8.4 Cross-Probe “Dispersion-Free Common Term”: The First Verdict Line for Redshift and Time Delay

I. Section Conclusion.

If probes as different as supernovae, standard candles, strong-lensing time delays, strong-gravity transients, and extreme transients—probes that do not share the same instrument chain or even the same source physics—still leave behind the same common term after their strictest subtraction of dispersion terms, medium terms, and instrument terms, and if that common term does not fan out with frequency, points in the same direction across different carriers, and can be reproduced across different pipelines, then EFT’s redshift main axis is elevated, for the first time, from something one can merely say to something that deserves explanatory priority.

Conversely, if the so-called common term only looks good in a single band, flips sign whenever the bandwidth changes, disappears whenever the pipeline changes, and requires a separate set of exceptions to be invented for every source class, then EFT can no longer hide behind verbal advantage on this line. What would have to retreat is not just one pretty case, but the entire working discipline that says Tension Potential Redshift (TPR) sets the Baseline Color and Path Evolution Redshift (PER) only refines the details.

Verdict Card

• Core commitment: the cross-probe common term must simultaneously be nearly dispersion-free, point in the same direction, fall in the same window, and preserve the same ordering; any “pretty residual” that works only in a single window may not be promoted to a main conclusion.
• Primary readouts: residual strength of T_common after strict subtraction; agreement in leading sign and leading ordering across frequencies / carriers; event same-window behavior (zero lag or a pre-registered short lag); amplification after environmental grouping.
• Minimum resolvable effect: the main text should not force in a universal constant, but the pre-registration must state three thresholds—no sign flips, no ordering breakdown, and a common term that rises above the noise and permutation background of each pipeline. Anything below threshold may be recorded only as “unresolved,” not counted as support.
• Key artifacts and alternative explanations: plasma dispersion (1/ν²), Faraday rotation (λ²), dust scattering and absorption, bandpass / timestamp errors, microlensing and environmental-modeling degeneracies, sample truncation, and selection effects. Anything governed mainly by these laws goes back onto the path / instrument ledger and may not masquerade as a dispersion-free common term.
• Destination of null results: if a stable T_common never emerges across probes, this section does not blur the null result; it must be rewritten as an “upper bound on TPR’s shared Baseline Color,” an “upper bound on PER / path-term weight,” or a narrowed-domain conclusion in which “TPR is valid only in local windows.”

II. Why the First Hard Verdict Has to Land Here

Volume 6 has already set out EFT’s working order for redshift: read the endpoints first, the path second; inspect the main axis first, the scatter second; TPR carries the Baseline Color, PER handles the edge refinements. At the same time, Section 6.15 sharply separates “different cadence at emission” from “energy worn away on the road,” refusing to stuff every non-expansion redshift back into the old pocket labeled tired light.

That is why the first hard verdict line in Volume 8 cannot rest on whether one Hubble plot looks plausible, nor on whether one batch of supernova residuals can be talked into place. It has to go one step further and ask directly: do different probes all read out the same common term that does not fan out with frequency?

Because a single probe always leaves too many escape hatches. Supernovae can be dismissed as source-side complexity, lensing time delays as modeling degeneracy, transients as environmentally dirty, and local anomalies as small-sample bias. Only when these heterogeneous readout chains begin pointing toward the same common structure does EFT truly leave the stage of “isolated curiosities” and enter the stage of “cross-probe consistency testing.”

III. What Counts as a “Dispersion-Free Common Term”

“Dispersion-free” has to be defined carefully here, or this whole section will be bent out of shape immediately.

It does not mean that the world contains no scattering, no absorption, no line broadening, and no medium perturbations. It means that after those subtractions—subtractions that should be performed anyway—if a dominant common term still remains stably present, then that dominant common term should not govern the result in a frequency-selective way. In other words, it should not scale, flip, and reshuffle itself according to 1/ν², λ², or any other typical dispersion law; it should look more like a Baseline Color shared across readout chains than like a loss term that hits one frequency band especially hard along one segment of the path.

Accordingly, the “dispersion-free common term” discussed in this section must satisfy at least three layers of requirements.

