The Real Reason Galaxies Recede
The observation that most galaxies show redshifted spectra, consistent with recession velocity proportional to distance, is the primary empirical basis for the claim that the universe is expanding. This piece proposes an alternative mechanism for that same observation: gravitational sorting, operating across cosmic timescales, producing a universe in which the surviving population of galaxies is predominantly on non-intersecting trajectories, and therefore predominantly moving apart from one another, without space itself needing to expand at all.
The Solar System Already Shows You This
The solar system provides a directly observable demonstration of the underlying mechanism. Early in solar system formation, planetesimals and proto-planets occupied a wide range of orbital planes and inclinations, moving in every direction. Objects on intersecting orbits collided, merged, were deflected, or were ejected entirely. Over hundreds of millions of years, the surviving objects, the planets observed today, ended up occupying orbits on approximately the same plane, moving in the same direction. Objects on incompatible orbits were eliminated from the observable population, not by any special force, but by ordinary collision and gravitational interaction, given enough time.
A direct proof-of-concept simulation of exactly this mechanism, a three-dimensional N-body simulation of 120 bodies around a fixed central star, implementing only Newtonian gravity, with correct orbital velocities and realistic mass ratios matching real solar systems, produces orbital plane alignment rising from 67% to 84% as retrograde bodies on incompatible orbits are eliminated through genuine three-dimensional collisions. This isn't a toy result tuned to match an expected answer. It's ordinary gravity, run forward, producing exactly the kind of sorting the solar system itself displays.
The Highway, and Why It Looks Like Everything Is Fleeing
A more intuitive version of the same idea: imagine a very wide highway on which vehicles are initially moving in every direction and at every speed, some heading north, some south, some at high speed, some slow, some already on direct collision courses with each other. Let ordinary physics unfold. Vehicles on collision courses collide and are eliminated. Vehicles on near-collision courses swerve and are deflected. Vehicles moving in genuinely incompatible directions at the same location cannot both survive indefinitely.
After enough time, look at what remains. The surviving vehicles are the ones that were never on collision courses with each other, moving in roughly compatible directions, and the faster ones have travelled further from wherever you're standing. An observer at any point on this highway, looking out at the survivors, would see a striking pattern: almost all vehicles are moving away, and the further away a vehicle is, the faster it appears to be receding, because the faster it was moving, the further it has had time to go. This isn't because some mysterious force is pushing every vehicle in the same direction, and it isn't because the highway itself is expanding. It's the straightforward, mechanical result of survival. The incompatible trajectories eliminated themselves over time. What remains looks, from any vantage point at all, exactly like universal recession proportional to distance.
This is Hubble's Law, reread. Not a law of universal expansion, but a law of survival. The universe, under this reading, is the highway after an amount of time current cosmology doesn't contemplate has already passed. The galaxies observed receding are the survivors, the ones whose trajectories were never going to intersect our own galaxy's path. The ones that were going to intersect it already have, across the trillions of years of the pre-ignition era and the billions of years since.
Why It Looks the Same in Every Direction
A serious objection to this mechanism is whether it would produce the isotropic recession pattern actually observed, meaning the same recession relationship in every direction, not just some. Under this framework, that isotropy is a natural, expected consequence of the process operating independently and identically across infinite space. Every region of an infinite universe has undergone the same gravitational sorting process, across the same essentially unlimited timescales. Every observer, located anywhere at all, sees the same end state of that process: a predominantly divergent surviving population. Isotropy is predicted directly, without needing to invoke any special central origin point for the universe as a whole.
The Formula, Derived, Not Assumed
This can be made precise using the mathematics of survival analysis. In an infinite field of galaxies with initially random trajectories, let the probability that any given galaxy pair has not yet collided or merged by time t follow the standard Poisson survival model, meaning that probability decays exponentially with a rate set by the mean collision frequency, which in turn depends on galaxy number density and velocity distribution. As time increases, the surviving population consists increasingly of galaxies on non-intersecting, divergent trajectories. The mean recession velocity of that surviving population turns out to be proportional to distance, because galaxies moving faster have naturally travelled further from wherever their last gravitational interaction occurred.
