transductive.science

Framework principle · 16 September 2026

The Symmetric Burden Principle

Alternative theories should not be required to achieve what established theories are allowed to leave unresolved.

Take every objection seriously. Then apply the same criterion, at the same explanatory level, to the incumbent baseline.

Alternative theories should not be required to achieve what established theories are allowed to leave unresolved

Transductive Science — Framework Essay — 16 September 2026

A new physical model is often asked to pass a strangely absolute examination.

Derive every important constant. Reproduce every established prediction. Supply a complete ontology. Prove every bridge between scales. Explain the detector from first principles. Avoid uncontrolled approximations. Avoid fitted parameters. Resolve every mathematical pathology. Match decades of precision before anyone is obliged to take the model seriously.

Each demand can sound reasonable in isolation.

The problem appears when the same demand is not applied to the framework already occupying the field.

That asymmetry is not rigor. It is asymmetric rigor: one framework is judged as a historically accumulated research programme, complete with approximations, fitted inputs, unresolved foundations and open problems, while its challenger is judged as though it must arrive already finished.

The Symmetric Burden Principle is a simple corrective:

Whenever an objection is raised against a challenger, take the objection seriously — and then apply the same criterion, at the same explanatory level, to the incumbent baseline.

This does not make competing theories equally good. It does not excuse failed predictions. It does not imply that every unconventional idea deserves attention. It says something narrower and more demanding: no criterion may function as a veto against challengers while remaining optional for incumbents.

The hidden gate

Established theories are not single equations floating outside history. They are mature ecosystems. They inherit generations of mathematical technique, experimental apparatus, calibration practice, software, specialist training, institutional memory, numerical tables, approximation methods, and accumulated repairs.

An alternative usually begins with almost none of that.

If entry requires the alternative to reproduce the mature ecosystem and solve foundational problems that the ecosystem itself leaves open, the requirement is not merely high. It can make alternative development structurally impossible.

This creates a vicious cycle. A candidate is denied attention because it is immature. Because it receives little attention, it receives little engineering, data, formalization or experimental development. Its continued immaturity is then cited as evidence that it never deserved attention.

The result can look like intellectual selection while partly reflecting accumulated-resource selection.

Same standard does not mean same score

The Symmetric Burden Principle is not a demand for false equivalence.

If an incumbent predicts a quantity to twelve decimal places and a challenger misses by ten percent, that difference matters. If an incumbent survives an experiment and a challenger fails it under a frozen test, the challenger has a real failure. Symmetry does not erase performance differences.

What symmetry changes is the meaning of objections.

If the objection is, “your model contains fitted constants,” the relevant comparison is not challenger-with-fitted-constants versus an imaginary incumbent-with-none. The comparison is: which quantities does each framework take as inputs, which does each derive, and how much independent evidence constrains those inputs?

If the objection is, “your scale bridge is not rigorously proven,” ask which scale bridges or foundational limits in the baseline are themselves proven, numerically supported, phenomenological, conjectural, or unresolved.

If the objection is, “your measurement story is incomplete,” the baseline's own measurement ontology belongs on the same table.

The principle does not guarantee the challenger wins. It guarantees that the contest is real.

Put the baseline under the same lights

Several examples make the asymmetry visible.

The Standard Model is extraordinarily successful, yet CERN itself notes that we still lack an understanding of the physical origin of its many free parameters. A challenger should therefore not be dismissed merely because it has not yet derived every numerical constant from deeper primitives. The proper question is comparative: what does each framework assume, fit, derive and predict?

Four-dimensional quantum Yang–Mills theory underlies modern particle theory, yet the rigorous existence and mass-gap problem remains unsolved. The Clay Mathematics Institute explicitly describes the mathematical foundation as still unclear. That does not make Yang–Mills useless. It demonstrates that an unresolved foundational derivation is compatible with a productive research programme.

QED achieved finite and spectacularly accurate predictions through renormalization after ultraviolet infinities had threatened the theory's development. Whatever view one takes of renormalization, its history makes one principle unavoidable: the mere existence of mathematical repair machinery cannot by itself disqualify a competing framework. The fair question is what machinery each representation needs, what it explains, and what remains dependent on fitted input.

Quantum theory also retains a genuine measurement problem and multiple competing accounts of what its formalism says exists. Again, that does not negate its predictive power. It means “supply a uniquely settled ontology before we take you seriously” cannot be imposed selectively on challengers.

And the Standard Model itself does not incorporate gravity and leaves major phenomena unexplained. Incompleteness is therefore clearly compatible with enormous usefulness.

These examples do not prove that an alternative theory is correct. They prove that unresolved foundations, fitted inputs, approximation machinery and limited domains of validity are not, by themselves, legitimate one-sided disqualifiers.

Baseline, not boundary

Transductive Science therefore treats existing physics as a baseline corpus, not a constitutional boundary.

The baseline contains an immense amount of hard-won constraint: experimental records, numerical regularities, successful predictive structures, mathematical results, engineering knowledge and documented failures. An alternative physical model should attempt to reproduce as much of that territory as it can, especially where the evidence is robust.

But reproduction does not require inheriting the incumbent ontology.

