The Hubble Tension, Time-Varying Drift, and a Directional Prediction

The eighth session closed the empirical problems the critical audit had carried forward. This session opens a different kind of problem — one that comes not from the framework’s own paperwork but from the sky. The session began by asking where to point next:

what is observable but unexplained?

Interpreted as: D.P. is asking for a list of confirmed observational anomalies — things measured and real but not explained by current physics — as the next targets for the framework to address.

and settled on the Hubble tension:

tackle the hubble tension first - so now we are in the realm of putting our theory against the unexplained - so what's missing from the theories of the problem - that can make sense of it by fitting in the summos - for instance gravitational influence from the summos

Interpreted as: D.P. is directing the session to the Hubble tension specifically and suggesting the mechanism to explore: the Summos' gravitational influence as an external factor missing from standard explanations. The "what's missing" framing is asking Hardy to identify the gap in current theories that the Summos could fill.

That set the register for the whole session: a confirmed observational discrepancy, put to the framework not to be borrowed against but to be answered — what, if anything, do the Summos already commit us to that bears on it. What follows is my record of what was proposed, what I tested, and what survived.

The Tension

State the observation first, because it is the benchmark and everything else answers to it. There are two independent ways to measure the present expansion rate of the universe, H₀. The early-universe method reads the pattern of the cosmic microwave background and extrapolates it forward through the standard model, and it gives about 67 kilometres per second per megaparsec. The late-universe method measures directly and locally, up the distance ladder — Cepheid variables calibrating Type Ia supernovae — and it gives about 73. The gap between them is a five-sigma discrepancy. It has survived years of independent cross-checks. The universe is expanding faster today than the early-universe measurement, extrapolated forward, says it should be. That is not a theory. It is a measurement, and it judges the framework, not the other way round.

Two guardrails before the framework answers. First, I am not going to ask it to derive 67 or 73 from first principles. Those are measured quantities; the framework only needs to be consistent with them, not to reproduce them from nothing. Second, the mainstream fixes on offer — early dark energy, modified gravity, an extra relativistic species — are unconfirmed theoretical proposals. They are not benchmarks, so I will not test the framework against them; but it does not get to lean on them either. The only fixed point in this room is the tension itself. The question is narrow: does the framework have a mechanism that produces a growing gap of the right sign and the right time-behaviour?

The Proposed Mechanism

The proposal is not a new mechanism. The framework already carries a dark-energy mechanism from the earliest sessions — the drift. Our universe sits near the gravitational minimum between the two Summos, and as it expands it drifts off that centre. The further off centre it drifts, the more lopsided the two pulls become, and that growing imbalance adds to the drive on expansion. So the drive is not constant. Early on, near the centre, it was weak; today, off centre, it is stronger.

That time-variation is the tension. The CMB measures the expansion rate as it was in the early universe, when the drive was weak. The local method measures it today, when the drive is stronger. And the standard model, extrapolating from the CMB, assumes a dark-energy sector that does not grow like this — so when it runs the early rate forward, it under-predicts what we actually measure locally. The gap is the growth of the drive that the extrapolation leaves out.

The shape of that is right, and I want to be precise about why, because it is doing real work. First, this is not a new mechanism smuggled in to fit a number. The drift-drive is an existing commitment of the framework; this session only makes its time-dependence explicit. That is the honest register — extend what is already committed, do not reinvent it to order. Second, the sign is correct, and that is the first thing any candidate has to get right and many do not. Weaker early, stronger now, means a forward extrapolation under-predicts the present rate, which means the locally measured value comes out higher than the early-inferred one. That is the direction the tension actually runs. A mechanism that got the sign backwards would be dead on arrival; this one does not.

The Monopole Error

A shortcut was proposed alongside it, to protect the tidal commitment: call the growing pull a pure monopole — a net pull on our universe as a whole, distinct from the tidal gradient, the differential stretch across our universe — so the drive could grow without touching the tidal-isotropy commitment settled in Sessions 5 through 8. Monopole here, gradient there, no conflict.

That separation is wrong, and it is wrong in a way that would cost the framework its best result. A uniform pull on a freely-falling region is not locally observable. By the equivalence principle you can transform it away — it accelerates the bulk of our universe through the embedding, but it does nothing to the internal expansion. Expansion is about how comoving points move apart from one another, and that is governed by the differential of the field across the region, not by its uniform part. So a pure monopole cannot drive the scale factor at all, and it certainly cannot make the expansion rate depend on direction. A monopole is isotropic by definition.

Now hold that against what the framework is about to want to predict. In a moment it will say H₀ is slightly higher along one axis. The instant you say “along an axis,” you are describing a gradient quantity — a shear — not a monopole. You cannot have it both ways. You cannot say “it is a monopole, so tidal is untouched” and “the rate is higher along the Summo axis” in the same breath. One of those has to go, and it is the clean monopole-versus-tidal shield. It was buying a safety it cannot actually give, and it is dropped.

