range. Observational equivalence is UNDERDETERMINED.
Survivor
F5-W and F5-N remain viable; the option ethic remains chosen rather than physical.
03 · Candidate law
A compact wager with an explicit zero.
This expression defines the comparison; it does not certify that nature uses it.
Pκ[h | I, BT] = Zκ−1 P0[h | I, BT] exp(κ U5[h; BT])
h
A complete, declared D4 history in the comparison ensemble.
I
Declared initial or preparation data.
BT
Physical boundary data on a declared later boundary ΣT; not a future causal cone.
P0
The frozen accepted-physics and control baseline.
U5
A declared option or viability functional over histories. It creates no value premise by itself.
κ = 0
The no-F5 arm.
κ ≠ 0
The strong F5-R wager.
Zκ
Normalization over the declared history ensemble.
Admission gate. “Force” remains withheld until the proposal supplies a physical coupling, units, carrier or boundary source, conservation accounting, Lorentz behavior, range, and a distinct observable. Until then it is a candidate history-selection law.
04 · Five independent wagers
A result for one does not promote the other four.
This separation prevents physics, cosmology, information, and ethics from laundering evidence across domains.
F5-P1 · boundary dependence
Present probabilities depend on declared later physical boundary data beyond signalling, common cause, and postselection.
F5-P2 · new interaction
Any dependence requires genuinely new dynamics rather than an existing time-symmetric or variational formulation.
F5-P3 · ontic plurality
Alternative histories physically coexist rather than merely appearing in a model or measure.
F5-P4 · option bias
The candidate law favors a declared option-preserving or ektropic class.
F5-P5 · normative teleology
Agents ought to preserve durable mutual option space under Justice and Exit.
A represented history bundle does not establish ontic plurality. A measured later-boundary dependence would not automatically establish a new interaction. No physical result by itself derives the chosen ought.
05 · Entropy and option separation
“More potential” is not one number.
The decision surface is a Pareto frontier unless a distinct measurement and value contract earns a scalar.
Dimension
Bearer or system
Required accounting
Agent options
Declared acting agent
Reachable and distinguishable actions, not merely imaginable states
Affected-whole options
Every materially affected bearer
Options widened, narrowed, transferred, or irreversibly removed
Subsystem thermodynamic entropy
Named subsystem
State variables and boundary; no moral metaphor
Total entropy production and export
System plus environment
Where dissipation and displacement go
Path or history entropy
Declared probability measure
Distributional diversity, not physical-world count
Empowerment or control capacity
Agent-environment channel
Horizon, channel, intervention set, and estimator
Viability cost
Maintained organization
Energy, attention, repair, and externalized burden
Consent · Justice · Exit
All consequence bearers
Authorization, reversible refusal, concentrated harm, and remedy
Chosen orientation: widen durable, bearer-complete option space under Justice and Exit. This remains a normative wager even if F5-R survives.
06 · Contact design
The later intervention must not rewrite the earlier record.
A credible discriminator must separate a residual from signalling, selection, postselection, common cause, and flexible analysis.
freeze baselineprepare systemrecord early outcomeseal recordrandomize later boundaryopen held-out resultkill or revise
A candidate design must specify the physical system, initial data, later boundary, separation against ordinary channels, preregistered statistic, power, negative controls, and the exact κ range it promises. The early record must be irreversibly bound before the later randomization, and all exclusions must be declared before contact.
A null in the promised range kills that parameterization. A residual that is equally predicted by an existing time-symmetric model does not establish a new interaction. Replication must come from independent custody.
07 · Neighbors before novelty
The wager enters an old and technically serious conversation.
Resemblance establishes a prior-art obligation, not support.
A “strange attractor” enters the physics only after a state space, flow, metric, invariant set, basin, and attraction test are declared. “Transcendental object” may remain metaphysical language, but it cannot substitute for those objects in a scientific claim.
08 · Current disposition
A sharpened conjecture, not an observed fifth force.
The achievement at this stage is separation: three branches, five propositions, a typed ledger, and tests able to return null.
Retained
F5-W floor
Present models can represent futures and alter present action through actual carriers.
Live null
F5-N
No additional interaction is required unless a distinct, replicated residual defeats the baseline.
Open wager
F5-R
Candidate law only. No truth bonus, no external validation, no empirical observation claimed.