• Same-direction consistency
  Residuals extracted from different frequency bands, different carriers, and different observational readout conventions should not flip their leading sign arbitrarily as frequency changes.
• Same-window coincidence
  In time-sequence observations, this common term should appear as near-zero-lag co-occurrence, or at least remain stably aligned within the pre-registered time window, rather than jumping to the other side whenever the band changes.
• Same ordering
  Even if the amplitude scales of different probes are not identical, the rank ordering of stronger and weaker signals should remain broadly consistent: which sightlines are stronger, which environments are more sensitive, and which subsamples are more likely to show the common term cannot be one ordering today and another tomorrow.

What really matters is not how large one particular number is, but whether these three kinds of consistency hold at the same time. Once all three stand together, the “common term” is no longer just a statistical leftover; it begins to look like a shared readout written by the Base Map itself.

IV. Why This Line Hurts EFT So Much

Because EFT itself has already separated the ledgers.

TPR keeps the endpoint-calibration ledger. The issue is not that light gets worn down along the way, but that the clock standards at the source end and the local end were never the same to begin with. PER keeps the path-evolution ledger. The issue there is not that light bleeds energy all the way, but that it crossed regions still undergoing extra evolution and therefore picked up finite edge refinements. Tired light is something else altogether: it presupposes a path-loss ledger—energy bleeding away all along the route, with side effects such as color dependence, blurring, broadening, polarization rewriting, and degraded coherence left behind at every step.

That is exactly why what EFT fears most is not being told, “you are not an expansion model.” What it fears is someone eventually proving that its supposed extra term is, at bottom, still just a variant of path fatigue. If that were true, EFT would have to pay the full side-account of path loss: why there is no stable color dependence, why there are no synchronized spectral scars, why there is no consistent polarization rewriting, and why there are no scattering-style fingerprints reproduced across probes.

So Section 8.4 is not asking only whether there is an extra term; it is asking what sort of term it is.
If it behaves like frequency-selective loss, EFT looks bad.
If it behaves like a non-dispersive Baseline Color shared across probes, only then has EFT truly separated TPR from tired light.

V. Why It Is the “First Verdict Line for Redshift and Time Delay”

Because redshift and time delay are the two kinds of readout most likely to preserve the same Baseline Color across different carriers.

Redshift records how cadence differences are read out by local Rulers and Clocks. Time delay records how arrival order is pulled apart in comparison. They look like two different kinds of quantity on the surface, but in fact both ask the same question: has the Base Map written the same common structure into different readout chains?

If EFT’s claim is right, then this common structure should not show itself on only one side. It should appear simultaneously as:

• On the redshift chain, the residuals can be read as a decomposition into “TPR Baseline Color + PER refinement,” not as a lump of patches drifting arbitrarily with source class.
• On the time-delay chain, after standard geometric and medium terms are removed, a stable, cross-frequency, cross-station, cross-method, dispersion-free common term still remains.
• In the joint comparison, redshift residuals and time-delay residuals need not have the same numerical value, but they should obey the same environmental ordering, the same subgroup amplification, and the same discipline of not following a dispersion law.

More concretely: on the one hand, two-station propagation scaling requires the common term’s timing offset to satisfy co-occurrence, linear delay with distance, and energy independence all at once; on the other hand, redshift decomposition requires the residuals to be writable as
Δz = z_TPR + z_PER, with TPR carrying a universal Baseline Color and PER occupying only the slot of discrete fine adjustments, not being forced to slide into a frequency-dependent dispersion law.

So the phrase “the first verdict line for redshift and time delay” does not mean crudely gluing two kinds of quantity together. It means that these are the two earliest windows through which the same Base Map can be jointly audited.

VI. Which Probes Are Best Suited to Carry This Verdict Line

The immediate task is not to dump every experimental detail onto the page, but to identify the probe families that can actually carry this verdict line.