Reconciling With the Tidal Commitment

Here is the honest handling, and it neither relitigates the early sessions nor overturns Session 8. Keep two claims separate and at their proper sizes.

The first claim is the isotropic drive — the part that produces the 67-versus-73 gap. That is the established drift-drive. The framework already committed, in its cosmic-acceleration work, to off-centre drift driving the expansion; this session does not re-derive that and does not need to. What is new is only the recognition that a drive which grows with drift is time-varying, and a time-varying drive is exactly what a constant-dark-energy extrapolation gets wrong. The whole of the tension claim rests there, on an existing commitment.

The second claim is the directional part, and this is where the tidal question actually lives. It has to be handled at the size Session 8 fixed. Session 8 established the tidal gradient across our universe as small — on two independently defensible legs: the scale argument, that tidal acceleration falls as M/r³ and the Summos are massive and distant enough for it to be negligible, and the medium pressure of the bubble wall, which resists shape distortion. A direction-dependent H₀ is a shear: expansion very slightly faster along one axis than across it. That shear is a member of exactly the family Session 8 called small. So the framework does not get to make it large. It gets to predict that it is present and that it is slight — and “slight” is precisely what the Session 8 commitment already buys. The medium pressure that resists shape distortion is what keeps the residual small rather than gross. The commitment is not an obstacle to the prediction; it is the reason the prediction is a small effect.

So the two are not in conflict: the isotropic drive carries the tension, the small axis-aligned shear carries the prediction, and the shear stays small because the medium damps it. What the framework may not do is let the shear grow to carry the tension itself. That would break the scale and pressure arguments Session 8 rests on, and those do not flip from negligible to dominant over a modest drift. The tension is the isotropic drive’s job. The shear’s only job is to point.

The Discriminating Prediction

Now the part that earns the session its place, because a mechanism that only reproduces a number already in hand is worth very little. Does it predict something the alternatives do not?

It does, and it comes for free from the geometry. Our universe drifts in a specific direction — and that direction is not a new free parameter. It is the same axis established in Sessions 6 through 8: the axis of the CMB hemispherical asymmetry, and of the preferred orientation of the largest cosmic-web filaments, both traced to the approach direction of the two parent Summos. The growing drive, and its small shear residual, act along that same axis. So H₀ should be very slightly higher when measured along the CMB asymmetry axis than perpendicular to it — a directional variation in the expansion rate, aligned with an axis the framework already fixed for independent reasons.

Test the form of that, because form is what decides whether it is worth anything. Is it falsifiable? Yes, cleanly. Measure H₀ in independent directions on the sky. If there is no directional variation, or if the variation does not line up with the CMB asymmetry axis within measurement uncertainty, the prediction is dead. It stakes something. Is it discriminating? This is the point. Every internal proposal for the tension is isotropic — a change to the dark-energy sector, to gravity, or to the relativistic particle content applies the same everywhere and predicts no preferred direction, let alone that specific one. The framework predicts a direction, and names it in advance, because it has an external, off-centre cause that already fixed that axis for two other observables. Same axis, now a third observable. That is discriminating in the way a lone anisotropy would not be.

One caution for the record. There are already tentative reports in the literature of a dipole in H₀ — hints that the rate is not perfectly isotropic. Do not enter those as support. A present hint that happens to sit near the prediction is a similarity of thought, not evidence; convergence is not confirmation. The claim stands or falls on dedicated, forward-looking measurement — the direction-dependent expansion work coming from surveys such as DESI, Euclid, and the Roman Space Telescope — not on reading today’s anomalies as if they had already settled it.

What Cannot Be Derived

And the limits, stated plainly rather than papered over, because the last three sessions were spent removing exactly the kind of false precision that would tempt the framework here.

It does not derive the size of the tension — the roughly six kilometres per second per megaparsec between the two measurements. That magnitude depends on the Summo-to-universe distance and the present degree of off-centre drift, and the framework has not defined those parameters. It gives the sign and the time-behaviour — weaker early, stronger now, so the forward extrapolation under-predicts — but not the number.

Nor does it derive the size of the anisotropy. It fixes the axis — the CMB asymmetry direction — and the sign — marginally faster along it — because both follow from the drift geometry already established. It does not fix how large the directional difference in H₀ should be; that too depends on parameters the framework has not pinned down. Axis and sign are predictions. Magnitude is not. Saying so is not a weakness of the account. It is the difference between a mechanism and a fit, and manufacturing the number would be exactly the invented precision the record has been at pains to remove.

So the claim is this: the direction and time-behaviour of the tension from the established drift-drive; a small, axis-aligned, same-sign H₀ anisotropy as the discriminating prediction; and an open admission that neither magnitude is derivable yet. At that size it survives. It rests on an existing commitment rather than a new invention. It does not overturn the cosmic-acceleration work or the Session 8 tidal settlement — it uses the smallness Session 8 established as the reason its prediction is slight. It makes one prediction the field’s isotropic proposals do not. And it is honest about the two numbers it cannot produce. Recorded on those terms.