Working through the mechanics directly: galaxies that have travelled the greatest distance from their last interaction have done so because they've been moving at higher velocities for longer, without a collision to interrupt them. In a population where survival itself is tied to divergence velocity, the expected distance travelled by any surviving galaxy comes out directly proportional to its velocity, distance equals velocity times a mean survival time. Rearranged, that gives velocity equals distance divided by that same mean survival time, which is exactly Hubble's Law, with the Hubble constant equal to one over that mean collision-free survival time. The linear form of the relationship isn't an assumption fed into the model. It falls directly out of the survival selection process itself. Faster galaxies travel further. Observed at any single moment, faster galaxies are further away. The proportionality is a geometric fingerprint left behind by the sorting, not a property of space itself stretching uniformly.
Tested at Small Scale
This mechanism has already been checked through direct simulation. An N-body simulation of 200 galaxies, initialized with random positions and random velocities in a cubic volume, using only Newtonian gravity and momentum-conserving mergers, no expansion term, no dark-energy term, no tuned initial recession field inserted anywhere in the physics, was evolved until the merger rate fell close to zero and the surviving population stabilized. At that point, the correlation between each surviving galaxy's distance and its recession velocity was measured directly: a Pearson correlation coefficient of 0.675, alongside 84% of the surviving galaxies found to be actively receding from the observer point, up from a starting figure of exactly 50% in the initial randomized population. This result has been independently reproduced on Google Colab, and the simulation itself is openly available for anyone to run.
The significance of that result isn't that every observed Hubble data point has thereby been re-derived from scratch. It's that the standard inference, that a Hubble-like recession pattern uniquely proves global metric expansion, loses its monopoly. A Hubble-like statistical relationship can emerge directly from a dynamically filtered survivor population, with no metric expansion required as the only possible cause.
No Universal Brake, So No Deceleration Either
One further question needs addressing directly: does gravitational deceleration, acting over cosmological timescales, introduce curvature into this velocity-distance relationship that shouldn't be there? Under this framework, the answer is no, and this follows as a direct prediction of spatial infinitude, not an extra assumption. In an infinite, isotropic universe with no preferred direction and no boundary, the gravitational pull on any given galaxy from the infinite matter distribution surrounding it in every direction cancels to zero net force at cosmological scale. There is no universal gravitational brake acting on the system as a whole. Individual galaxies experience local interactions, mergers, deflections, accretion, but none of these remove momentum from the system overall. When two galaxies merge, the resulting body simply inherits the combined momentum of both. When a galaxy is deflected, its speed is conserved even as its direction changes. Galactic velocities, at universal scale, can only be maintained or increased over time under this mechanism, never decreased, which guarantees that the linear velocity-distance relationship is the permanent geometric signature of the sorted population, not an approximation that degrades as time passes. As mergers consolidate momentum into fewer, faster surviving bodies over time, the velocity distribution of the sorted population may even shift toward higher values, producing an apparent acceleration in the recession data that requires no dark energy to explain at all.
This mechanism also offers a direct account of the Hubble Tension addressed in Papers One and Sixty. If the proportionality between recession velocity and distance is an emergent statistical property of a sorted population, instead of a fundamental constant of spacetime itself, then different measurement methodologies, probing different scales, different galaxy populations, and different epochs of that sorting process, would naturally be expected to yield slightly different values, precisely the pattern actually observed across the supernova ladder, the Cosmic Microwave Background, and gravitational lensing time-delay measurements. Under the standard model, that disagreement is a tension requiring resolution, evidence that something is being measured incorrectly somewhere. Under gravitational sorting, some of that disagreement is exactly what the mechanism predicts, since there was never a single, universal constant for every method to agree on in the first place.
All DOIs linked below.