Observed agreement does not determine a causal mechanism. SMM/DTT must derive stable dependencies among completed records and identify which interventions its source dynamics allows.
Three questions that causality can name
Why does a world exist? Why did one physical state change into another? What would happen if an experimenter altered one variable? These are questions about source dependence, dynamical mechanism, and intervention. They can be related without being interchangeable.
SMM principally motivates a common origin of physical multiplicity. A functional theory must also explain the ordinary dependencies by which its inhabitants act and infer. DTT adds the question of which generative dependencies survive completion as accessible causal relations. A statement that everything depends on one source does not by itself determine which laboratory manipulation changes which detector outcome.
A correlation with two different explanations
Let U be a fair binary variable. Model one assigns X=U and Y=U. Model two assigns X=U and Y=X. In passive observation both produce only (0,0) and (1,1), each with probability one half. Their observed joint distribution is identical.
Now intervene by setting X=0 while leaving U’s distribution unchanged. In model one, Y still equals U and is zero only half the time. In model two, Y becomes zero with certainty. The example demonstrates how a controlled intervention can distinguish a common cause from a directed influence when the intervention is correctly represented. The inference depends on the model of what the manipulation changes.
Earlier records are not automatically causes
A thermometer reading may change before an expanding rod is measured, yet both can respond to a common temperature change. The order of recorded detections also depends on travel times and instrument delays. Temporal precedence alone does not identify the mechanism.
An inner update ordering is similarly richer or different from the observer’s causal ordering. A source operation may precede a record without being separately controllable by the observer. The model should define its allowed interventions and their locations in the dynamics. This is especially important when a proposed hidden construction has an order that is absent from the effective relativistic description.
Repeated success and the domain of a law
If repeated preparations give a stable outcome frequency, that supports a probabilistic model within the tested conditions. It does not logically prove that the same rule holds under every possible preparation. The inference relies on assumptions about sampling, environmental stability, and relevant dependence between trials.
This leaves a positive strategy: estimate parameters in one set of conditions, freeze them, and calculate outcomes in others. A model earns explanatory reach when those additional results agree. A recreation account can permit unique complete events while retaining the same effective relation across a controlled class. Exact duplication of the entire universe is not required for reproducible science.
Make renewal carry the causal dependence
Write a source update as sn+1=F(sn,an), where a denotes an allowed intervention. Completed records are rn=R(sn). The model can calculate the consequences of replacing one action sequence by another, including effects delayed until later completion.
If the visible record omits predictive memory, the apparent transition law may depend on earlier records. Enlarging the retained state can restore a local description. SMM/DTT should explain this relation rather than declare all intermediate causes unreal. Effective causes can remain reliable descriptions of how interventions propagate through the shared source under a specified observation interface.
What a common-source explanation would add
One useful target is a relation among several causal responses that cannot be adjusted independently. If one parameter fixes the clock response, a transport delay, and a record’s relaxation rate, an experiment changing that parameter tests their joint dependence. Simply assigning each response its own temporal mechanism would provide less unification.
Read A69 with A82–A83 on future tests, A147 on projected memory, and A190–A194 on operational distinctions. Historical discussions such as al-Ghazali’s analysis of fire and cotton sharpen the difference between observed association and a claim of necessity. The modern intervention example here is a separate construction, not a mathematical result attributed to that historical text.
Continue with the articles
- Causality and Induction
Causality concerns dependence, while induction concerns inference from observed cases. Their distinction clarifies why temporal order and repeated records need a lawful generative relation before they support explanation and prediction.
- When two histories count as the same observation
Completion identifies histories that leave the same record, while observation identifies records that available tests cannot distinguish. Their combined quotient is consistent only with the declared operations and contexts.
- Test languages and observable quotients
A test language determines which records an observer can distinguish. Enlarging that language refines the observable quotient and reveals when apparently different instruments retain the same information.
- Eliminating hidden variables creates memory
Eliminating a relaxing hidden sector yields an exact memory equation for the retained variables. The resulting kernel and initial-state term preserve causal dependence while exposing information absent from the current readout.
- When memory can be represented by one time
Finite auxiliary states can generate memory while evolving in one laboratory time. Rational-kernel realization identifies when that reduction is exact and which additional conditions make the auxiliary system a physical reservoir.
- Empirical equivalence is not full structural equivalence
Identical prepared behavior does not require identical full bundle structure. Concrete counterexamples separate empirical equivalence, connection-preserving trivialization and changes of the laboratory clock.
- How a protocol can reveal dual-time structure
A temporal signature must be attached to a preparation, control sequence and readout that exposes retained information. Loop-response and predictive-state examples give exact witnesses while specifying the reduced model classes they can exclude.
- A checklist for claims of two physical times
Claims about two physical times need operational definitions, a complete interface and an explicit comparison class. This checklist combines minimal-state reduction, memory and holonomy tests with noise-aware requirements for an observable distinction.