• Supernova and standard-candle family
  Here the audit asks whether redshift residuals, luminosity residuals, width–luminosity relations, color-correction terms, and host-environment groupings still leave behind a stable common Baseline Color. These probes are not meant to close the case by themselves; they are meant to test whether TPR can really carry the main axis.
• Strong-lensing time-delay family
  Here the question is whether, after mass models, environmental structure, microlensing, and instrumental calibration have all been accounted for, the arrival-time differences of multiple images still preserve a common residual that is cross-frequency consistent and robust across pipelines. This is the core gateway for pulling “time delay” into the same audit framework.
• Microlensing and image-timing puzzle family
  What matters most here is not the light-curve variability itself, but whether complex light curves can be used to reconstruct a smooth common term that is nearly dispersion-free across frequencies and coincident at zero lag across stations. This family forces the question of whether the common term is a Base Map readout or a pipeline artifact.
• Strong-gravity transients and extreme-transient family
  This includes fast radio bursts (FRBs), gamma-ray bursts, tidal disruption events, and gravitational-wave / electromagnetic counterpart events. What makes them valuable is not the word “extreme” itself, but the fact that they provide short-duration, high-contrast, high-pressure windows with strong environmental differences, making it easier to separate dispersion terms from common terms.
• Solar-system common-source multipath and solar-grazing sequences
  These probes function more like a calibration court. They may not be the main cosmological battlefield, but they are exceptionally well suited to testing, with real strictness, whether a dispersion-free common term remains after de-dispersion, because both the geometric chain and the path chain are more controllable.
• Knife-edge occultations, lunar occultations, and controllable near-field events
  The value of these platforms is that they move common-term auditing from “waiting for heaven to hand us an event” to “building a high-pressure test field with designed controls.” They do not replace cosmology; they provide the methodological foundation for cosmology’s verdict line.

These probes are not laid out side by side as equals.
The first two families pull out the cosmological main axis.
The middle two pull high-pressure transients into the same language.
The last two harden, at the methodological level, the question of whether the common term is real at all.

VII. One Common Verdict Protocol for Different Probes

No field gets to judge this question in isolation, so the cross-probe protocol has to be fixed in advance. At minimum, it requires the six steps below.

• Freeze the standard subtraction terms first
  Dust, plasma, Faraday rotation, troposphere, ionosphere, instrument bandpass, timestamps, microlensing, environmental structure, mass-sheet transformation, beam curvature, template residuals … whatever should be subtracted must be subtracted first, and the standard must be frozen before the results are seen.
• Keep at least two frequency bands or two carriers
  Without frequency splitting or carrier splitting, one cannot talk about “dispersion-free” at all. A pretty residual inside a single band counts only as a clue, not a verdict.
• Accept only common terms that are same-direction across frequencies, same-window across stations, and robust across methods
  Even if amplitudes differ slightly, if the leading sign, the leading ordering, or the event alignment falls apart as soon as the pipeline changes, the result may not be promoted to a main conclusion.
• Explicitly exclude canonical dispersion laws
  If a result scales mainly with 1/ν², λ², or another known path-dispersion law, or flips sign as soon as the bandwidth changes, it goes back onto the medium ledger and may not masquerade as an EFT common term.
• Run null tests, holdouts, and permutations
  Label permutation, time reversal, station permutation, off-axis controls, reference windows far from the knife edge, holdout events, holdout stations, and holdout frequency bands are not accessories; they are part of the main verdict criteria.
• Across probes, compare structure only; do not demand the same numerical scale
  Section 8.4 is not trying to crush all probes onto the same absolute number. It is asking whether they share the same structural discipline: non-dispersion, same direction, same window, same ordering, and stronger signal under environmental grouping.

With those six steps in place, the experiments that follow no longer collapse into a contest of mutually insulated stories.

VIII. What a Result That Supports EFT Should Actually Look Like

Support here does not mean one striking figure in one paper. It means the simultaneous appearance of the following.

• After strict subtraction, multiple probes all leave behind a dominant common term that is nearly dispersion-free.
• Those common terms preserve the same direction and the same ordering across different frequency bands, different stations, and different processing chains.
• Residuals on the redshift chain can be written stably as TPR Baseline Color + PER refinement, rather than forcing PER into the main seat.
• Residuals on the time-delay chain show zero-lag co-occurrence across frequencies or an equivalent same-window structure.
• Environmental grouping works: along more extreme paths, in hosts at higher hierarchical levels, or in stronger lensing environments, the common term appears more strongly, more stably, and more predictably.
• All of these conclusions survive null tests, holdout sets, and cross-team replication.