A new model is not obliged to contain the same primitive objects, the same decomposition into fields and particles, the same intermediate infinities, the same renormalized quantities, or the same vocabulary of explanation. It owes an account of the resistant phenomena, not ritual duplication of the incumbent's internal furniture.

When a baseline result depends strongly on interpretive or model-dependent reconstruction, that dependency should be exposed rather than silently promoted into an ontological requirement.

The developmental asymmetry

There is another asymmetry that is easy to miss: maturity itself is cumulative.

An established framework may have had eighty years, thousands of researchers, bespoke instruments, specialist journals, graduate programmes, industrial-scale compute, curated datasets, and generations of debugging. An alternative may begin with one person, one model, and one workstation.

It is therefore unreasonable to confuse “not yet mature” with “demonstrated impossible.”

This does not justify lowering the final empirical burden. It changes the developmental question. Early alternatives should be allowed to earn progressively stronger forms of survival:

  1. demonstrate that a supposedly impossible construction is actually constructible;
  2. reproduce one robust phenomenon quantitatively;
  3. reproduce several independent phenomena with the same frozen mechanism;
  4. reduce the number of separately fitted quantities;
  5. survive tests not used during construction;
  6. produce discriminating predictions;
  7. eventually compete with the baseline in the regions where the baseline is strongest.

Demanding step seven before permitting steps one through six is not methodological toughness. It is a way of preventing competitors from reaching maturity.

The symmetric burden test

For every serious objection, record five things.

1. The criterion. What exactly is being demanded? Prediction precision? Derivation of a parameter? Mathematical proof? Ontological completeness? Computational tractability? A scale bridge?

2. The target level. Is the criterion being applied to an observable, an effective description, a microscopic mechanism, a detector model, or an interpretation? Comparisons across different levels are often misleading.

3. The challenger status. Does the challenger satisfy the criterion, partially satisfy it, or leave it unresolved?

4. The baseline status. Apply the same question to the incumbent without changing the vocabulary to make its unresolved point sound more respectable.

5. The consequence. Does the difference actually discriminate between the frameworks, expose a maturity gap, identify an open problem on both sides, or reveal that the proposed objection was never a legitimate one-sided veto?

A short form is:

Objection → decompose the criterion → apply it to challenger → apply it to baseline → compare at the same level → preserve the asymmetry if it is real.

The last clause matters. The point is not to erase differences. The point is to discover the real differences.

No rescue by symmetry

The principle also cuts against alternative theories.

A challenger cannot answer every failed prediction by saying that mainstream physics also has unresolved problems. A frozen failure remains a failure.

If the challenger rejects an auxiliary assumption to rescue its mechanism, it inherits the consequences of that rejection everywhere the assumption mattered. It cannot replace a scattering interpretation in one inconvenient experiment while continuing to borrow the old interpretation everywhere else without declaring the inconsistency.

Symmetric burden therefore comes with propagated burden:

You may challenge any assumption, but you inherit the full dependency cost of changing it.

This prevents the principle from becoming a universal escape hatch.

Why this matters beyond fairness

The larger issue is the ecology of inquiry.

A field that makes alternatives pay impossible entrance costs can become increasingly good at refining one family of representations while becoming increasingly poor at generating genuinely different ones.

That matters especially in an AI era. It is becoming cheap to construct, simulate and attack candidate mechanisms. The historical reason for requiring enormous prior disciplinary investment before a model could even be made concrete is weakening.

If institutions retain an asymmetric burden structure, AI may produce millions of possible alternatives while almost none can survive long enough to accumulate the refinement required for serious comparison.

The result would be an abundance of candidate generation combined with a bottleneck of conceptual permission.

Transductive Science is designed to attack that bottleneck.

A constitutional rule for Transductive Science

The Symmetric Burden Principle should therefore sit beside Evidence Debt as a core constitutional rule.

Evidence Debt: open entry does not mean free survival. Every retained candidate accumulates obligations to expose assumptions, implementation, parameters, provenance, failures and verification boundaries.

Symmetric Burden: those obligations apply to incumbent baselines too whenever they are used to reject a challenger.

Together they produce a much healthier rule:

Open the gate. Raise the evidential price. Charge the same currency to everyone.

No alternative receives immunity because it is unconventional.

No incumbent receives immunity because it is established.

Same floor. Same lights.

You want mathematical rigor? Good. Apply it symmetrically.

You want every fitted constant counted? Count the incumbent's too.

You want every scale bridge justified? Label the baseline's bridge status too.

You want every unresolved measurement assumption exposed? Put all of them on the table.

You want the alternative to explain exactly what exists? Ask what ontological claim the incumbent has actually established, rather than what its notation encourages us to imagine.

You want the challenger to reproduce the world? Absolutely.

But do not require it to reproduce every unresolved problem of the framework it is challenging.

Same floor. Same lights. No inherited exemptions.

Framework note

This article defines the Symmetric Burden Principle / Symmetric Burden Test for Transductive Science. It is intended to be applied whenever established theory is used as a baseline against an alternative model.

The principle is not a claim that all theories are equivalent. It is a procedure for preventing asymmetric standards from deciding the contest before evidence does.

Sources and starting points