At that point EFT cannot claim the case is closed, but it has at least earned the most important prize in the first round: explanatory priority.
It has shown that what it proposed is not a rhetorical flourish from one narrow domain, but a shared claim that can be developed across readout chains.

IX. What Kinds of Results Would Force EFT to Tighten

The verdict is not simply black or white. Many results would not kill EFT outright, but they would force a visible narrowing of its domain.

The following kinds of results should be recorded as tightening, not smuggled back in as “support anyway.”

• The common term appears only in one class of probe and remains absent over the long run when one moves to other probes.
• The common term holds only inside a very narrow environmental window and becomes unstable outside it.
• TPR’s Baseline Color coefficient cannot remain universal; different source classes each have to keep their own parameter set.
• PER keeps having its amplitude raised until it no longer looks like a residual slot at all, but instead eats into the explanatory space of the main axis.
• Dispersion-free behavior stands only under very specific pipelines and very specific subtraction standards, and drifts significantly as soon as the algorithm changes.

When results like these appear, EFT has not necessarily lost yet, but it must retreat honestly:
what was written as a “common Baseline Color” can only retreat to “locally valid”;
what was written as a “cross-probe main axis” can only retreat to an empirical regularity for special scenarios.

X. What Kinds of Results Would Strike Directly at the Main Axis

What really counts as structural damage is not “this figure does not quite look right,” but the stable, repeated, cross-pipeline appearance of the following situations.

• Systematic absence of the common term
  After strict subtraction, different probes consistently fail to show any stable, shared, dispersion-free residual.
• Results governed mainly by dispersion laws
  The supposed common term ends up scaling mostly as 1/ν², λ², or some other frequency-dependent law, showing that path-medium terms are the real main actors.
• Instability in leading sign and leading ordering
  One band is positive today and another negative tomorrow; one sample looks stronger today and the order reverses as soon as the pipeline changes.
• One rule set per source class becomes necessary
  Supernovae need one PER, lenses need another PER, and transients need a third PER, with no translation possible among them.
• Null tests and holdouts cannot break it
  If after label permutation, station permutation, frequency holdout, and time reversal the so-called common term remains significant at the same level, then it looks more like a pipeline artifact than a physical Baseline Color.

If several of these kinds of results hold over the long run, EFT can no longer maintain that redshift and time delay share one main line of a dispersion-free common term. What would have to retreat then is not just one case, but the status of Section 8.4 as the priority verdict line.

XI. What Still Cannot Be Judged Today

“Not yet judged” must also have boundaries, or it collapses into unlimited life support.

In this section, there are only three cases that genuinely deserve a “not yet judged.”

• Frequency coverage is still insufficient, so dispersion-free behavior cannot really be distinguished from weak dispersion.
• The standard subtraction terms have not yet been frozen, leaving too much model freedom and making it easy for common terms and system terms to be swapped into one another.
• Samples and signal-to-noise are still insufficient, so cross-probe evidence remains only scattered hints rather than a reproducible structure.

But once frequency splitting has been done, null tests have been done, holdouts have been done, and cross-pipeline checks have also been done, if the result still goes the opposite way, then “not yet judged” no longer holds. At that point the issue is not that “the instrument is still not good enough.” It is that the theoretical commitment is being weakened by reality.

XII. Section Summary.

What this section has to pin down is the first hard verdict line:

If multiple probes all read out the same common term that does not fan out with frequency, then it looks more like a shared cause rooted in the source end and the Base Map than like a frequency-selective loss accumulated along the path; conversely, if the so-called common term always fragments into one private version per probe and always has to be propped up by dispersion and patches, then EFT’s redshift main axis has to retreat.

That is why 8.4 comes before 8.5. Here the question is whether a dispersion-free common term is genuinely shared across probes. In 8.5 the question becomes harder: can that same source-end line survive a joint audit of the redshift main axis, the distance-calibration chain, and the local residual ledger, with PER still kept in the residual slot?